IO L

PO W ER
KEN N ETH J. H O FFER

SLACK Incorporated
 Ke nne th J. Ho ffe r, MD
Clin ical Professor of Ophthalmology, Ju les Stein Eye Institute
Un iversity of Californ ia, Los Angeles
Fou nd ing President, American Society of Cataract and Refractive Su rgery
Fou nd ing Ed itor, Jou rnal of Cataract & Refractive Su rgery
Past-President, IOL Power Clu b (2005-2007)
St. Mary’s Eye Center
Santa Mon ica, CA
www.slackbooks.com
ISBN : 978-1-55642-988-0

Copyright © 2011 by SLACK Incorp orated

All rights reserved . No part of th is book m ay be reprodu ced, stored in a retrieval system or
tran sm itted in any form or by any m ean s, electron ic, m echan ical, photocopying, record ing
or otherw ise, w ithout w ritten p erm ission from the pu blisher, except for brief quotation s
em bod ied in critical articles and review s.

The procedu res and practices d escribed in th is book shou ld be im plem ented in a m an ner
con sistent w ith the professional stand ard s set for the circu m stances that ap ply in each sp ecif-
ic situ ation. Every effort h as been m ad e to con firm the accu racy of the in form ation presented
and to correctly relate generally accep ted practices. The authors, ed itor, and pu blisher can not
accept resp on sibility for errors or exclu sion s or for the outcom e of the m aterial presented
herein. There is no expressed or implied w arranty of th is book or in form ation imparted by
it. Care h as been taken to en su re th at d rug selection and dosages are in accord ance w ith
cu rrently accepted/ recom m end ed practice. Du e to continu ing research, ch anges in govern-
m ent p olicy and regu lation s, and variou s effects of d rug reaction s and interaction s, it is
recom m end ed that the read er carefu lly review all m aterials and literatu re provided for each
d rug, esp ecially those that are new or not frequ ently u sed. Any review or m ention of sp ecific
compan ies or products is not intended as an endorsem ent by the author or pu blisher.

SLACK Incorp orated u ses a review process to evalu ate subm itted m aterial. Prior to pu blica-
tion, edu cators or clin ician s provid e important feed back on the content that we pu blish. We
welcom e feed back on th is work.

Pu blished by: SLACK Incorp orated

6900 Grove Road
Thorofare, N J 08086 USA
Telephone: 856-848-1000
Fax: 856-848-6091
w w w.slackbooks.com

Contact SLACK Incorp orated for more in form ation about other books in th is field or about
the availability of ou r books from d istributors outside the Un ited States.

Hoffer, Ken neth J.

IOL p ower / Ken neth J. Hoffer.
p. ; cm .
Intraocu lar len s p ower
Includ es bibliograph ical references and ind ex.
ISBN 978-1-55642-988-0 (alk. pap er)
1. Intraocu lar len ses. I. Title. II. Title: Intraocu lar len s p ower.
[DN LM: 1. Len ses, Intraocu lar. 2. Biometry. WW 358]
RE988.H64 2011
617.7’524--d c22
2010053713

For p erm ission to reprint m aterial in another pu blication, contact SLACK Incorp orated .
Authorization to photocopy item s for internal, p ersonal, or acad em ic u se is granted by
SLACK Incorp orated provid ed that the appropriate fee is paid d irectly to Copyright Clearance
Center. Prior to photocopying item s, please contact the Copyright Clearance Center at 222
Rosewood Drive, Danvers, MA 01923 USA; phone: 978-750-8400; web site: w w w.copyright.
com ; em ail: in [email protected]
 De dicatio n
I ded icate th is book to my w ife Marcia (Figu re D-1), who has stood beside m e
for these 40 years wh ile all th is cataract/ IOL and IOL power h istory took place.
I also ded icate it to my preciou s grandch ild ren: Kaylin (Figu re D-2), Anabella
(Figu re D-3), Wesley (Figu re D-4), and Erik Hoffer (Figu re D-5). May they forgive
m e for the time I took from being w ith them to work on th is book.

Ken neth J. Hoffer, MD

Figure D-1. Marcia. Figure D-2. Kaylin. Figure D-3. Anabella.

Figure D-4. Wesley. Figure D-5. Erik.

 Co nte nts
Dedication ................................................................................................................................v
A cknowledgments .................................................................................................................. xi
A bout the Author .................................................................................................................. xii
Contributing Authors ............................................................................................................xv
Foreword by Emmanuel Rosen, BSc, M D, FRCSEd, FRCOphth, FRPS ........................... xix
Foreword by Bradley R. Straatsma, M D, JD ...................................................................... xxi

S ECTION I BASICS AN D ACCURATE BIOMETRY ....................................................... 1

Chapter 1 Section I: Introduction ................................................................................... 3
Kenneth J. Hoffer, M D
Chapter 2 Axial Length: Ultrasou nd : Backgrou nd ..................................................... 9
Kenneth J. Hoffer, M D
Chapter 3 Im mersion Versu s Applanation Ultrasou nd ............................................ 13
Kenneth J. Hoffer, M D
Chapter 4 Ultrasou nd Velocities for Axial Length Measu rement .......................... 17
Kenneth J. Hoffer, M D
Chapter 5 A-Scan Biometry: Im m ersion Tech n ique .................................................. 25
H. John Shammas, M D
Chapter 6 A-Scan Biom etry: Avoid ing Pitfalls With Im mersion Tech n ique ......... 29
Karl Ossoinig, M D
Chapter 7 Available A-Scan Instru mentation ............................................................ 39
Kenneth J. Hoffer, M D
Chapter 8 Im mersion Using the Prager Shell ............................................................. 53
Thomas C. Prager, PhD, M PH
Chapter 9 Axial Length: Laser Interferometry: Basics of the IOLMaster .............. 63
Wolfgang Haigis, M S, PhD
Chapter 10 IOLMaster Exam ination .............................................................................. 67
Wolfgang Haigis, M S, PhD
Chapter 11 IOLMaster in Difficu lt Eyes ........................................................................ 71
Wolfgang Haigis, M S, PhD
Chapter 12 Axial Length: Laser Interferometry: The LENSTAR LS900
Instru ment ..................................................................................................... 75
Kenneth J. Hoffer, M D; H. John Shammas, M D; and Jaime A ramberri, M D
Chapter 13 Corneal Power: Diopters Versu s Rad iu s ................................................... 89
Wolfgang Haigis, M S, PhD
Chapter 14 Corneal Power: Manu al Keratometry and In stru mentation ................. 93
Kenneth J. Hoffer, M D
Chapter 15 Autom ated Keratometry for IOL Power Calcu lation .............................. 97
Jaime A ramberri, M D
viii Contents

Chapter 16 Corneal Power: Corneal Topography for IOL Power Calcu lation ...... 107
Jaime A ramberri, M D
Chapter 17 Corneal Power: Measu ring Corneal Power With the Pentacam ..........115
Giacomo Savini, M D
Chapter 18 IOL Position : ACD and ELP .......................................................................119
Kenneth J. Hoffer, M D
Chapter 19 IOL Position : Measu ring the ACD by Optical Pachymetry ................. 123
Kenneth J. Hoffer, M D
Chapter 20 IOL Position : Double-AL Method for Scleral Buckle Eyes ................... 131
Kenneth J. Hoffer, M D

S ECTION II FORMULAS AN D SPECIAL CIRCUMSTANCES................................ 133

Chapter 21 Section II: Introduction .............................................................................. 135
Kenneth J. Hoffer, M D
Chapter 22 Formu las and Program s: Formu la H istory and Basics ........................ 137
Kenneth J. Hoffer, M D
Chapter 23 Formu las and Program s: Regression and Theoretic Formu las .......... 143
Kenneth J. Hoffer, M D
Chapter 24 Formu las and Program s: Olsen Formu la ............................................... 149
Thomas Olsen, M D
Chapter 25 Formu las and Program s: Accessing Modern IOL Power Formu las ... 155
Kenneth J. Hoffer, M D
Chapter 26 Formu las and Program s: Formu la Personalization .............................. 163
Kenneth J. Hoffer, M D
Chapter 27 Special Circu m stances: AL Measu rem ent in Staphylom a Eyes ...........167
H. John Shammas, M D
Chapter 28 Sp ecial Circu m stances: Silicone Oil–Filled Eyes ................................... 169
Wolfgang Haigis, M S, PhD
Chapter 29 Special Circu m stances: Un ilateral H igh Myop es and Hyperopes ..... 171
Kenneth J. Hoffer, M D
Chapter 30 Special Circu m stances: Penetrating Keratoplasty and
Scarred Corneas .......................................................................................... 173
Kenneth J. Hoffer, M D
Chapter 31 Sp ecial Circu m stances: Rad ial Keratotomy Eyes .................................. 175
Giacomo Savini, M D
Chapter 32 Sp ecial Circu m stances: Post-Laser Refractive Su rgery Eyes ............... 179
Kenneth J. Hoffer, M D
Chapter 33 Sp ecial Circu m stances: H aigis-L IOL Formu la ..................................... 195
Wolfgang Haigis, M S, PhD
 Con ten ts ix

Chapter 34 Special Circu m stances: Dou ble-K Method ............................................ 199

Jaime A ramberri, M D
Chapter 35 Sp ecial Circu m stances: In fluence of Spherical Aberration
on IOL Power .............................................................................................. 207
Sverker N orrby, PhD
Chapter 36 Sp ecial Circu m stances: Mu ltifocals and Toric IOLs ..............................211
John M oran, M D, PhD
Chapter 37 Special Circu m stances: Ped iatric Eyes .................................................... 215
Scott K. M cClatchey, CA PT, M C, USN, M D
Chapter 38 Sp ecial Circu m stances: Piggyback IOLs................................................. 219
Kenneth J. Hoffer, M D
Chapter 39 Special Circu m stances: Silicone Oil Power ............................................ 221
Kenneth J. Hoffer, M D
Chapter 40 Special Circu m stances: Effect of IOL Tilt on Astigm atism ................. 223
Susana M arcos, PhD, FOSA , FEOS
Chapter 41 Sp ecial Circu m stances: An iseikon ia and An isom etropia ................... 231
Kenneth J. Hoffer, M D
Chapter 42 Preventing IOL Power Errors ................................................................... 233
Kenneth J. Hoffer, M D
Chapter 43 Diagnosing and Treating IOL Power Errors .......................................... 237
Kenneth J. Hoffer, M D
Chapter 44 Futu re Directions in IOL Power Calcu lation:
Intraoperative Refractive Biometry ......................................................... 241
Tsontcho [Sean] Ianchulev, M D, M PH

A ppendix A : IOL Power Club ............................................................................................ 247

Financial Disclosures .......................................................................................................... 249
 Ackno w le dg m e nts
I first w ish to acknowledge the pioneering work in IOL p ower calcu lation done by
Rob Van der Heidje, PhD (Fig. A-1) from Am sterd am. H is original formu la, nomo-
gram (Fig. A-2), and d isk calcu lation device (Fig. A-3) (Biometer, Med ical Workshop,
Gron ingen, Netherland s) was the earliest to attempt to m ake it easier for su rgeons
to select the prop er lens power.

Figure A-1. Rob G.

L. Van der Heidje, Figure A-3. 1975 Van der
PhD Heidje Biometer using 3.5
mm as a standard ELP.

Figure A-2. 1975 Van der Heidje

Nomogram using 3.5 mm as a
standard ELP.

I mu st personally acknowledge the m any people who have help ed me over these
37 years, but it is imp ossible to note them all. I first note the original encou rage-
m ent of my early (1972-1977) practice partners, Joh n E. Gilmore, MD (deceased) and
Donald E. Dickerson, MD; as well as the tireless effort and devotion of my nu rses,
Patricia Th irsk, RN and Florence Brau n, RN in helping to assu re accu racy and p er-
fection in the care of my patients and d ata collection in my research. It wou ld have
been d ifficu lt in those early years w ithout the support of the UCLA Santa Mon ica
Hospital in giving me qu arters for the Eye Lab and provid ing staff, equ ipment, and
m aterials to start IOL power calcu lation s the right way. In th is endeavor, I have to
than k my first A-scan tech n ician, Mr. Don Allen (deceased), who proved tech n ical
p erson nel cou ld do as good a job as a physician. I also than k my m any other tech n i-
cians th rough the years who were originally trained by Mr. Allen.
On a personal note, I w ish to than k Joh n Sham m as, MD and Wolfgang H aigis,
PhD for their m any years of friend sh ip and support; as well as Jack Hollad ay, MD, a
you ng m an in the 70s who help ed me prove that theoretic formu las were better than
regression and then for keeping me on my toes.
Finally I acknowledge the patience and acceptance of my w ife Marcia and ou r
ch ild ren, Kevin, Jeffrey, and Kristin, who sacrificed time w ith me wh ile I spent much
p ersonal time on these endeavors.
 Abo ut the Autho r
Kenneth J. Hoffer, M D was born in New York City; grew up in upstate N Y; and
received h is education at Siena College, Loudonville, N Y, SUN Y Upstate Med ical
Un iversity in Syracu se, N Y, UCLA Santa Mon ica Hospital, and Wayne State
University Kresge Eye In stitute in Detroit, MI. He im med iately set up practice in
Santa Mon ica, CA in Ju ly 1972.
H is interest in cataract su rgery began early in h is clin ical career by begin n ing
phacoemu lsification after h is first two months of practice, and IOL implantation two
years later in 1974. Being in an extremely comp etitive environ ment, and to create
superior patient outcom es, it stimu lated h im to introduce u ltrasou nd axial length
IOL power calcu lation. He created the American Intra-Ocu lar Implant Society (now
ASCRS) and fou nded the Journal of Cataract & Refractive Surgery (previou sly AIOISJ).
He w rote h is original Hoffer formu la in 1974 and, based on a pachymetry study in
1982, proposed the first pred iction of ELP based on the eye’s axial length. The Hoffer
Q formu la followed in 1993, u sing a tangent of K as an add itional ELP pred ictor. To
ease the calcu lation of formu las he first u sed early program m able calcu lators (H P-
65), and in 1993 produced the first program for IOL power calcu lation for personal
computers (Hoffer Program s®). He also participated in the fou nd ing of the IOL
Power Clu b in 2005 that has help ed stimu late h is interest in th is su bject.
After m any years of teach ing cou rses on the su bject of IOL power at the Santa
Mon ica Hospital IOL cou rses (2800 su rgeon s attended over 10 years); at m ajor m eet-
ings such as AAO, ASCRS, and ESCRS; perform ing research in th is field ; and pub-
lishing papers and book chapters, he was finally stimu lated to w rite h is first book
on the subject to wh ich he has devoted h is professional career. Th is book is based
prim arily on the cou rses he has taught w ith the add itional help of m any friend s in
the field.
 Co ntributing Autho rs
Jaime A ramberri, M D
BEGITEK Clín ica Oftalmológica
OKULAR Clín ica Oftalmológica
San Sebastián, Spain
Vitoria-Gasteiz, Spain

Wolfgang Haigis, M S, PhD

Professor of Ophthalm ic Biometry
Un iversity of Wü rzbu rg
Dept. of Ophthalmology
Laboratory for Biometry
Wü rzbu rg, Germ any

Kenneth J. Hoffer, M D
Clin ical Professor of Ophthalmology, Ju les Stein Eye In stitute
Un iversity of Californ ia, Los Angeles
Fou nd ing President, American Society of Cataract and Refractive
Su rgery
Fou nd ing Ed itor, Jou rnal of Cataract & Refractive Su rgery
Past-President, IOL Power Clu b (2005-2007)
St. Mary’s Eye Center
Santa Mon ica, CA

Tsontcho [Sean] Ianchulev, M D, M PH

Clin ical Assistant Professor
Un iversity of Californ ia, San Francisco
San Francisco, Californ ia

Susana M arcos, PhD, FOSA , FEOS

Professor of Research
Instituto de Óptica “Daza de Valdés”
Mad rid, Spain
xvi Contributing Auth ors

Scott K. M cClatchey, CA PT, M C, USN, M D

Adju nct Associate Professor of Ophthalmology, Lom a Lind a
Un iversity
Assistant Professor, Un iform ed Un iversity of the Health Sciences
Navy Med ical Center
San Diego, CA

John R. M oran, PhD, M D

President/ CEO
Moran Research and Consu lting, Inc.
Hou ston, TX

Sverker N orrby, PhD

Vice-President, IOL Power Clu b (2009-2011)
Netherland s

Thomas Olsen, M D
Assistant Professor, d r. med.
Un iversity Eye Clin ic
Århu s Hospital, Den m ark
President, Scand inavian Society of Cataract and Refractive Su rgery
(SSCRS)
Board Mem ber, Eu ropean Society of Cataract & Refractive Su rgery
(ESCRS)
Secretary, IOL Power Clu b (2005-2011)
Den m ark

Karl Ossoinig, M D
Honorary President of SIDUO
Professor Emeritu s
Department of Ophthalmology
Un iversity of Iowa
Iowa City, IA
 Con tributin g Auth ors xvii

Thomas C. Prager, PhD, M PH

Vale Asche Ru ssell Professor of Ophthalmology
Un iversity of Texas Med ical School
Hou ston, TX
Clinical Professor, Department of Ophthalmology and Visual Science
Un iversity of Texas Med ical Branch, Galveston
Clin ical Professor, Department of Ophthalmology
Weil Cornell School of Med icine
Hou ston, TX

Giacomo Savini, M D
Stud io Ocu listico d’Azeglio
Mem ber, IOL Power Club
Bologna, Italy

H. John Shammas, M D
Clin ical Professor of Ophthalmology
The Keck School of Med icine at USC
Los Angeles, CA
Med ical Direcor
Sham m as Eye Med ical Center
Lynwood, CA
 Fo re w o rd
Ken Hoffer is pre-em inent in the field of ocu lar biometry. After hu nd red s of
p eer-reviewed presentations and contributions to the ophthalm ic literatu re, over 50
chapters in other textbooks, he has at last produced h is ow n authored textbook, IOL
Power, for what I am su re w ill be en lighten ing for all ophthalmologists engaged in
lens and refractive su rgery.
H is academ ic affiliations attest to h is peer recogn ition and h is pleasu re in
sharing h is knowledge as a gifted teacher. He has represented the Department
of Ophthalmology and the Ju les Stein Eye Institute, and Clin ical Facu lty, at the
University of Californ ia in Los Angeles as a Clin ical In structor from 1976 to 1979,
Assistant Clin ical Professor from 1979 to 1983, Associate Clin ical Professor from 1983
to 1990, and Clin ical Professor from 1990 to the present tim e.
H is contributions to ophthalmology have been more w ide-ranging than h is very
well-know n and heavily utilized Hoffer IOL Power formu lae, as he has been a con-
stant in novator in aspects of cataract su rgery instru mentation and refractive su rgery
tech n iques.
Ken Hoffer was the fou nd ing president of ASCRS from 1974 to 1975 and fou nd ing
ed itor of the Journal of Cataract & Refractive Surgery in 1975. I have particu lar reason to
enjoy Ken’s continu ing contributions to JCRS as author and peer reviewer where he
never refu ses a request to review and always produces thorough and expert review s
for gu id ance of the ed itors.
Since Ken introdu ced h is “Intraocu lar Len s Power
Calcu lation” lectu re at the first Intraocu lar Lens Symp osiu m
held in the USA at Long Beach Memorial Hospital in Long
Beach, CA on Novem ber 16, 1974, he has been prolific in
design ing and publish ing h is formu lae and advice for the
betterm ent of ophthalm ic patient care. In so doing, he has
attracted the than ks of cou ntless ophthalmologists world-
w ide.
I w ish Ken the success that h is endeavor, gen iu s, and
productivity deserve w ith success of th is volu m e.

Professor Emanuel Rosen, BSc, M D, FRCSEd, FRCOphth, FRPS

Director, Rosen Eye A ssociates
Co-Editor, Journal of Cataract and Refractive Surgery
Past-President, European Society of Cataract & Refractive Surgeons (ESCRS)
M anchester, UK
 Fo re w o rd
Ken neth J. Hoffer, MD, and a group of internationally recogn ized exp erts, bring
together the physical sciences, m athem atical formu las, and clin ical experience
requ ired to calcu late the intraocu lar lens power for cataract/ intraocu lar lens su r-
gery and refractive phakic/ intraocu lar lens su rgery. As a service to readers, the
authors provide h istorical perspective, hone the m essage to essentials, and ad here
to Albert Einstein’s d ictu m that “Everyth ing shou ld be as simple as possible, but not
simpler.”
System atically, the chapters progress from biometry w ith focu s on axial length,
corneal power, and intraocu lar lens p osition; to formu las for power calcu lation
includ ing formu la selection and personalization; and on to clin ical factors such as
patient preference, concu rrent ocu lar cond itions, and “intraocu lar len s su rprises.”
Each segment is presented authoritatively w ith the strength of extensive exp eri-
ence.
The su bject of preoperative intraocu lar lens p ower calcu lation is more important
than ever before, becau se of the requ irement for accu racy related to mu ltifocal,
accom mod ative, and toric intraocu lar len ses. To th is mu st be added the complexities
of intraocu lar lens power calcu lation after corneal refractive
su rgery and the increased expectations of patients.
As a basic text and a reference for u se in special circu m-
stances, IOL Power is of d istinct value to ophthalmologists,
para-ophthalm ic person nel, and m anu factu rers of intraocu lar
lens devices. Ad m irably, Dr. Hoffer and co-authors succeed in
the stated objective of provid ing in form ation to improve the
accu racy of intraocu lar lens power calcu lations in ophthalm ic
su rgery.

Bradley R. Straatsma, M D, JD
Chairman Emeritus
Jules Stein Eye Institute
University of California, Los A ngeles
 Bas ics a n d
Accu ra te Bio m e try
I
 1

Se ctio n I:
In tro d u ct io n
Kenneth J. Hoffer, M D (Fig. 1-1)

T
h is book has been w ritten to provide all the latest
in form ation available regard ing the calcu lation of
intraocu lar len s (IOL) power. It is form atted along the
lines of the IOL power cou rses I have taught at the an nu al
m eetings of the Am erican Acad emy of Ophthalmology
(AAO) and the Am erican Society of Cataract & Refractive
Su rgery (ASCRS) over the past 36 years (Fig. 1-2). After years
of presenting these cou rses, it seemed th is was a good tim e
to put th is m aterial into a textbook to increase the availability
of the in form ation to everyone.
When I performed the first u ltrasou nd A-scan for IOL
p ower in the Un ited States in 1974, I also w rote the original
Hoffer formu la. I was u nable to get it pu blished in American
jou rnals becau se it was concern ing IOLs and I was ju st two
Figure 1-1. Kenneth J. Hoffer, years out of my residency and completely u n know n. The
MD first time I was able to publish anyth ing was in 19751 (Fig. 1-
3), after I started the American Intra-Ocu lar Implant Society
(AIOISJ), wh ich is now the Journal of Cataract & Refractive Surgery (JCRS). It took me 7
years to finally get the Hoffer formu la 2 pu blished in 1981. A sim ilar th ing happ ened
when I perform ed a study that showed that the u ltim ate position of the IOL was d irectly
prop ortional to the axial length (AL) of the eye 3 (Fig. 1-4). Th is led me to propose a
3
Hoffer KJ. IOL Power (pp. 3-8).
© 2011 SLACK Incorporated.
4 Ch apter 1

Figure 1-2. Title slide for IO L Power courses given

at AAO and ASCRS since 1975.

Figure 1-3. First published paper on IO L power

calculation in the US in AIO ISJ, 1975.

A B

Figure 1-4. First published paper establishing

direct relationship between IO L position and
the AL of the eye, 1975.
 Section I: In troduction 5

Figure 1-5. Collage of textbooks containing chapters on IO L

power by the author.

Figure 1-6. Kwitko-Kelman book con-

taining chapter on the history of IO L
power calculation in North America.

Figure 1-7. IO L Power Club formed in San Sebastian, Spain

on September 5, 2005.

formu la to better pred ict the IOL position u sing AL. Ju st as I was preparing to su bm it
it for publication, the Chairm an of my residency department at Kresge Eye Institute
strongly requested me to su bm it som eth ing to th is new, short-lived jou rnal he had been
asked to ed it. I thu s subm itted it to h im and it was pu blished in a short-lived jou rnal that
is d ifficu lt to reference.
Over the past 35 years, I have been asked to w rite a textbook but have never been
able to get to it. On the other hand, I have w ritten chapters on th is subject in more than
50 textbooks of others (Fig. 1-5). I was asked to w rite the h istory of IOL power calcu lation
in North America for the book on the h istory of modern cataract su rgery by Kw itko and
Kelm an (Fig. 1-6).
I w ish to than k my colleagues, the fellow mem bers in the IOL Power Club (IPC) (Fig. 1-
7), and others who have graciou sly taken their tim e to m ake contributions based on their
6 Ch apter 1

A B

Figure 1-8. (A) Sir Harold Ridley, MD, London,

UK. (B) Eye with a Ridley implant with UCVA of
20 /20, implanted by Ridley in 1952, photographed in 1979 (photo by Hoffer). (C) The author (center)
with Lady Elizabeth and Sir Harold Ridley (right) in 1999. (Photo by Marcia Hoffer.)

exp ertise. The IPC was formed in 2005 in San Sebastian, Spain to foster cam araderie, com-
mu n ication, and scientific interaction in th is subject by those who have labored in th is
field for m any years. The IPC was fou nded by Jaime Aram berri of San Sebastian, Spain;
Wolfgang H aigis, PhD, of Wü rzbu rg, Germ any; Sverker Norrby, PhD, of Gron ingen,
Holland ; Thom as Olsen, MD, of Århu s, Den m ark; H. Joh n Sham m as, MD of Lynwood,
CA, USA; and myself. The joint efforts of these pillars of th is field have led to new com-
mu n ication and m any new ideas.
Since Sir H arold Rid ley (Fig. 1-8A and C) experienced a 21 d iopter (D) “su rprise” in
lens p ower calcu lation on h is first two cases in 1949-1950, we have been seeking ways to
calcu late intraocu lar len s (IOL) power w ith greater accu racy. Fig. 1-8B (photo by author
in 1979) proves he u ltim ately worked th is out, since th is is an eye w ith a Rid ley posterior
cham ber (PC) implant w ith u ncorrected visu al acu ity (UCVA) of 20/ 20, implanted by
Rid ley in 1952.
Th is science is rather d ry and does not stimu late great interest on the part of the m ajor-
ity of cataract su rgeons. To m ake the su bject more interesting it m ight be advantageou s
to break it dow n into its comp onent parts. The th ree m ajor components of IOL power
calcu lation are Biometry, Formu las, and Clinical Variables. Biometry can be d ivided into
its components needed to calcu late IOL power: the A L, the Corneal Power (K), and the
IOL Position (ELP). The su bject of Formu las involves their Generations, their Usage, and
their Personalization. Clin ical Variables deals w ith the su bjects of Patient N eeds and Desires,
Special Circumstances, and Problems and Errors. Th is book w ill present all the in form ation
necessary to fu lly u nderstand all the aspects of calcu lating IOL powers.
 Section I: In troduction 7

l Biometry
© Axial length (AL)
© Corneal p ower (K)
© IOL position (ELP)
l Formu las
© Generation s
© Usage
© Personalization
l Clin ical Variables
© Patient Need s and Desires
© Sp ecial Circu m stances
© Problem s and Errors

We recom mend for references in th is su bject the textbooks by H. Joh n Sham m as 4 and
Sand ra Frazier-Byrne.5

Re fe re nce s
1. Hoffer KJ. Mathem atics and computers in intraocu lar len s calcu lation. A m Intra-Ocular
Implant Soc J. 1975;1(1):3.
2. Hoffer KJ. Intraocu lar len s calcu lation: The problem of the short eye {Hoffer formu la}.
Ophthalmic Surgery. 1981;12:269-272.
3. Hoffer KJ. The effect of axial length on p osterior cham ber len ses and p osterior cap su le
p osition. Current Concepts in Ophthalmic Surgery. 1984;1:20-22.
4. Sham m as H J, ed . Intraocular Lens Power Calculations. Thorofare, N J: SLACK Incorporated ;
2003.
5. Byrne SF. A -scan A xial Eye Length M easurements. Mars H ill, NC: Grove Park Pu blishers;
1995.
 2
Ax ia l Le n g t h :
Ultra so u n d
Backg rou n d
Kenneth J. Hoffer, M D

W
hen the hu m an lens is replaced w ith an IOL (Fig. 2-1), the optical situation
becomes a two-lens system (cornea and IOL) projecting an im age onto the retina
(m acu la). The d istance between the two lenses (Estim ated Lens Position, or ELP)
affects the refraction, as does the d istance between the two-lens system and the m acu la (Y).
ELP is defined as the d istance from the anterior su rface (vertex) of the cornea to the effective
principle plane of the IOL in the visual axis. Y is defined as the d istance from that principle
plane of the IOL to the photoreceptors of the m acu la in the visual axis. It is easy to see that
ELP + Y is equal to the visual axis AL of the eye. Therefore, know ing the ELP and the AL
w ill allow the calcu lation of Y (Y = AL – ELP).
Also to calcu late the IOL power (P), we mu st know the optical effective power of the
cornea (K) as well as the vergence of the light rays entering the cornea (refractive error,
or R). For em metropia, R is zero. The relation sh ip of the factors P, AL, K, ELP, and R are
such that a formu la can be w ritten to describe it. The first such formu la for IOL power
was published in Ru ssia by Fyodorov and Kolon ko 2 in 1967. Know ing the values of any
fou r of these values w ill allow for the calcu lation of the 5th .
Originally it was thought that it was good enough to get the patient back to the refractive
error they had prior to the cataract. It became standard to use the same power for all eyes
to yield this resu lt. As lens implantation became more popu lar in the US, a standard 18.0 D
Binkhorst prepupillary lens was used. Some thought that they cou ld adjust this power up or
dow n depend ing upon the previous refractive error, and charts were developed to aim clos-
er to em metropia. The first IOL formu las became available in the late 1960s and early 1970s
but they requ ired the need to measu re the AL of the eye and the power of the cornea (K).
9
Hoffer KJ. IOL Power (pp. 9 -12).
© 2011 SLACK Incorporated.
10 Ch apter 2

Figure 2-1. Basics of all formula’s calculation of

IO L power.

Figure 2-2. Jan Worst MD, Holland.

Figure 2-3. Karl O ssoinig, MD, Iowa,

the father of accurate immersion A-scan
ultrasound for axial length measurement.

Early attempts were m ade to measu re the AL u sing u ltrasou nd in Ru ssia and Holland.
In the early 1970s, Jan Worst of Gron ingen, Holland (Fig. 2-2) u sed a sm all A-scan u ltra-
sou nd w ith the Colenbrander 1 formu la that was also u sed by several of the early Dutch
IOL implanters. I becam e aware of th is and, having knowledge of the work done by
Ossoin ig 2 (Fig. 2-3) to develop the Kretz 7200MA A-scan u ltrasou nd in Vien na, Au stria,
decided to u se th is instru ment to perform the first such AL m easu rements in the US
in April 1974.3 Ossoin ig had improved the accu racy of AL m easu rement w ith a precise
im m ersion cup tech n ique and a specific calibration of the instru ment.
Becau se the tech n ique requ ired photograph ing the screen and m aking m easu rements
of the Polaroid photograph s w ith precise calip ers, the process was qu ite ted iou s. I trained
the first Am erican A-scan tech n ician (Donald Allen) and then performed a study that
showed that tech n ician accu racy was equ al to that of a physician.
I then set out to persu ade Sonom etrics, Inc (Boston, MA) to produce an A-scan that
was specific for IOL power and produce an autom atic readout of the AL. In 1975, they
produced the Sonometrics DBR-100 (Fig. 2-4), wh ich u sed an applanation cone applied
d irectly to the cornea and a fixation light as I had d irected. The instru ment read out the
AL d igitally w ithout need of photos and calipers. Un fortu nately, though it became p opu-
 Axial Len g th : Ultrasoun d : Backg roun d 11

Figure 2-4. O riginal 1975 Sonometrics

DBR-100 A-scan ultrasound using the con-
tact applanation method.

lar th roughout the world, it u sed an applanation contact method, later proven by u s to be
in ferior to im m ersion.
I compared the resu lts of th is instru ment w ith the stand ard im m ersion tech n ique and
d iscovered that the applanation tech n ique led to an u npred ictable shorten ing of the read-
ing from 0.25 to 0.33 m m, wh ich cou ld not be offset by a correction factor. After in form ing
others of th is fact,4 Sham m as 5 and others 6 repeated my study and valid ated th is find ing
w ith sim ilar resu lts. In the m eantim e, m any adopted the in stru m ent and applanation
tech n ique and it became the stand ard in the US, as it was con sidered much easier than
the Kretz u n it and the Ossoin ig im mersion tech n ique.
Fyodorov 7 published the first IOL power formu la in 1967. I8 developed a formu la, based
upon the Colenbrander formu la, in 1974 and u sed it w ith the im mersion tech n ique. The
accu racy of the resu lts led to a stand ard of aim ing for em m etropia in all cataract patients.
Cou rses and lectu res over the ensu ing years gradu ally d eveloped a stand ard of not u sing
ju st one IOL power for all eyes, but rather ind ividu ally calcu lating an IOL for each patient.
Ultrasou nd instru ments were continuou sly develop ed over the next 20 years. Everyth ing
changed in 1999 when Zeiss (Jena, Germ any) introduced the first instru ment that u ses a
laser to m easu re the AL. The IOLMaster was specifically set so that it m atched the resu lts
w ith the im m ersion u ltrasou nd tech n ique and not applanation. Th is in stru m ent has now
been show n to be easier to u se, more accu rate, and more reproducible than any other
instru ment or tech n ique. In 2009, H aag-Streit (Koen iz, Sw itzerland) produced a sim ilar
u nit they call the LEN STAR LS900.
In the 1980s, becau se of in accu racies du e to the ap planation m ethod, m any were
looking for w ays to im prove IOL calcu lation. Th is led to the d evelopm ent of regression
formu las, w h ich soon coalesced into the SRK formu la.9 Becau se it w as so easy to u se, it
becam e rapid ly p opu lar worldw id e. I first showed that it w as the lead ing cau se of IOL
rem ovals becau se of IOL p ower error.7 H ollad ay’s introdu ction of h is formu la 10 th at
im proved the pred iction of the ELP in 1990 help ed lead m any back to the u se of theoret-
ic formu las. Th is led to the d evelopm ent of the SRK/ T11 (the “T” stand s for Theoretic)
formu la by Retzlaff, the H offer Q 12 formu la in the early 1990s, and the H aigis13 formu la
in 2000.
Modern IOL p ower calcu lation has been the product of m any sm all steps and a few
step s backward s (regression formu las), but tod ay we can assu re patients that their chance
of getting an improp er IOL power is qu ite sm all.
12 Ch apter 2

Re fe re nce s
1. Colenbrand er MC. Calcu lation of the p ower of an iris clip len s for d istance vision. Br J
Ophthalmol. 1973;57(10):735–740.
2. Ossoin ig KC. Stand ard ized echography: Basic principles, clin ical application s, and
resu lts. Int Ophthalmol Clin. 1979;19(4):127–210. Review. No abstract available.
3. Hoffer KJ. The h istory of IOL p ower calcu lation in North America. In : Kw itko ML,
Kelm an CD, ed s. The History of M odern Cataract Surgery. The H ague, Netherland s: Kugler
Pu blication s; 1998;193–208.
4. Comparison of Kretz 7200 MA, Sonom etrics DBR, and Storz Ocu lom eter for Len s
Calcu lation, First U.S. Intraocu lar Len s Symp osiu m . Am erican Intra-Ocu lar Implant
Society (AIOIS), Los Angeles, CA, March 1978.
5. Sham m as H J. A comparison of im m ersion and contact tech n iques for axial length m ea-
su rem ents. J A m Intraocul Implant Soc. 1984;10(4):444–447.
6. Schelen z J, Kam m an n J. Comparison of contact and im m ersion tech n iques for axial m ea-
su rem ent and implant p ower calcu lation. J Cataract Refract Surg. 1989;15(4):425–428.
7. Fyodorov SN, Kolon ko AI. Estim ation of optical p ower of the intraocu lar len s. Vestn ik
Oftalmologic (Moscow). 1967;4:27.
8. Hoffer KJ. Intraocu lar len s calcu lation: The problem of the short eye. Ophthalmic Surg.
1981;12(4):269–272.
9. Sand ers D, Retzlaff J, Kraff M, et al. Comparison of the accu racy of the Bin khorst,
Colenbrand er, and SRK implant p ower pred iction formu las. J A m Intraocul Implant Soc.
1981;7(4):337–340.
10. Hollad ay JT, Prager TC, Chand ler TY, et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–24.
11. Retzlaff J, Sand ers DR, Kraff MC. Developm ent of the SRK/ T intraocu lar len s implant
p ower calcu lation formu la. J Cataract Refract Surg. 1990;16(3):333–340. Erratu m in :
1990;16(4):528.
12. Hoffer KJ. The Hoffer Q formu la: A comparison of theoretic and regression formu las.
J Cataract Refract Surg. 1993;19(6):700–712. Erratu m in : 1994;20(6):677 and 2007;33(12):2–3.
13. H aigis W. The H aigis Formu la. In: Sham m as, H J, ed. Intraocular Lens Power Calculations.
Thorofare, N J: SLACK Incorporated ; 2003:41–57.
 3
Im me rsio n Ve rs us
Ap p la n atio n Ultra so u n d

Kenneth J. Hoffer, M D

T
he im mersion tech n ique of Ossoin ig 1 has been show n to be more accu rate than
the stand ard applanation tech n ique in several stud ies2-6 over the past 3 decades.
They report a mean average shorten ing of the AL of 0.21 m m (range from 0.11 to
0.36 m m) u sing applanation compared to im m ersion, and the longer the eye the greater
the shorten ing4 (Table 3-1).
Theoretically, if th is iatrogen ic error was con sistent it cou ld be comp ensated for by the
add ition of a simple constant or by formu la p ersonalization, but th is is not possible since
the error varies so much from eye to eye.
Argu ments against im mersion are that it is expen sive, time-consu m ing, messy, and
requ ires the patient to be supine. On the contrary, the exam ination can be perform ed
very easily in a stand ard ophthalm ic exam ination chair (Fig. 3-1A) reclined back at a 45º
angle w ith the head rest set back so that the patient’s AL is p erp end icu lar to the floor (Fig.
3-1B).
To m aintain a non-leaking flu id bath in the Ossoin ig scleral shell (Fig. 3-2A) (H an sen
Ophthalm ic Developm ent Labs, Coralville, IA, w w w.H an senLab.com), we u se a 50/ 50
d ilution of 2.5% hyd roxypropyl methycellu lose (Gon iosol, CIBA Vision Ophthalm ics,
Atlanta, GA) in Dacriose solution (Fig. 3-2B th rough D). Once the eye is anesthetized topi-
cally, the scleral shell is gently placed between the lid s and filled th ree-qu arters fu ll w ith
the solution. Any air bubbles shou ld be vacuu med w ith a short silicone tu be attached
to a syringe. The latter can also be u sed to remove the solution at the completion of the
procedu re. The u ltrasou nd probe is placed into the solution and p ositioned parallel to the

13
Hoffer KJ. IOL Power (pp. 13-16).
© 2011 SLACK Incorporated.
14 Ch apter 3

Ta b le 3 -1 .

APPLANATIO N VS IMMERSIO N STUDIES IN THE LITERATURE SINCE 1981

Report by # Eyes Applanation US Immersion US Shortening

Hoffer 1981 28 22.77 23.10 0.33
Shammas 1984 180 23.28 23.52 0.24
Artaria 1986 131 23.13 23.44 0.31
Schelenz and 46 (<23.3) 22.39 22.59 0.20
Kammann 1989 54 (>23.3) 24.06 24.38 0.32
Olsen 1989 60 23.35 23.49 0.14
Watson and 225 23.24 23.55 0.11
Armstrong 1999
TOTAL 724 0.21

A
B

Figure 3-1. Patient sitting (A) and then reclined (B) in standard ophthalmic exam chair for immersion
technique.

A B C

D
Figure 3-2. Immersion setup. (A) Gonisol cut 50% with Dacriose,
(B) O ssoinig cup placed between lids, (C) solution fills cup, (D)
ultrasound probe placed in solution and parallel to the eye’s axis.
 Immersion Versus Applan ation Ultrasoun d 15

Figure 3-4. Prager (left) and Kohn shells are an

alternative immersion technique.

Figure 3-3. Axiality is indicated by spike patterns

on the oscilloscope screen as the probe position
is adjusted.

axis of the eye. Axiality is judged by watch ing for the correct spike patterns on the oscil-
loscop e screen as the probe position is adju sted (Fig. 3-3).
The first spike is the echo em ission from the end of the probe, the second is the dou ble-
spike from the cornea, the th ird is the anterior of the lens, the fou rth is the posterior of the
lens, the fifth is the retinal spike, and the rem ain ing are the spikes from the sclera and
orbital tissues. It is imp ortant to adju st the sou nd inten sity so that the spikes can be seen
to change their height. Then the probe is adju sted so that the corneal and retinal spikes
are about equally h igh and straight. Then the read ing shou ld be taken.
Many find the Prager or Koh n shells (ESI, Inc, Plymouth, MN ) (Fig. 3-4) easier to u se
for im m ersion. I have no experience u sing it; therefore see Chapter 7 on its u se.

N OTE: If the A L is very difficult to obtain and the eye appears to have a length greater than
25 mm, suspect a STA PHY LOM A . Shammas recommends direct ophthalmoscopy (with patient
fixating on cross-hair target); measure the distance from the target (macula) to the edge of the optic
nerve (in disc diameters). B-scan exam is then performed to measure the A L at that distance from
the edge of the optic nerve shadow (personal communication) (see Chapter 27).

N OTE: W hen measuring an eye containing an IOL, ignore multiple reduplication echoes noted
in the vitreous space that are caused by the IOL.

N OTE: If planning silicone oil injection into the vitreous space, perform an accurate A L mea-
surement before doing so, and make this information available to the patient. It is practically impos-
sible to measure a silicone oil eye. If your US instrument allows you to do so, set the vitreous gate to
the corresponding sound velocity of the silicone oil (either 1000 cSt or 5000 cSt. Otherwise, using
16 Ch apter 3

a velocity of 1000 m/s may not achieve the desired accuracy. The Z eiss IOLM aster is the best way
to get a measurement in silicone oil-filled eyes (see Chapter 11).

NO TE: Mea sur ing t he AL of BO TH eyes is pr udent a nd cust oma r y.

Always measu re AL to the nearest hu nd redth of a m illimeter and record it carefu lly.
Errors in AL are the most sign ificant and amou nt to ~2.5 D/ m m in IOL p ower. But, it is
important to be aware that th is error d rops to ~1.75 D/ m m in very long eyes (30 m m) and
ju mps to ~3.75 D/ m m in very short eyes (20 m m). Greater care mu st be taken in m easu r-
ing short eyes.

Re fe re nce s
1. Ossoin ig KC. Stand ard ized echography: Basic principles, clin ical application s, and
resu lts. Int Ophthalmol Clin. 1979;19(4):127–210.
2. Sham m as H JF. A comparison of im m ersion and contact tech n iques for axial length m ea-
su rem ents. J A m Intraocul Implant Soc. 1984;10(4):444–447.
3. Artaria LG. Axial length m easu rem ents w ith d ifferent u ltrasou nd d evices. Klin M onast
Augenheilkd. 1986;188:492-494.
4. Schelen z J, Kam m an n J. Comparison of contact and im m ersion tech n iqu es for axial m ea-
su rem ent and implant p ower calcu lation. J Cataract Refract Surg. 1989;15(4):425–428.
5. Olsen T, Nielsen PJ. Im m ersion versu s contact tech n ique in the m easu rem ent of axial
length by u ltrasou nd. A cta Ophthalmol (Copenh). 1989;67(1):101–102.
6. Watson A, Arm strong R. Contact or im m ersion tech n ique for axial length m easu rem ent?
Aust N Z J Ophthalmol. 1999;27(1):49–51.
 4
Ultra so u n d Ve lo cit ie s
fo r Ax ia l Le n g t h
Me a su re m e n t
Kenneth J. Hoffer, M D

M
easu ring the AL of the eye u sing an A-scan is dependent upon the sou nd velocity
the instru ment is set at for the measu rement. Some instru ments u se an average
velocity for the entire eye while others u se ind ividual velocities for each part of
the eye. It wou ld be valuable for you to find out what method you r instru ment u ses.
The u ltrasou nd velocities for the variou s parts of the eye and intraocu lar lens m aterials
are show n in Table 4-1 and the average pseudophakic velocities that the author 1 calcu lated
in 1974 are show n in Table 4-2. The effect of AL error is 2.35 D for every m m in a 23.5 m m
eye, but d rops to 1.75 D/ m m in a 30 m m eye. It rises d ram atically to 3.75 D/ m m in a 20 m m
eye, which is the reason short eyes are the problem eyes in IOL power (Fig. 4-1).
To calcu late the average sou nd velocity in a norm al PH AKIC eye of 23.5 m m, the aver-
age th ickness of the cornea (0.55) and lens (4.63) mu st be u sed to ind ividu ally calcu late
the tim e it takes the sou nd wave to travel th rough each of them . Th is is done by d ivid ing
them ind ividu ally by their respective velocity (1641 m/ sec) (Fig. 4-2). The rem ain ing flu id
parts of the eye (23.5 - 0.55 - 4.63) mu st then be d ivided by the flu id velocity (1532 m/ sec).
The total time it takes for the sou nd to traverse the entire eye is the su m of the “solid”
time and the “liqu id” time. We can u se the formu la Velocity = Distance/ Tim e (V = d/ t).
Since the eye has a given AL of 23.5 m m, that value is d ivided by the total time (15.1148
m/ sec) yield ing an average velocity of 1555 m/ sec. If th is same exercise is p erform ed on
a sample long eye and short eye (Fig. 4-3) the calcu lation s reveal that the average velocity
is not the sam e as for a norm al length eye. A 30 m m eye thu s has a velocity of 1549 m/ sec
and a 20 m m eye has a velocity of 1561 m/ sec.
17
Hoffer KJ. IOL Power (pp. 17-24).
© 2011 SLACK Incorporated.
18 Ch apter 4

Ta b le 4 -1 .

ULTRASO UND VELO CITIES (METERS/SECO ND ) AT BO DY TEMPERATURE2

Substance Velocity (m/sec)
Cornea and lens 1641
Aqueous and vitreous 1532
PMMA IO L 2660
Silicone IO L 980
Acrylic IO L 2026
Glass IO L 6040
Silicone oil 987

Ta b le 4 -2 .

HO FFER CALCULATED AVERAGE SO UND SPEEDS (METERS/SECO ND )

FO R VARIO US C O NDITIO NS O F A 23.5 MM EYE1

Eye Status Velocity (m/sec)

Phakic eye 1555
Aphakic eye 1534
PMMA pseudophakic 1556
Silicone pseudophakic 1476
Acrylic pseudophakic 1549
Glass pseudophakic 1549
Phakic silicone oil 1139
Aphakic silicone oil 1052

Figure 4-1. Effect of an error in axial length on Figure 4-2. Calculations to determine the average
resultant refractive error. sound velocity of a normal size phakic eye.
 Ultrasoun d Velocities for Axial Len g th Measuremen t 19

A B

Figure 4-3. Average velocity for long (A) and short (B) eyes.

Figure 4-4. Graph of the

average velocity (Y-axis)
based on the AL (X-axis).

Graph ing the average velocity against AL reveals an inverse relationsh ip to the AL
(Fig. 4-4). As can be seen, the slope is steeper in short eyes than it is in long eyes.
Now if we do the same exercise on an example norm al length aphakic eye (Fig. 4-5A,)
we obtain a d rop in average velocity from 1555 to 1534 m/ sec, wh ich is totally due to the
loss of the crystalline lens. Calcu lating the sam e for long and short aphakic eyes (Fig. 4-
5B) reveals that AL does not seem to affect the average velocity. Th is tells u s that it is the
lens in the eye that cau ses th is average velocity variation.
It is imp ortant to p erform th is calcu lation on pseudophakic eyes w ith IOLs of d ifferent
m aterials. We can p erform th is calcu lation on eyes of norm al AL and thu s for a PMMA
len s (Fig. 4-6A) of average th ickness of 0.74 m m we obtain an average velocity of 1555
m/ sec, wh ich is the same as a phakic eye. If we replace the lens w ith one of Collamer
(Staar Su rgical Company, Mon rovia, CA) (Fig. 4-6B) of average th ickness of 0.74 m m, the
average velocity d rop s to 1540 m/ sec.
20 Ch apter 4

A B

Figure 4-5. Calculation of average velocities in a normal size aphakic eye (A) and long and short apha-
kic eyes (B).

A B

Figure 4-6. Calculation of average velocities in a normal size PMMA pseudophakic eye (A) and a
Collamer pseudophakic eye (B).

Figure 4-7. Calculation of average velocity in a nor-

mal size Silicone pseudophakic eye.

Repeating th is for a norm al eye w ith a silicone len s (Fig. 4-7) of average th ickness 1.34
m m wh ich has a sou nd velocity of on ly 980 m/ sec, we see a d ram atic d rop of the average
eye velocity to 1487 m/ sec. Th is w ill have a substantive effect on the IOL power calcu la-
tion lead ing to clin ically sign ificant errors if 1555 m/ sec is u sed to m easu re the eye.
 Ultrasoun d Velocities for Axial Len g th Measuremen t 21

N OTE: M easuring an eye containing a silicone IOL with a standard phakic velocity (1555 m/
sec) can amount to an error of 3 to 4 D.

CORRECTING AN ERROR IN VELOCITY

If an eye has been m easu red u sing the w rong velocity, it can be easily corrected w ith-
out rem easu ring the eye by u sing the follow ing formu la:
ALC = (ALM ) x (VC ) ÷ V M
where AL = axial length, V = u ltrasou nd velocity, C = correct and M = m easu red.
Th is is becau se the u ltrasou nd instru ment does not measu re length or d istance (d)
d irectly. In stead it measu res the time (t) it takes the sou nd to traverse the eye and converts
it to a linear value u sing the velocity (V) formu la where d = V x t.

OPTIONAL CALF METHOD

Hollad ay 3 pu blished an optional method to m easu re the AL that attempts to decrease
the error in herent in changes in average velocity due to the length of the eye. All eyes,
regard less of statu s, are measu red at a velocity of 1532 m/ sec (as if the eye was a bag of
water) and to th is value is added the Corrected AL Factor (CALF). The CALF value repre-
sents the th ickness of a lens in the eye whether it is the crystalline lens or an IOL(s). The
formu la for the CALF of any lens (includ ing the cornea) is:
CALF = TL x (1-1532/ VL)
where TL = the axial th ickness of the len s and VL = the sou nd velocity th rough that
lens.
Hollad ay computes the th ickness of the hu m an cataractou s len s u sing:
TL = 4 + Age/ 100
and the sou nd velocity th rough the cataract u sing:
VL = 1659 - [(Age - 10)/ 2]
Su bstituting these two formu las into the CALF formu la above, the CALF formu la for
the crystalline lens yield s:

Age 1532
CALF = 4 + × 1 −
100 Age − 1
0
10
1659 −
2

CALF formu la is dep endent on ly upon the age of the patient.

As can be seen, the CALF for the cataractou s len s is calcu lated u sing on ly the age of
the patient. Hollad ay recom m end s u sing a CALF value of 0.28 (value for 70 year-old) for
all ages becau se the value for a 1-year-old is 0.306 and that for a 100-year old is 0.224. The
m aximu m error in CALF for those you nger than 70 is 0.026 (~0.07 D) and for those older
than 70 it is 0.056 (~0.14 D). The reason ing beh ind th is method is that, if an “average” eye
velocity is incorrect, it affects the entire AL measu rement. However, if the estim ate of the
CALF value is w rong, it on ly affects a sm all percentage of the overall AL, ie, on ly the lens
portion.
22 Ch apter 4

Ta b le 4 -3 .

FO RMULAS FO R C ALCULATING BIOMETRIC PARAMETERS

A. CALF factors for pseudophakic eyes (using CALF = TL × (1 – 1532/VL)
where VL = the sound velocity for the IO L material in the eye:
• CALFPMMA = TL × (1 – 1532/2660) = +0.424 × TL
• CALFSILICO NE = TL × (1 – 1532/980) = –0.563 × TL
• CALFACRYLIC = TL × (1 – 1532/2026) = +0.243 × TL

B. The correction for the cornea:

• CALFCO RNEA = TC × (1 – 1532/1641) = 0.55 × (0.066423) = 0.037

C. Knowing the thickness of the implanted IO L (which can be obtained from the manu-
facturer) the following formulas can be used:
• PMMA Eye AL = AL1532 + 0.424 × TL + 0.037
• Silicone Eye AL = AL1532 – 0.563 × TL + 0.037
• Acrylic Eye AL = AL1532 + 0.243 × TL + 0.037
• Piggyback IO Ls AL = AL1532 + T1 × (1 – 1532/V1) + T2 × (1 – 1532/V2) + 0.037
where T1 and T2 are the thickness and V1 and V2 are the velocity of each IO L.

CORNEA MUST BE CONSIDERED

H is formu lation, however, ignores the factor of the corneal th ickness (0.55 m m). To
correct th is, I recom mend u sing a CALF of 0.32 (0.28 + 0.037). The 0.37 correction for the
cornea is calcu lated in Table 4-3B.
A sim ilar method can be u sed for p seudophakic eyes u sing CALF = TL x (1-1532/ VL)
and the know n VL for each IOL m aterial (Table 4-3A). Know ing the th ickness of the
implanted IOL, the formu las in Table 4-3C can be u sed. If the IOL th ickness can not be
obtained, Hollad ay 3 published a table to u se based on the p ower of the IOL. The resu lting
formu las based on IOL m aterial are show n in Fig. 4-8. The AL of an eye contain ing two
IOLs of d ifferent m aterials can be obtained u sing the formu la in Table 4-3C.

BIPHAKIC EYES (PHAKIC EYES WITH A PHAKIC IOL)

A phakic eye w ith a phakic IOL implanted (biphakic) m ay develop a cataract. The
problem here is elim inating the effect of the sou nd velocity th rough the phakic lens when
m easu ring the AL u sing u ltrasou nd. The A-scan assu med that sou nd travels th rough the
phakic refractive lens (PRL), for example, at 1555 m/ sec; however, sou nd travels th rough
the PRL at 980 m/ sec. To calcu late the true AL of th is eye, we mu st calcu late and su btract
the erroneou s (E) d istance of the PRL and add back the true (T) d istance (Fig. 4-9):
AL + T – E
where E = T × 1555/ 980.
The phakic IOL th ickness can be obtained from the publication s 4,5 wh ich have tables
show ing the phakic IOL central th ickness based on their d ioptric p ower for each phakic
IOL on the m arket tod ay.
 Ultrasoun d Velocities for Axial Len g th Measuremen t 23

Figure 4-8. CALF formulas for pseudophakic eyes Figure 4-9. Axial length with ultrasound in bipha-
containing different materials. kic eyes.

Then the
AL + T - E = AL + T – (T × 1555/ 980) = AL + T × (1 – 1555/ 980).
The author prop osed a method 4 to correct for th is potential error by u sing the follow-
ing resu ltant formu la:
ALCORRECTED = AL1555 + C * T
where AL1555 = the measu red AL of the eye at sou nd velocity of 1555 m/ sec, T = the central
th ickness of the phakic IOL and C = the m aterial specific correction factor of +0.42 for PMMA,
-0.59 for silicone, +0.11 for collam er, and +0.23 for acrylic.

Re fe re nce s
1. Hoffer KJ. Ultrasou nd velocities for axial length m easu rem ent. J Cataract Refract Surg.
1994;20(5):554–562.
2. Mark H F, Bikales N, Overberg CG, Menges, Kroschw itz G JI. Encyclopedia of Polymer
Science and Engineering, Volu m e 1, 2nd ed. New York: Wiley and Son s. 1985;147–149.
3. Hollad ay JT. Stand ard izing con stants for u ltrason ic biom etry, keratom etry, and intraocu-
lar len s p ower calcu lation s. J Cataract Refract Surg. 1997;23(9):1356–1370.
4. Hoffer KJ. Ultrasou nd axial length m easu rem ent in biphakic eyes. J Cataract Refract Surg.
2003;29(4):961–965.
5. Hoffer KJ. Add endu m to u ltrasou nd axial length m easu rem ent in biphakic eyes:
Factors for Alcon L12500–L14000 anterior cham ber phakic IOLs. J Cataract Refract Surg.
2007;33(5):751–752.

Sug g e s te d Re a ding
Hollad ay JT, Prager TC. Accu rate u ltrason ic biom etry in p seudophakia. A mer J Ophthalmol.
1989;107(2):189–190.
 5

A-Sca n Bio m e try:

Im me rsio n Te ch n iq u e
H. John Shammas, M D

T
he im mersion tech n ique described herein can be u sed w ith any u ltrasou nd u n it
equ ipp ed w ith a solid A-scan probe and mobile electron ic gates.1-5

l The patient is placed in a supine position on a flat exam ination table or in a reclin ing
exam ination chair and a d rop of local anesthetic is instilled in both eyes.
l An Ossoin ig scleral shell is applied between the lid s. The most com mon ly u sed
scleral shells are the H ansen shells, the Prager shell, and the Koh n shell (Fig. 5-1).
l The Ossoin ig shell is filled w ith gon ioscopic solution. Methylcellu lose 1% is pre-
ferred over the 2.5% concentration (wh ich is too th ick) and over saline solutions (too
liqu id). The solution shou ld be free of air bu bbles; the presence of bu bbles cau ses
variations in the speed of sou nd and is resp onsible for noise form ation w ith in the
u ltrasou nd pattern. The easiest way to avoid bubbles is to remove the bottle’s n ipple
and to pou r the solution in the cup. If bu bbles do form w ith in the solution, they are
removed w ith a syringe, and, if u nsuccessfu l, the cup has to be emptied, cleaned,
repositioned and refilled w ith gon ioscopic solution.
l The Prager and the Koh n shells are designed to hold the probe tightly and allow a
better fit on the eye. Becau se of th is tight fit, the coupling flu id u sed in these shells
does not have to be methylcellu lose; instead, balanced salt solution or artificial tears
cou ld be u sed.

25
Hoffer KJ. IOL Power (pp. 25 -28).
© 2011 SLACK Incorporated.
26 Ch apter 5

Figure 5-1. The Kohn, O ssoinig (Hansen), and Prager

scleral shells are displayed from left to right.

Figure 5-2. The ultrasound probe

is immersed in the solution keep-
ing it 5 to 10 mm away from the
cornea.

Figure 5-3. Ultrasound display of the different echospikes dur-

ing immersion A-scan biometry, identifying from left to right:
the initial spike, the anterior and posterior corneal surfaces, the
anterior and posterior lens surfaces, the retina, sclera, and orbital
tissues.

l The u ltrasou nd probe is inserted in the Koh n shell that keeps the tip 5 to 8 m m away
from the cornea (Fig. 5-2). The patient is asked to look, w ith the fellow eye, at a fixa-
tion point placed at the ceiling. Gently move the probe u ntil it is properly aligned
w ith the optical axis of the eye and an acceptable A-scan echogram is d isplayed on
the screen.
The A-scan pattern of a norm al phakic eye exam ined w ith an im mersion tech n ique
d isplays the follow ing echo spikes from left to right (Fig. 5-3):
l The in itial spike is produced at the tip of the probe. It has no clin ical sign ificance.
l The corneal spike is double-p eaked, representing the anterior and posterior su rfaces
of the cornea.
l The anterior lens spike is generated from the anterior su rface of the lens.
l The posterior len s spike is generated from the posterior su rface of the lens.
l The retinal spike is generated from the anterior su rface of the retina. It is straight,
h igh ly-reflective, and tall whenever the u ltrasou nd beam is perp end icu lar to the
retina (as it shou ld be du ring AL measu rem ent).
 A-Scan Biometry: Immersion Tech n ique 27

l The scleral spike is another h igh ly-reflective spike generated from the scleral su r-
face right beh ind the retinal spike, and shou ld not be con fu sed w ith it.
l The orbital tissues create low reflective spikes beh ind the scleral spike.
Many older biometers give the read ings d irectly in m illimeters u sing an average sou nd
velocity of 1550 to 1555 m/ s.6,7 The most accu rate velocity to u se is noted in Table 4-2 of
the Ultrasou nd Velocities chapter. Most modern biometers u se ind ividu al velocities for
each of the eye’s comp onents (cornea, aqueou s, lens, and vitreou s), but it is imp ortant to
m ake su re the values are set to the appropriate ones show n in Table 4-1 of the Ultrasou nd
Velocities chapter. It helps to replace older US u n its w ith more modern ones or an optical
m easu ring device such as the IOLMaster or LENSTAR LS900.
l The th ickness of the cornea is measu red u sing a velocity of 1641 m/ sec.
l The anterior aqueou s depth is measu red between the posterior corneal su rface and
the anterior len s su rface u sing a velocity of 1532 m/ sec.
l The len s th ickness is measu red between the anterior len s su rface and the posterior
lens su rface u sing a velocity of 1641 m/ sec. Actu ally 1640.5 m/ sec is the calcu lated
sou nd velocity in the norm al crystalline lens. The sou nd velocity varies in catarac-
tou s eyes w ith a slower velocity (average 1590 m/ sec) in the intu mescent cataracts
due to their h igh water content, and a h igher velocity in the p osterior capsu lar cata-
racts. In most cases of nuclear sclerosis w ith or w ithout subcapsu lar changes, the
sou nd velocity averages 1641 m/ sec.
l The vitreou s cavity’s depth is measu red between the posterior len s su rface and the
anterior su rface of the retina u sing a velocity of 1532 m/ sec.

Re fe re nce s
1. Byrne SF. Stand ard ized echography, Part I: A-scan exam ination procedu res. Int Ophthalmol
Clin. 1979;19(4):267–281. No abstract available.
2. Ossoin ig KC. Stand ard ized echography: Basic principles, clin ical application s and resu lts.
Int Ophthal Clin. 1979;19(4):127–210.
3. Sham m as H J. Axial length m easu rem ent and its relation to intraocu lar len s p ower calcu-
lations. J A m Intraocul Implant Soc. 1982;8(4):346–349.
4. Sham m as H J. Manu al versu s electron ic m easu rem ent of the axial length. In : H illm an JS,
LeMay MM, ed s. Ultrasonography in Ophthalmology. Proceedings of the 1982 N inth SIDUO
Congress. The H ague: Dr. W. Ju n k Pu blishers; 1983:225–229.
5. Sham m as H J. A-scan biom etry of 1000 cataractou s eyes. In : Ossoin ing KC, ed. Ophthalmic
Echography. Proceedings of the 10th SIDUO Congress. The H ague: Dr. W. Ju n k Pu blishers;
1987:48,57–63.
6. Oksala A, Lehtinen A. Measu rem ent of the velocity of sou nd in som e parts of the eye. A cta
Ophthalmol. 1958;36(4):633–639.
7. Hoffer KJ. Ultrasou nd velocities for axial length m easu rem ent. J Cataract Refract Surg.
1994;20(5):554–562.
8. Colem an DJ, Lizzi FL, Fran zen LA, Abram son DH. A d eterm ination of the velocity of
u ltrasou nd in cataractou s len ses. Bibl Ophthalmol. 1975;83:246–251.
 6
A-Sca n Bio m e try:
Avo id in g Pitfa lls Wit h
Im me rsio n Te ch n iq u e
Karl Ossoinig, M D

His to rical Ba ckg ro und

E
arly in the 1960s, the Swed ish ophthalmologist F. Janson,1,2 th rough h is extensive
experim ental and clin ical work, introduced precise and accu rate im mersion AL
measu rements. He also gave u s the exact sou nd velocities w ith in the clear len s and
the vitreou s body as well as the anterior cham ber flu id. Hoffer introduced im mersion
A-scan for IOL power calcu lation in the USA in 1974 and stimu lated Sonometrics to pro-
duce a stand-alone u nit that introduced applanation contact A-scan. He later proved (see
Chapter 1) that applanation shortened the eye compared to im mersion and abandoned
its u se. Th is m ethod however gained im m ense popu larity becau se it was considered an
easier and faster m ethod. It spread rapid ly and w idely for m ain ly two reasons:
l It requ ired no add itional space in ophthalm ic offices to recline a patient as needed
for the im mersion m ethod.
l It was easier to learn and qu icker to p erform.

Why Im m e rs io n Me tho d
Over the past decade the increasing interest in the im mersion method is due to the
follow ing reason s:
29
Hoffer KJ. IOL Power (pp. 29 -38).
© 2011 SLACK Incorporated.
30 Ch apter 6

l Increased expectation s from patients and med icolegal concern s.

l Increasing competition among su rgeon s.
l The availability of improved IOLs for treating presbyopia.
l The IOLMaster has raised the bar in AL accu racy.
Tod ay the IOLMaster sets the stand ard for the accu racy and precision of AL measu re-
m ents needed for IOL p ower calcu lation s. The Partial Coherence Interferometer (PCI),
however, is u nable to obtain a read ing in 8% to 20% of the patients due to dense cataracts
(esp ecially those w ith su b-cap su lar plaques) and fixation problem s (esp ecially in eyes
w ith m acu lar degeneration).
There are two m ain featu res that m ake the im m ersion method superior to applana-
tion:
l No u ncontrollable and u ncorrectable applanation wh ich shortens the eye on the
average approxim ately 0.2 m m (shorten ing up to 0.8 m m has been observed).
l Objective align ment of the u ltrason ic beam w ith the optical axis of the eye.

Ba s ic Principle s o f the Im m e rs io n Me tho d

l Ultrason ic probe rem ains remote from the cornea (no touch, no applanation) (Fig.
6-1).
l Display of both corneal signals together w ith both lens signals and the retinal (+ sclera)
signals.
l Align ment of the u ltrason ic beam w ith the optical axis of the eye (objective align-
ment much more reliable than subjective align m ent).
l Optim al resu lts are comparable to those obtained w ith the IOLMaster (requ ires,
however, adequ ate instru ment parameters and optim al tech n iques).

Ex am inatio n Te chnique : Pre pa ra tio n

l Patient in sem i-supine position.
l Local anesthetic.
l Flawless im m ersion (no leaking):
© Im m ersion shells (u ltra-light, adequ ate sizes, no need to be hand-held) (Fig. 6-2).
© Saline as im m ersion flu id (no air bubbles in spite of rep eated “in and out” of the
probe) (Fig. 6-3).
© Prior sealing of the in ner shell edges w ith th icker flu id (eg, 1.6 % m ethyl-cel-
lu lose).

Ex am inatio n Te chnique
l H and-held probe im m ersed and grossly placed over corneal center. Observation
need s to occu r from above (th rough the flu id) and not from the side so as to avoid
m islead ing refraction.
 A-Scan Biometry: Avoidin g Pitfalls With Immersion Tech n ique 31

Figure 6 -1. Immersion A-scan remains remote from the

cornea preventing eye compression.

Figure 6-2. Hansen O ssoinig immersion

A-scan shells.

Figure 6 -3. Preparation steps for performing immersion

A-scan.

l In itially, put the A-scan on h igh in stru m ent gain (h igh ocu lar signals from the cor-
nea, both lens su rfaces, and the fu ndu s) to more easily recogn ize a su itable beam
d irection.
l Align ment of the u ltrason ic beam w ith the axis of the eye at decreasing gain set-
tings. There are two ways to ach ieve align ment:
© Subjective: Th is is the trad itional m ethod u sing a light sou rce in the tip of the
US probe so the patient can fixate on the light. Th is approach is prone to error
becau se the acou stic and optical pathways are not identical and m ay deviate
from each other sign ificantly. Th is can be due to d ifferent refractions, especially
in lenses w ith den se su bcapsu lar plaques. These are one of the sou rces of failu re
w ith the IOLMaster. Other patients u nable to benefit from the IOLMaster are
those not able to fixate on the light sou rce at all.
© Objective: Th is is optim al becau se the beam is objectively aligned w ith the opti-
cal axis of the eye. Th is approach secu res optim al measu ring accu racy. Th is is
ach ieved when the u ltrason ic beam is aimed perp end icu larly at the cornea and
both len s su rfaces simu ltaneou sly. The retina signal is not assessed at th is point.
32 Ch apter 6

Th is simu ltaneou s perpend icu larity autom atically gu ides the beam toward the
m acu la, and th is is especially important in long eyes w ith posterior staphylom as
and in short eyes w ith atypical shap es.
Th is objective align m ent of the u ltrason ic beam w ith the optical axis of the eye does
not requ ire patient cooperation. To obtain th is align m ent the probe mu st be freely mov-
able by the hand of the exam iner (do not u se probe holders), wh ich allow s for such an
objective beam align ment. The instru ment gain is slowly d ecreased du ring the scan n ing
procedu re in order to reduce the height of the cornea and len s spikes as much as p ossible
(w ithout losing the presu med retinal signal). At low spike height, their statu s of being
m axim ized (p erp end icu lar sou nd beam) is recogn ized much better as well as adju sted
(increased d isplay sen sitivity).
Th is optim al align ment of the u ltrason ic beam w ith the optical axis of the eye (ie, aim -
ing the beam from the corneal center to the fovea and thu s measu re the correct AL) lead s
to optim ized “measu ring accu racy.” Is th is d ifficu lt? Yes it is, but on ly in the begin n ing.
With some practice an acqu ired reflex sets in wh ich allow s the hand of the exam iner
to gu ide the beam almost autom atically into the needed simu ltaneou sly p erp end icu lar
d irection accord ing to the behavior of the signal height noted by the exam iner wh ile
observing the screen d isplay. Th is acqu ired reflex is comparable to how we are gu ided
wh ile walking, autom atically activating and controlling the relevant mu scles.
l Freeze the optim ized (regard ing beam p osition and d irection) echogram, eg, by
depressing a foot sw itch.
l Electron ic m easu ring gates are u su ally set autom atically once the echographer
accepts an AL echogram as optim al (eg, by depressing a foot sw itch). They need,
however, to be checked for precise setting and, if necessary, mu st now be corrected
before the echogram is processed for calcu lation of the measu red AL.
l When possible, peak-to-peak measu rements shou ld be applied. The p eaks of the
echo spikes tru ly represent the u nderlying acou stic interfaces and are not affected
by phase sh ifts stem m ing from d ifferences in echo intensity between the measu red
signals (eg, between the anterior and posterior lens signals). Wh ile such phase sh ifts
cau se on ly m in im al errors in the d istance measu rements, they shou ld be avoided for
the sake of optim al m easu ring precision. Most biometric instru ments, however, do
not yet allow th is approach. It is then best to get the measu ring m arkers as close as
possible to the peaks of the relevant echo spikes. The worst scenario is to m easu re
from signal base to signal base.
l Next, check on the retinal signal by increasing the instru ment gain and thu s all of
the echo signals to con firm that the presu med retinal signal indeed represents the
retinal su rface and not the stronger reflective sclera su rface.
l Finally, send th is echogram for processing the calcu lation of the measu red AL in
mm.
l Repeat th is entire procedu re at least th ree times for optim al resu lts by repeating all
the above steps from the begin n ing.
l The measu ring procedu re is completed when th ree acceptable echogram s have been
obtained. Rep eating the entire procedu re each time independent from each other
one avoid s m aking the same align ment m istake each time, wh ich m ay also resu lt in
close (but false) values.
l The final resu lt is the average of the th ree measu rements.
 A-Scan Biometry: Avoidin g Pitfalls With Immersion Tech n ique 33

Figure 6 -4. New software facilitates

the use of the immersion method.

Figure 6 -5. Example of results of new

software facilitating the immersion
method.

Ne w S o ftw are to Im pro ve

the Im m e rs io n Me tho d
An experimental design of a new software (develop ed by th is author and Jean Abascal
of Qu antel Med ical) was tested on a nu m ber of cataract patients in the Un iversity of
Vien na Eye Clin ic. The author perform ed the u ltrason ic im m ersion measu rem ents u sing
the new software wh ile Prof. Dr. Lu kas measu red the ALs of the same patients u sing the
IOLMaster. Both measu rements were then compared. In those cases where the IOLMaster
was successfu l, the resu lts were almost identical w ith the echograph ic measu rements,
w ith a tendency of the PCI m easu rem ents to be m in im ally longer.
With th is software, all the echographer has to do is angle and sh ift the US beam,
attempting to aim a perp end icu lar beam at both cornea and lens su rfaces (Fig. 6-4). The
instru ment then “remem bers” wh ich were the optim al read ings and, upon simu ltaneou s
d isplay of m axim ize corneal and lens signals, freezes and measu res th is optim al echo-
gram (beam precisely aligned w ith the optical axis of the eye) (Fig. 6-5).
34 Ch apter 6

Sum m ary o f Re quire m e nts fo r

an Optim al Im m e rs io n Te chnique
l Im m ersion shells for d ifferent eye sizes.
l H and-gu ided probe.
l Start at h igh system sen sitivity for gross beam align ment.
l Decrease system sensitivity for selection of optim al beam align m ent th rough simu l-
taneou s d isplay of m axim al signals from the corneal and lens su rfaces. Th is objec-
tively aligns the acou stic beam w ith the optical axis of the eye, wh ile decreasing
system sen sitivity to optim al (low) setting.
l Optim al align ment determ ined by exam iner vs. instru m ent software.
l Optim al echogram is frozen “m anu ally“ by the echographer or autom atically by the
new software.
l Five measu ring gates u sing correct tissue-dep endent sou nd velocities (cornea 1641
m/ sec; anterior cham ber and vitreou s cavity 1532 m/ sec; lens 1641 m/ sec).
l If necessary, correct the autom atically set optim al gate settings over the 5 peaks rep -
resenting the corneal epitheliu m and endotheliu m, the anterior and posterior len s
su rfaces, and the retinal su rface.
l Increase system sensitivity to verify retinal signal in frozen echogram .
l Obtain 3 independent measu rem ents w ith close resu lts repeating each time the
entire procedu re begin n ing w ith probe im m ersion and includ ing optim al beam
align ment.
l Use the average AL measu red th is way (calcu lated and d isplayed autom atically by
the in stru m ent).

Ins trum e nt and Pro be De s ig ns

fo r the Im m e rs io n Me tho d
Becau se of its easier application (both facility- and tech n ique-w ise) the Applanation
Method dom inated the m arket worldw ide. As a con sequence, the tech n ical requ irements
for biometric instru ments were mostly designed and continuou sly improved for the
Applanation Method w ith all its fau lts. The requ irements for the Im mersion Method were
largely ignored by the m ajority of the instru ment m akers. Tod ay that trend is in reverse,
but most instru mentation in u se is partly (some entirely) u n fit to ach ieve fu ll success
u sing im mersion. Therefore, at th is time, m any echographers w ill not yet be able to utilize
all the advantages of the Im mersion Method.

Ma nda to ry Re quire d Para m e te rs

l Correct sou nd velocities (see Table 4-1).
l Adju stable instru ment gain at least prior to freezing of echogram.
l Dynam ic Range (“acou stic field”) between 25 and 45 d b.
 A-Scan Biometry: Avoidin g Pitfalls With Immersion Tech n ique 35

Stro ng ly Pre fe rre d Para m e te rs

(All Re alize d in Standardize d Ins trum e nts)
l Use a non-focu sed 8 MH z probe w ith parallel beam to allow easy recogn ition of
beam perp end icu larity and application of the objective beam align ment w ith the
optical axis of the eye. H igher frequencies and sm aller probes all requ ire focu sing
their beam s and consequently cau se d ifficu lties in recogn izing beam perp end icu-
larity.
l S-shap ed amplifier characteristics (an optim al comprom ise between linear and
logarith m ic amplifiers, each of wh ich has critical advantages and shortcom ings).
l Dynam ic range (“acou stic field”) of 36 dB.
l Observe spike con figu ration to assu re optim al d isplay (see Fig. 6-5).

Optim al Pro be Pro pe rtie s

l A parallel (non-focu sed) u ltrason ic beam provides more accu rate and precise
measu ring resu lts than a focu sed beam becau se signals rise sharply on ly when
the u ltrason ic beam reaches an acou stic su rface p erp end icu larly. It thu s allow s the
exam iner to recogn ize p erp end icu larity wh ile m axim izing the signal height.
l Frequency of the transducer in the probe is of second ary importance. Wh ile it is
true that h igher frequencies provide better resolution, they mu st be focu sed to avoid
d ivergent beam s and so far they have never been focu sed to the point of em itting a
parallel beam.
l In stand ard ized instru ments an 8 MH z stand ard probe is always u sed. Its trans-
ducer d iam eter of 5 m m provides an em itted parallel beam. By reducing the instru-
ment gain, th is parallel beam can be narrowed as needed for optim al resolution and
reducing the spike height brings the peaks into the range w ith greatest sensitivity
(ie, low to m ed iu m spike height). Thu s th is clearly bigger probe provides the facility
of recogn izing beam perp end icu larity wh ile the sm aller focu sed probes do not. By
reducing the instru ment gain, the beam can be m ade very narrow yet still ach ieve
the advantages of a focu sed beam w ithout losing the parallel featu re.

Optim izing Re s ults Us ing No n -Standardize d

A-Scans : Optim izing Me a s uring Pre cis io n
l Use parallel rather than focu sed u ltrason ic beam (to be able to assess correct beam
d irection).
l Use logarith m ic rather than linear amplification (to avoid false m axim al appearance
of signals).
l Start w ith h igh instru ment sensitivity. Reduce instru ment sensitivity to a m in imu m
still d isplaying all pertinent signals before m easu rement.
l Maxim ize signals from cornea and lens simu ltaneou sly to optim ize beam d irec-
tion.
36 Ch apter 6

Figure 6-6. Differentiation between

lens and iris signals when the pupil is
not dilated.

l Do p eak-to-p eak m easu rements 3 of relevant spikes or, if not possible, set the mea-
su ring gates on the left side of the peaks as close to them as possible.
l Avoid devices for easier probe hold ing wh ich are likely to hamper free probe move-
ment and optim al beam align ment.

Avo iding Pitfalls o f the Im m e rs io n Me tho d

l When u sing a focu sed probe (m ade for the applanation method) removed from the
corneal su rface, the retinal su rface m ay sh ift into the more remote d ivergent part
of the beam. To correct th is, acqu ire a probe from the m anu factu rer m ade for the
im m ersion method. The best is a stand ard 8 MH z non-focu sed probe em itting a
parallel u ltrason ic beam.
l When the pupil is not d ilated, the anterior lens su rface signal m ay be con fu sed w ith
the iris signal wh ich is not necessarily received first. Also, in add ition, a lenticu lar
su bcapsu lar plaque signal m ay appear, fu rther con fu sing the situ ation. To correct
th is, d ilate the pupil before doing an im m ersion tech n ique. If th is is not possible
(Fig. 6-6), try to d ifferentiate between the signal from the iris su rface (wh ich is
extremely coarse) and the signal from the anterior lens su rface (wh ich is extremely
smooth). Th is is relatively easy when u sing a stand ard ized in stru m ent. Slight
angling of the beam reveals each of these two comp eting signals. The lens signal
read ily d isappears whereas the iris signal persists du ring the angling. When u sing
a focu sed u ltrason ic beam th is d ifferentiation m ay becom e qu ite d ifficu lt or even
impossible.
 A-Scan Biometry: Avoidin g Pitfalls With Immersion Tech n ique 37

Re fe re nce s
1. Jan son F, Koch E. Determ ination of the velocity of u ltrasou nd in the hu m an len s and
vitreou s. A cta Ophthalmol (Kbh). 1963;40:420-433.
2. Jan son F. Measu rem ents of intraocu lar d istances by u ltrasou nd. A cta Ophthalmol (Kbh).
1963;74(Suppl):1-51.
3. Ossoin ig KC. Basics of Standardized Ophthalmic Echography. 2009 (CD); available from
Echography Teach ing Services at w w w.echography.com .
 7

Ava ila b le A-Sca n

In stru me n tat io n
Kenneth J. Hoffer, M D

I
t is the pu rpose of th is chapter to provide in form ation and sp ecification s for a variety
of A-scans available and in u se th roughout the world. They are show n in Table 7-1.
Som e of these in stru ments listed in the table m ay no longer be available from the
original m anu factu rer, but there is an international second ary m arket of u sed and refu r-
bished equ ipment where those instru m ents m ay be an option.
Since the introduction of the Zeiss IOLMaster in 1999, interest in u ltrasou nd A-scans
for measu ring AL has declined. However, A-scan exam ination is still necessary in 10%
to 17% of eyes (dep end ing upon the patient popu lation) becau se the IOLMaster can not
obtain an accu rate AL read ing due to den se opacity, d en se PSC cataract, or inability to
fixate. Also the exp ense of the IOLMaster precludes its u se in m any parts of the world.

39
Hoffer KJ. IOL Power (pp. 39 -52).
© 2011 SLACK Incorporated.
40 Ch apter 7

Ta b le 7 -1 .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
Accutome AccuSonic A-scan Hoffer Q , Large 7.5 inch, VGA res-
Holladay, olution. LCD screen with
SRK/T intuitive user interface.
Rotary user input knob
and dedicated tactile but-
tons. Compact, portable
unit. O ptional keyboard,
footswitch and external
printer.
Accutome Advent A/B Combo Hoffer Q , A-scan: Tonometer
System Holladay, mounted, hand-held and
SRK, immersion techniques.
SRK/T Multi-tone A-scan lock-
on signal. Customized
surgeon lens profiles.
Calculation of 3 IO Ls at
once. Defaulted velocity
settings. Patient archive
data.
B-scan: Pan and zoom
feature. Window mode,
update mode. 3 defaulted
range values. Memory
storage /image download.
Alcon O cuscan RxP A-scan Hoffer Q , N OT AVAILABLE IN USA
Holladay, Biometry/pachymetry:
SRK/T desktop, computer inter-
facec, ontact or immer-
sion, velocity program-
mable for each segment,
personalization, and
allows up to 5 users
each.

(continued)
 Available A-Scan In strumen tation 41

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
Alcon UltraScan Combo Hoffer Q , N OT AVAILABLE IN USA
Holladay, B-scan, A-scan Biometry
SRK/T plus unique B-Biometry.
Multifunctional system.
Standard 10 MHz probe,
or optional 20 MHz
probe that brings 100
microns of resolution
(depths of 30 to 35 mm)
to posterior segment
images.
DGH Technology DGH 5000e A-scan Binkhorst Solid tip probe, 3 mm
II, diameter, fully auto-
Holladay, matic, internal printer,
SRK II, real time graphic dis-
SRK/T play. Pachymeter mode
(5100e) simultaneously
shows corneal thickness,
current measurement
position, and the selected
corneal map. Multiple
doctor’s configurations.
Data storage. Automatic
measurement mode. No
foot pedal is required.
Exclusive corneal com-
pression detection soft-
ware.

(continued)
42 Ch apter 7

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
DGH Technology DGH 5100e A-scan Binkhorst Solid tip probe, 3 mm
II, diameter, fully auto-
Holladay, matic, internal printer,
SRK II, real time graphic dis-
SRK/T play. Pachymeter mode
(5100e) simultaneously
shows corneal thickness,
current measurement,
position and the selected
corneal map. Multiple
doctor’s configurations.
Data storage. Automatic
measurement mode. No
foot pedal is required.
Exclusive corneal com-
pression detection soft-
ware.
Innovative I3 System- Combo Hoffer Q , A-scan: Measurement
Imaging Inc ABDv1 Holladay accuracy: 0.05 mm, store
up to 20 A-scans inter-
nally.
AL Biom etry: Pre-pro-
grammed velocities,
manual/auto modes,
immersion or contact set-
tings (ABDv2), solid
10 MHz probe.
Innovative I3 System- Combo Hoffer Q , Standardized diagnostic:
Imaging Inc ABDv2 Holladay, Two gate measurements
SRK/T display, tissue sensitivity
value stored, 8 MHz par-
allel beam probe.
B-scan: True geometry
with stepped zoom, pan,
freeze frame, post image
processing. Posterior
segment 10 MHz probe.
Wide-field anterior seg-
ment 20 MHz probe.

(continued)
 Available A-Scan In strumen tation 43

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
Nidek Inc Echoscan US A- scan Binkhorst Gate function; 40 mm
800 II, measure range; por-
Holladay, table, compact, and
SRK II, lightweight. Speedy and
SRK/T highly accurate measure-
ment. Five IO L power
calculations.
Nidek Inc Echoscan US A- scan Binkhorst, Adjustable amplifier gain.
1800 Hoffer Q , Measuring range 12 to
Holladay, 40 mm. Ten MHz solid
SRK, SRK with internal red fixation
II, SRK/T axial length measurement
probe.
Nidek Inc Echoscan US Combo Binkhorst Interfaces with US 2500
2520 II, for A/B combination.
Holladay,
SRK II

Nidek Inc Echoscan US Combo Binkhorst Versatile A/B-scan. Post

3300 II, processing function.
Holladay, User-friendly opera-
SRK II tion. Biometry probe is
optional.

O cuserv DB-3000 Hoffer Q , Ten MHz focused trans-

Holladay, ducer with fixation light.
SRK II, Stores 5 scans per eye,
SRK/T built-in printer.
Can be upgraded to B-
scan.
O cuserv DB-3000C Hoffer Q , Light and affordable. Ten
Holladay, MHz focused transducer
SRK II, with fixation light. Stores
SRK/T 5 scans per eye, printer
with results of ACD and
lens thickness.

(continued)
44 Ch apter 7

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
O cuserv DB-3000CG Hoffer Q , Light and affordable. Ten
Holladay, MHz focused transducer
SRK II, with fixation light. Stores
SRK/T 5 scans per eye, printer
with results of ACD and
lens thickness.
O cuserv DB-3100 Hoffer Q , Ultrasound software on
Holladay, a laptop computer with
SRK II, probe attached. Ten MHz
SRK/T focused transducer. Uses
computer printer.

O phthalmic OTI-A2000 Hoffer Q , Compact, portable 13

Technologies Inc Holladay, Mhz A-scan connects
SRK II, to any laptop or PC
SRK/T computer. Easy-to-use
software automatically
selects and ranks scans.
Up to 9 scans can be
displayed side by side
for fast comparison.
Advanced algorithms for
high myopia, pseudopha-
kia, and dense cataracts.
All major IO L formulas
and unlimited IO Ls.
O phthalmic OTI-2000 Combo Hoffer Q , Ten or 20 MHz un-
Technologies Inc Holladay, focused transducer A-
SRK II, scan connects to any
SRK/T laptop or PC computer.
Easy-to-use software
automatically selects
and ranks scans. Up to 9
scans can be displayed
side by side for fast
comparison. Dynamic
Digital Recording B-scan.
Advanced algorithms for
high myopia, pseudopha-
kia and dense cataracts.

(continued)
 Available A-Scan In strumen tation 45

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
O phthalmic OTI-B/A /3D Combo Hoffer Q , Real time dynamic movie
Technologies Inc 1000 Holladay, recording, Integrated 3D,
SRK II, 12 MHz B-scan, 13 MHz
SRK/T A-scan, UBM option, 50
frames/sec, portable and
compact. Desktop or lap-
top models available.
O ptikon Bioline A-scan Binkhorst, Compact, transportable
Haigis, unit with built-in printer.
Holladay, Large screen LCD blue
SRK II, display. Waterproof key-
SRK/T board. Immersion biom-
etry. Stores 15 IO Ls.

O ptikon Hi-line A-scan Binkhorst, Biometric A-scan is pro-

B-scan, Haigis, vided through 10 MHz
UBM Holladay, biometry probe. The
SRK II, biometry probe can be
SRK/T hand-held, mounted on
a tonometer, or provided
with accessories for
immersion. The software
ensures differential mea-
surements for pseudo-
phakic eyes in relation to
various lenses.
Paradigm P20 A-scan Binkhorst, Maintenance-free,
Medical Hoffer Q , solid tip, lighted probe.
Industries Inc Holladay, Automatic, semi-auto-
SRK II, matic and manual mode
SRK/T operations. Hands-free
operation. Complete
display of scan and intra-
ocular dimensions, with
select screening and
editing. Q uiet, built-in
printer.

(continued)
46 Ch apter 7

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
Paradigm A/B P37 Combo Binkhorst, Customized and user-
Medical Hoffer Q , friendly. Superior Q uad
Industries Inc Holladay, Imaging. Built-in flex-
SRK II, ibility. Standard biometry
SRK/T and diagnostic echogra-
phy. Precision A and B
probe.
Q uantel Medical Aviso B Combo Hoffer Q , AVISO remote control
Holladay, touch screen, 24” Dell
SRK II, desktop or laptop (pre-
SRK/T configured with AVISO
software), 10 MHz B-
Scan probe, foot switch,
power strip, keyboard,
and printer.
Q uantel Medical Axis II A-scan Hoffer Q , Carrying case, biometry
Holladay, probe with extension
SRK II, handle, foot-switch,
SRK/T Prager immersion shell,
printer with cable, power
strip.
Q uantel Medical Compact 2B Combo Hoffer Q , Ten MHz B Probe,
Holladay, power strip, foot switch,
SRK II, integrated keyboard /
SRK/T trackball/mouse, 9”
cube monitor and video
printer.
Q uantel Medical Cinescan B Combo Hoffer Q , Built-in monitor, 10 MHz
or S Holladay, B probe, foot switch,
SRK II, integrated keyboard /
SRK/T trackball/mouse, video
printer. Precision biom-
etry software, probe,
extension handle (for
applanation), Prager
immersion shell, printer
S = Standardized A-scan
by Karl C. O ssoinig, MD.

(continued)
 Available A-Scan In strumen tation 47

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
Sonogage EyeScan A-scan Holladay, Solid tip probe 10 MHz
SRK II, transducer, automatic
SRK/T pattern recognition with
tonal alert, internal dot
matrix printer. Accuracy
± 0.034 mm claimed.
Sonomed Inc Microscan A-scan Binkhorst, Live A-scan display. Five
100 A+ Holladay, different examination
Hoffer Q , modes. Measure review
SRK II, capability. Clinical accu-
SRK/T racy ± 0.1 mm. O ptional
printer. Two probe styles
available.
Sonomed Inc PacScan A-scan Binkhorst, Touch screen operation.
300AP Hoffer Q , Large, high-resolution,
Holladay, backlit LCD. Live A-scans
SRK II, display. Storage of 5 dif-
SRK/T ferent user profiles. Five
different examination
modes. Measurement
review capability.
Immersion capabilities.
Clinical accuracy ± 0.1
mm. O ptional printer.
Sonomed Inc A5500 A-scan Binkhorst, Three different probes.
Hoffer Q , Four different modes.
Holladay, Personalized A-constants
SRK II, and surgeon factors.
SRK/T Clinical accuracy ± 0.1
mm. Built-in calibration
cylinder.
Sonomed Inc A/B5500 Combo Binkhorst, Incorporates all of the
Hoffer Q , features of the A and B
Holladay, units with advantage of
SRK II, compactness.
SRK/T

(continued)
48 Ch apter 7

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
Sonomed Inc Combo Binkhorst, Incorporates all of the
Hoffer Q , features of the A and B
Holladay, units with advantage of
SRK II, being housed in a laptop
SRK/T, computer.
Storz CompuScan A-scan Binkhorst Company was purchased
O phthalmics LT Biometric II, by B+L and no longer
Holladay, sells A-scans.
SRK II,
SRK/T
Storz CompuScan Combo Binkhorst, Company was purchased
O phthalmics AB Hoffer Q , by B+L and no longer
Holladay, sells A-scans.
SRK/T
Tomey AL-100 A-scan Haigis, Measures axial length
Corporation Haigis and calculates IO L
optimized power. Easy to use touch
Hoffer Q , screen. Automatic tone-
Holladay, assisted measurement.
SRK II, Compact and lightweight.
SRK/T Built-in printer. O ptional
memory card. Contact or
immersion modes.
Tomey AL 2000 A-scan Haigis, Measures axial length,
Corporation Haigis corneal thickness and
optimized calculates IO L power.
Hoffer Q , Ultrasound technology
Holladay, provides precise mea-
SRK II, surements. Easy to use
SRK/T touch screen maximizes
efficiency. Compact and
lightweight. Contact
and immersion mode.
Multipoint pachymeter
map. Wide measurement
range. Solid state probes.

(continued)
 Available A-Scan In strumen tation 49

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD

Manufacturer Instrument Type Software Features

Formulas
Tomey AL 3000 A-scan Haigis, Measures axial length
Corporation Hoffer Q , and calculates IO L
Holladay, power. Easy to use
SRK II, touch screen maximizes
SRK/T efficiency. Contact and
immersion mode. Wide
measurement range.
Solid state probes.
Tomey UD 1000 & Combo Haigis, High resolution annu-
Corporation 6000 Haigis opt lar array B-scan probe.
Hoffer Q , Automated video record-
Holladay, ing. Area and distance
SRK II, measurement. Touch
SRK/T screen. Memory card.
O ptional diagnostic A-
scan. A-scan (UD 6000)
measures axial length
and calculates IO L
power.
Wuxi Kangming KN-1800 A-scan SRK I,SRK Solid probe 10 MHz with
Medical Device II, SCDK, focus light. Equipped
Co (China) & KO RA with special hardware
and software to compose
Callan myopia forecast-
ing and forewarning
system.
Some of the above instruments may no longer be available for new purchase but may be
available on the secondary/used /refurbished market.
The reference to the Holladay formula m eans the Holladay 1 formula; no instrum ent has
the Holladay 2 formula installed at this tim e.
(continued)
50 Ch apter 7

Ta b le 7 -1 co n t in u e d .

ULTRASO UND INSTRUMENT SPECIFICATIO NS FO R A VARIETY O F A-SCANS

FROM ARO UND THE W O RLD
Optical (Non-US) Instruments for AL Measurement

Manufacturer Instrument Type Software Features

Formulas
Zeiss Meditec IO L Master O ptical Haigis, Advantages of laser
Jena, Germany Haigis L, optical biometry in a
Hoffer Q , non-contact technique:
1999 Holladay, Measuring precision and
SRK II, repeatability are high. It
SRK/T is fast and easy for the
patient.
Measures AL, K, ACD,
CD.
Haag-Sreit LENSTAR LS O ptical Hoffer Q , Most recent pub-
Koeniz, Switzer- 900 Holladay, lished studies demon-
land SRK II, strate equivalence of
SRK/T the LENSTAR to the
2009 IO LMaster in accuracy
and repeatability. It is not
as fast, but in one pass
it measures: AL, K, CT,
AQ D, LT, CD, RT.
 Available A-Scan In strumen tation 51

Figure 7-1. Immersion method; probe dipped

into O ssoinig shell.

Figure 7-2. Contact or applanation method.

The authors of th is book emphasize the importance of u sing the im m ersion method
(Fig. 7-1) rather than the contact or applanation m ethod s (Fig. 7-2). That is the reason there
is no chapter on applanation in th is book. It is important to know wh ich tech n iques the
instru ment you are considering can p erform, otherw ise it m ay not enable you to step up
to the im m ersion method.
One of the option s w ith A-scan s is the type of probe that is u sed and whether the
tran sducer is focu sed or non-focu sed. It appears to m ake sen se that focu sing the sou nd
wave at a d istance of 23 to 24 m m to reach the retina wou ld be a good idea. However,
Ossoin ig strongly recom mend s a non focu sed tran sducer becau se the sou nd wave m ay
be focu sed for a 23.5 m m eye but the wave w ill be a converging wave in a shorter eye
and a d iverging one in a longer eye. Th is w ill create a greater error than is seen w ith a
nonfocu sed tran sducer.

N ote: In form ation on sp ecifications were obtained from resp ective m anu factu rers.
 8

Im me rsio n Usin g
t h e Pra g e r Sh e ll
Thomas C. Prager, PhD, M PH

Appla na tio n Bio m e try

C
ontact, or applanation biometry, begins w ith the u ltrasou nd probe being placed
d irectly on the su rface of the cornea. In contrast, the im mersion tech n ique utilizes
a liqu id interface between the eye and the u ltrasou nd probe. Wh ile d irect appla-
nation on the cornea seem s simple enough and is probably u sed more often than the
im m ersion tech n ique, contact biometry is the least accu rate method of measu ring axial
length. When trying to place the probe tip on the center of the cornea, even experienced
tech n ician s m ay encou nter parallax problem s, resu lting in a measu rement that is off
axis by sm all or large amou nts. Rem em ber that the eye is the longest when m easu ring
th rough the center of the cornea. Second ly, when attempting to center the probe d irectly
on the cornea, there w ill always be som e degree of corneal compression, esp ecially in
patients w ith reduced intraocu lar pressu re. Short eye measu rem ents as a resu lt of corneal
compression associated w ith the applanation tech n ique w ill read ily reduce the accu racy
requ ired for presbyopia-correcting IOLs.

53
Hoffer KJ. IOL Power (pp. 53-62).
© 2011 SLACK Incorporated.
54 Ch apter 8

The Im m e rs io n Te chnique Us ing

a Fixe d Im m e rs io n She ll
The Prager Shell was designed in 1982, and represented an improvement upon an
im m ersion shell first created by Jackson Colem an, MD.1 The im mersion tech n ique, a one-
handed procedu re that is easy to m aster, elim inates or m in im izes tech n ician variables
(such as corneal compression, align ment of the u ltrasou nd beam, and probe insertion
depth) and lead s to more reproducible resu lts. Each biometer m anu factu rer has sp ecified
an optim al d istance from the cornea for d ata acqu isition; having an im mersion shell w ith
a fixed shelf or a probe auto-stop en su res reliable probe placement. The fixed im mer-
sion tech n ique is much easier to m aster than the open shell tech n ique, wh ich requ ires a
certain level of dexterity in order to position the probe at the appropriate d istance from
the cornea wh ile simu ltaneou sly being perpend icu lar to the retina and d irecting sou nd
waves th rough the center of the cornea and lens. Im mersion biom etry utilizing the Prager
Shell has been reported to be faster than the contact m ethod.2 Aside from ease in getting
on axis, time is saved by not having to review ind ividu al scans for corneal compression
errors.
In the quest of greater accu racy in su rgical outcom e, there have been m any compari-
son s between applanation and im mersion tech n iques first investigated by Hoffer in 1981
and pu blished w ith Sham m as 3 in 1984, who rep orted that im m ersion scan s consistently
resu lt in longer axial lengths and less variability than the contact tech n ique. Th is has
been replicated in m any stud ies 2-10 over the past 30 years. The Prager Shell and other
im m ersion devices have been d irectly compared to the IOLMaster (non-contact coherent
light m ethod of AL determ ination) and no clin ical measu rem ent d ifferences have been
fou nd between the two methodologies, although there is a sign ificant d ifference in the
cost of the equ ipment.2,5,11 Fu rther, becau se the IOLMaster was specifically set to m atch
im m ersion and not applanation, there is a h igh correlation between im mersion biometry
and the IOLMaster u n its in AL measu rements (Pearson correlation coefficient = 0.996).2,12
AL m easu rements u sing the IOLMaster are u nobtainable in 8% to 17% of the cataract
p opu lation due to reduced visual acu ity, corneal and med ia opacities, as well as dense
cataracts.11-16
Thu s, im m ersion biom etry w ill always have clin ical utility.

Co ns ide ra tio n Whe n Us ing

the Fixe d Im m e rs io n She ll
The balance of th is chapter w ill d iscu ss practical h ints and pearls when u sing the fixed
im m ersion Prager Shell and suggestions on reducing measu rement error. It is important
to emphasize that ju st utilizing an im m ersion shell does not gu arantee perfect resu lts in
every patient. There is no substitute for u nderstand ing the sou rces of error when per-
form ing im mersion biometry. One mu st be able to recogn ize an u nu su al scan resu lt and
be interested and compu lsive enough to resolve any apparent d iscrepancy. I often say,
“An inexperienced tech n ician, who does not u nderstand the basic principles u nderlying
AL scan n ing, can u nderm ine the efforts of the most skillfu l cataract su rgeon.”
 Immersion Usin g th e Prag er Sh ell 55

Figure 8-1. When perpendicular to the

retina, A scanspikes are of equal height
and both retinal and scleral heights are
within 80% of one another with no
stair stepping of the retinal spike.

Note that all spikes are of equ al height and both retinal and scleral heights are w ith in
80% of one another w ith no stair stepping of the retinal spike (Fig. 8-1). If one does not see
spikes associated w ith orbital fat, the scan m ay be going th rough the optic nerve and not
the fovea. A B-scan is requ ired when encou ntering a w id e separation between retinal and
scleral spikes wh ich cou ld ind icate edem a. Any eye wh ich does not have a clear cornea
and/ or vitreal med ia w ill requ ire a sonogram .

REVIEW THE CHART PRIOR TO MEASURING THE EYE

Most refractive su rprises occu r in u nu sual length eyes and are more prevalent in short
eyes. Proportionately a 1 m m error in a 20 m m eye has a greater postoperative refrac-
tive con sequence than the same 1 m m error in a 30 m m eye. Although staphylom as m ay
m ake it more d ifficu lt to locate foveal spikes in a long eye, the short eye requ ires greater
m easu rement precision. Prior to an ocu lar measu rem ent it is the respon sibility of the
biometrist to determ ine if the eye is u nu sual in any way. Peru se the chart becau se a previ-
ou s scleral buckle can change the shape of the eye producing a sign ificant AL d ifference
between the eyes. For almost all patients both eyes are approxim ately the sam e length,
typically w ith in 0.3 m m of one another. Replicate m easu rements several times and if a
0.3 m m or greater d ifference rem ain s, note in the chart that the “measu rements exceed
norm al physiological find ings.” To verify th is d ifference in axial length, obtain a B-scan
th rough the optic nerve of both eyes wh ich w ill allow a d irect comparison. Display the
two scans (w ith x-axis grad ation s) one above another, then d raw a line from the optic
nerve of one sonogram dow n to the x-axis and continue to the sam e axis location on the
other echogram (Fig. 8-2). Su btle d ifferences are read ily seen and th is docu mentation
shou ld be included in the patient chart.
Determ ine if either eye is aphakic, as th is w ill requ ire a change in sou nd velocity to
comp ensate for the m issing lens. Sim ilarly, a pseudophakic eye w ill requ ire a change
in tissue velocity. Most biometers have settings for the d ifferent typ es of IOL m aterials
which mu st be know n for an accu rate axial length (see Chapter 4).
56 Ch apter 8

Figure 8-2. Comparing the B-scan

through the optic nerve of both eyes
to verify a reported difference in AL.

SILICONE OIL EYES

Measu ring eyes that contain silicone oil in the p osterior vitreou s is a d ifficu lt task w ith
im m ersion biometry and, in th is author’s exp erience, seldom resu lts in an accu rate AL
m easu rement. Silicone oil in the eye slow s the speed of sou nd th rough the eye, resu lt-
ing in an artifactual long eye measu rement. A fu rther complication is that the two most
com mon types of silicone each have d ifferent tissue velocities, 1050 or 980 m/ s, and the
clin ician mu st d iscern wh ich velocity to select to prevent a measu rement m istake. Sand ra
Frazier Byrne add resses the problem of silicone16 in AL measu rem ents and outlines a
 Immersion Usin g th e Prag er Sh ell 57

A B

C
Figure 8-3. Prager shell with disposable tubing.

corrective procedu re. Each ocu lar component of the eye, cornea, anterior cham ber, lens,
and posterior vitreou s cavity mu st be measu red ind ividu ally and then su m m ed. Wh ile
the anterior cham ber does not requ ire a m athem atical adju stment, to obtain true values
for the lens, mu ltiply X (the measu red len s th ickness) by 1641/ 1532 and mu ltiply the mea-
su red vitreou s length by 980 (most cases)/ 1532. Final IOL power determ ination is m ade
more con fu sing becau se the index of refraction d iffers in the eye w ith silicone versu s the
norm al intact eye and requ ires the add ition of more refractive p ower. In the silicone oil
eye, the IOLMaster produces the more accu rate measu rem ent.
Carefu lly review the chart for add itional anom alies that can potentially affect the su r-
gical outcome. What is the IOL power requested? If there is an anticipated 2 d iopters d if-
ference or more in the final refraction and the other eye does not requ ire cataract su rgery,
patients m ay not be able to tolerate the an isometropia. Prior to a measu rement u se the
cu rrent glasses prescription to estim ate the anticipated axial length. The average eye is
23.5 m m, so given that 1 m m = ~2.35 D, a 24-m m eye shou ld be rough ly myopic by ~1.2 D.
If the glasses have a hyperopic optical correction of +3.00 d iopters and the AL measu re-
ment is 26 m m, th is shou ld alert the biom etrist that there is a potential m istake.

PRACTICAL TIPS IN OBTAINING IMMERSION ANTERIOR-POSTERIOR

LENGTH MEASUREMENTS
The Centers for Disease Control (CDC) gu idelines requ ire thorough ly soaking the
shell and probe in a beaker of alcohol or hyd rogen peroxide for at least 5 m inutes prior to
m easu ring the eye.16 Su bsequently, allow the im mersion shell to completely d ry or flu sh
w ith BSS, since alcohol can read ily remove corneal epithelial cells (Fig. 8-3). To reduce
the likelihood of transm itting pathogen s from patient to patient, change the con necting
58 Ch apter 8

Figure 8-5. Examination of one-eyed patient.

Figure 8-4. Fixation light on extension pole.

tu bing w ith each patient. Tu bing w ith a check valve prevents reflu x and w ill p erm it the
u se of the sam e bottle of BSS over several patients. A study 17 conducted at 34 ophthalmol-
ogy clin ics showed w ide variability in shell/ probe/ tubing clean liness, w ith on ly 14% of
the centers follow ing CDC gu idelines. Fu ngu s was cu ltu red in 12% of the samples and
m icroorgan ism s associated w ith endophthalm itis or keratitis were fou nd in 53% of the
sites tested. The Prager Shell has a Luer fitting to facilitate changing the tu bing.
If there is an outer plastic sheath for applanation, pu ll it away from the probe. Since the
acqu isition software w ill captu re scans only if the probe (m ain bang) is at the specified
d istance from the cornea specified by each u ltrasou nd m anu factu rer, in sert the probe
into the Prager Shell u ntil it seats at the auto-stop, then gently tighten the external set
screw. Internal centering gu ides hold the probe in place to ensu re probe perp end icu larity
w ith in the shell. Routine topical anesthesia is ad m in istered to each eye. Seat the patient
w ith their head tilted slightly back again st a cou nter or u se a reclin ing exam ination chair.
A fixation light on a flexible stem facilitates th is biometry (Fig. 8-4) but m ake su re that it
is far enough from the patient’s eyes to avoid convergence, wh ich increases the d ifficu lty
of locating the fovea du ring biometry. The patient’s attention is natu rally d raw n to the
fixation light, wh ich is beneficial if you do not sp eak their langu age.
When exam in ing the one-eyed patient, prop er fixation can be a problem. A helpfu l
h int is to have the patient extend their arm, m ake a fist and have them stare at their
thu m b (Fig. 8-5). Even if the patient is completely blind, th rough proprioceptive feed back
the eye w ill be able to locate and follow. Always support and move the arm to m in im ize
fatigue.
 Immersion Usin g th e Prag er Sh ell 59

Figure 8-6. Proper placement of the

Prager shell between the lids.

Exam ining Te chnique

Place a towel on the patient’s shou lder; secu re the tu bing and BSS bottle and syringe
to the shell Luer fitting. Ask the patient to look dow nward, toward their feet; then lift the
patient’s upper eye lid and insert the flared rim u nderneath the lid (the upper p ortion of
the shell w ill m ake contact w ith the sclera wh ile the lower part of the shell w ill be held
away from the eye); ask the patient to look straight ahead w ith the u ncovered eye, toward
the fixation light. Pu ll the patient’s lower eyelid dow n and gently pivot the lower portion
of the shell into the lower forn ix, m aking su re by close in spection that it is in the forn ix
and not sitting atop a fold in the conju nctiva. Th is pivotal motion avoid s contact w ith the
cornea and insu res centration of the device arou nd the lim bu s.

HOW TO HOLD THE SHELL

The goal is to put m in im al pressu re on the eye. In fact, it is qu ite in structive for YOU to
be the patient (at least once) and experience firsthand the benefits of a light touch.
Note the Luer filler port is faced temporally (Fig. 8-6). The left hand/ palm is resting on
the forehead (given the biometry in stru m ent is to you r left), and is u sed to reduce shell
pressu re on the eye. Try to keep the A-scan instru ment in you r d irect line of sight. It is
important to position the biom eter screen so that it can be seen easily du ring the proce-
du re. Looking over you r shou lder wou ld need lessly complicate the exam ination. Note
that the palm acts as the fu lcru m or pivot point for the shell. The exam iner m ay want
to stabilize the shell w ith the right hand to m ake m icro-movements. With practice, most
practitioners u sually w ill hold the shell w ith ju st the hand resting on the forehead. The
right hand is free to m ake instru ment adju stments, if necessary.
To m ake a measu rement, slowly inject the saline into the shell. Qu ickly jetting flu id
into the shell m ay remove corneal epithelial cells, resu lting in patient d iscom fort. As
soon as the liqu id fills the shell su fficiently to reach the tip of the probe (about 2 cc), the
60 Ch apter 8

typical waveform s of im mersion biom etry, as previou sly illu strated, w ill be seen on the
screen. Most com m ercial biom eters captu re at least 10 scan s and d isplay an average and a
stand ard deviation. The stand ard deviation, a measu re of variability, shou ld be less than
0.05. If th is value is larger, review the ind ividu al waveform s and delete the outliers. Do
not remove the shell from the patient’s eye du ring th is review, as any measu rem ent that
is deleted w ill need to be im m ed iately replaced w ith a new and acceptable read ing.
To remove the shell from the eye, raise the patient’s upper eyelid, wh ich releases the
top part of the shell from u nder the eyelid. Next, pivot the shell dow nward, d irecting the
patient to continue to look straight ahead. Then pu ll away from the eye w ithout contact-
ing the cornea. Up on the in itial release, the rem ain ing contents of the shell (1 to 2 cc of
liqu id) w ill spill dow n the patient’s cheek. Be prepared w ith a towel or facial tissue.

CRITICAL BIOMETRY TIPS

Although the Prager Shell completely elim inates corneal compression as a complicat-
ing factor, and greatly assists in the align m ent of the probe w ith the m acu la, it is still
necessary to review and analyze waveform s to insu re optimu m read ings.
Again, be su re to accept on ly steeply rising retinal spikes and reject any that have a
stair-step app earance. The corneal, anterior len s, and retinal spikes shou ld be of approxi-
m ately equ al height. If the spikes demonstrate a dow nward trend, th is suggests that the
scan is off-axis. With dense cataracts, the tendency is to increase the gain thereby elevat-
ing the spikes. If the top s of the spikes app ear flattened, th is m ay ind icate that the ampli-
fiers are satu rated, resu lting in an inaccu rate read ing.
With very long eyes, such as in those w ith staphylom a, the m acu la m ay be located
on the sloping portion of the staphylom a and the retinal spike m ay not rise to the same
height as the corneal spikes. However, in norm al eyes, the retinal/ scleral spikes equ al the
height of the corneal spikes. Detection of orbital fat spikes is a requ irem ent. A norm al
scan has a series of orbital fat echoes w ith descend ing amplitudes. If they are absent, or
m arked ly attenu ated, the probe m ay be m isaligned and the biometrist m ay have d irected
the sou nd beam to the optic nerve instead of the fovea.

Sum m ary o f Critical Po ints

l Always m easu re both eyes in every patient.
l Re-m easu re both eyes if:
© There is >0.03 m m d ifference between eyes.
© AL is <22 m m or >25 m m in either eye.
© AL does not correlate well w ith patient’s spectacle refraction.
© There is d ifficu lty obtain ing correctly positioned, h igh, steeply rising echoes that
are not stair-stepp ed. The retina and scleral spikes are w ith in an 80% height of
one another.
© There are problem s w ith patient cooperation or fixation.
© Perform a B-scan to docu m ent the d ifference.
l When doing the second eye later, always measu re the first eye again and m ake a
post-cataract comparison to adju st the su rgeon factor in the IOL formu la.
 Immersion Usin g th e Prag er Sh ell 61

Once the biom etrist is fam iliar w ith the basics d iscu ssed in th is chapter, the learn ing
cu rve is probably 5 to 10 patients and you w ill have m astered the Prager shell im mersion
tech n ique and have confidence in you r measu rements.

Re fe re nce s
1. Colem an DJ, Silverm an RH, Lizzi FL, et al. Ultrasonography of the Eye and Orbit, 2nd ed .
Ph ilad elph ia: Lippincott William s & Wilkin s; 2005.
2. Packer M, Fine IH, Hoffm an RS, Coffm an PG, Brow n LK. Im m ersion A-scan com -
pared w ith partial coherence interferom etry: Outcom es analysis. J Cataract Refract Surg.
2002;28(2):239–242.
3. Sham m as H J. A comparison of im m ersion and contact tech n iqu es for axial length m ea-
su rem ents. J A m Intraocul Implant Soc. 1984;10(4):444–447.
4. Fries U, Hoffm an n PC, Hut WW, Echardt H B, Heu ring A. IOL-calcu lation s and u ltrason ic
biom etry: Im m ersion and contact procedu res. Klin M onatsbl Augenheilkd. 1998;213:162–165.
Germ an. No abstract available.
5. Giers U, Epple C. Com parison of A-scan d evice accu racy. J Cataract Refract Surg.
1990;16(2):235–242.
6. Hoffm an n PC, Hutz WW, Eckhardt H B, Heu ring AH. Intraocu lar len s calcu lation and
u ltrasou nd biom etry: Im m ersion and contact procedu res. Klin M onatsbl Augenheilkd.
1998;213(3):161–165. Germ an.
7. Olsen T, Nielsen PJ. Im m ersion versu s contact tech n iqu e in the m easu rem ent of axial
length by u ltrasou nd. A cta Ophthalmol (Cop en h). 1989;67(1):101–102.
8. Schelen z J, Kam m an n J. Comparison of contact and im m ersion tech n iques for axial length
m easu rem ent and implant p ower calcu lation. J Cataract Refract Surg. 1989;15(4):425–428.
9. Watson A, Arm strong R. Contact or im m ersion tech n ique for axial length m easu rem ent?
Aust N Z J Ophthalmol. 1999;27(1):49–51.
10. H aigis W, Lege B, Miller N, Sch neider B. Comparison of im m ersion u ltrasou nd biom etry
and partial coherence interferom etry for intraocu lar len s calcu lation accord ing to H aigis.
Graefes A rch Clin Exp Ophthalmol. 2000;238(9):765–773.
11. Narvaez J, Cherwek DH, Stu lting RD, et al. Comparing im m ersion u ltrasou nd w ith
partial coherence interferom etry for intraocu lar len s power calcu lation. Ophthalmic Surg
Lasers Imaging. 2008;39(1):30–34.
12. H asem eyer S, Hugger P, Jonas JB. Preop erative biom etry of cataractou s eyes u sing partial
coherence laser interferom etry. Graefes A rch Clin Exp Ophthalmol. 2003;241(3):251–252. No
abstract available.
13. Rajan MS, Keilhorn I, Bell JA. Partial coherence laser interferom etry vs conventional u ltra-
sou nd biom etry in intraocu lar len s power calcu lations. Eye. 2002;16(5):552–556.
14. Teh ran i M, Kru m m enauer F, Ku m ar R, Dick H B. Com parison of biom etric m easu rem ents
u sing partial coherence interferom etry and applanation u ltrasou nd . J Cataract Refract
Surg. 2003;29(4):747–752.
15. Verhu lst E, Vrijghem JC. Accu racy of intraocu lar len s p ower calcu lation s u sing the Zeiss
IOL m aster. A prosp ective study. Bull Soc Belge Ophtalmol. 2001;(281):61–65.
16. Centers for Disease Control. Recom m end ation s for preventing p ossible tran sm ission of
hu m an T-lymphotroph ic viru s typ e III/ lymphad enopathy-associated viru s from tears.
M orb M ortal W kly Rep. 1985;34(34):533–534. No abstract available.
17. Velázquez-Estades LJ, Wanger A, Kellaway J, H ardten DR, Prager TC. Microbial contam i-
nation of im m ersion biom etry u ltrasou nd equ ipm ent. Ophthalmology. 2005;112(5):e13-e18.
62 Ch apter 8

Sug g e s te d Re a ding s
Kah n H A, Leibow itz H M, Gan ley JP, et al. The Fram ingham Eye Study. I. Outline and m ajor
prevalence find ings. A m J Epidemiol. 1977;106(1):17–32.
Fou r costliest outpatient procedu res. Hospital Health N ews. 1998;72:32–33.
Steinberg EP, Javitt JC, Sharkey PD, et al. The content and cost of cataract su rgery. A rch
Ophthalmol. 1993;111(8):1041–1049.
Hollad ay JT, Prager TC, Ru iz RS, Lew is JW, Rosenthal H. Improving the pred ictability of intra-
ocu lar len s p ower calcu lation s. A rch Ophthalmol. 1986;104(4):539–541.
Javitt JC, Bren ner MH, Cu rbow B, Legro MW, Street DA. Outcom es of cataract su rgery.
Improvem ent in visu al acu ity and su bjective visu al fu nction after su rgery in the first,
second, and both eyes. A rch Ophthalmol. 1993;111(5):686–691.
Mangione CM, Ph illip s RS, Law rence MG, et al. Improved visu al fu nction and attenu ation
of declines in health-related qu ality of life after cataract extraction. A rch Ophthalmol.
1994;112(11):1419–1425.
Prager TC, Chu ang AZ, Slater CH, Glasser JH, Ru iz RS. The Hou ston Vision Assessm ent
Test (H VAT): An assessm ent of valid ity. The Cataract Outcom e Study Group. Ophthalmic
Epidemiol. 2000;7(2):87–102.
 9
Ax ia l Le n g t h :
Lase r In te rfe ro m e try
Basics of th e IO LMaster
Wolfgang Haigis, M S, PhD

Ze is s IOLMa s te r Optical Bio m e try

T
he IOLMaster by Carl Zeiss Med itec (Fig. 9-1), introduced in 1999, is an all-in-one
device allow ing all measu rem ents necessary for the calcu lation of IOL powers to
be performed w ith one instru m ent.1 It includes an autom atic keratometer to deter-
m ine central corneal cu rvatu res as well as a slit-im age-based setup to m easu re anterior
cham ber depth. The IOLMaster’s most important featu re, however, is the u se of partial
coherence interferometry (PCI)—also named laser Doppler interferometry (LDI), optical
coherence biom etry (OCB), or laser interference biometry (LIB)—to measu re AL. PCI
biometry was develop ed by the Au strian physicist A. F. Fercher,2 who p erformed the first
AL measu rement in vivo in 1986.
All IOLMaster measu rements are noncontact procedu res, easily perform ed and well
accepted by patients. The instru ment’s operating software includes d atabases for IOL and
su rgeon d ata and offers IOL power calcu lations w ith all p opu lar formu las.
The reproducibility for AL m easu rements w ith the IOLMaster is 22 µm . The in stru-
m ent is h igh ly observer-indep endent, as the inter- and intra-observer variabilities show:
AL: 10-12 µm, ACD: 31-38 µm, and corneal rad ii: 11-14 µm (Fig. 9-2).

63
Hoffer KJ. IOL Power (pp. 63-66).
© 2011 SLACK Incorporated.
64 Ch apter 9

A B

Figure 9-1. Zeiss IO LMaster. (A) side view, (B) exam-

iner view, (C) computer keyboard. (Reprinted with
permission from Carl Zeiss.)

Figure 9-2. AL measure-

ment with IO LMaster (soft-
ware version 5.x, compos-
ite signal): AL = 22.80 mm;
SNR = 43.5. The signal-
to-noise ratio (SNR) is a
measure of signal quality:
the higher the better.

Me a s uring AL w ith
Partia l Co he re nce Inte rfe ro m e try
The AL measu ring principle of the IOLMaster is based on du al-beam partial coher-
ence interferometry. The term “coherence” describes the physical property of waves hav-
ing a temporally constant or regu larly varying phase d ifference at every point in space.
Coherence is a necessary requ irement for interference. Partially coherent light rays can
interfere w ith each other if they m eet w ith in their coherence length s.
In the IOLMaster, a laser d iode em itting partially coherent light (coherence length ≈160
µm) in the near in frared (at a wavelength of 780 n m) is part of a Michelson interferom-
eter setup, wh ich (by means of a moving m irror) produces two partial beam s of d ifferent
optical path lengths. The patient fixates onto the light of the laser d iode, thu s offering
h is visu al axis to the measu ring laser. Both partial beam s are reflected at the cornea
 Axial Len g th : Laser In terferometry: Basics of th e IOLMaster 65

and at the retina. An interference signal is obtained, eg, when the optical path length of
the d isplacement of the moving m irror in the Michelson interferometer is equ al to the
optical path length between cornea and retina (ie, the AL of the eye). The optical path
length equ als the geometrical length times the (group) refractive index of the med iu m
th rough wh ich the light travels. The p osition of the interferometer’s moving m irror can
be m easu red very precisely determ in ing the accu racy of the AL measu rement w ith the
IOLMaster.

Co rre latio n Be tw e e n
Optical and Ultra s o und Bio m e try
It was already mentioned that the IOLMaster m easu res along the visual axis of the eye.
To be more precise, it m easu res the optical path length from the anterior corneal vertex
to the retinal pigment epitheliu m (RPE) along the visu al axis. Ultrasou nd, on the other
hand, measu res from the anterior corneal vertex to the internal lim iting mem brane (ILM)
along the eye’s optical axis, becau se it is th is axis wh ich allow s all interfaces to be m et at
right angles thu s producing a good A-scan echogram . Since clin ical experience bu ilt-up
for decades is based on u ltrasou nd biom etry, it was necessary to translate optical AL m ea-
su rem ents into equ ivalent u ltrasou nd read ings. Therefore, the IOLMaster was internally
calibrated again st h igh precision im mersion u ltrasou nd.3 Con sequently, the IOLMaster
d isplays AL resu lts as if it were a h igh ly precise A-scan in stru m ent perform ing segmental
im m ersion m easu rements w ith laser precision.
As a consequence of th is equ ivalence between IOLMaster and im mersion AL, it was
necessary to ad apt the existing IOL constants, wh ich are given by lens m anu factu rers
and u sually meant to be u sed for contact u ltrasou nd. For most of tod ay’s IOLs, optim ized
constants can be dow n loaded d irectly into the IOLMaster from the website of the User
Group for Laser Interference Biometry (ULIB).4

Re fe re nce s
1. H aigis W. Optical coherence biom etry. In : Koh nen T, ed. M odern Cataract Surgery. Dev
Ophthalmol. Basel: Karger; 2002;34:119–130.
2. Fercher AF, Roth E. Ophthalm ic laser interferom eter. Proc SPIE. 1986;658:48 –51.
3. H aigis W, Lege B, Miller N, Sch neid er B. Comparison of im m ersion u ltrasou nd biom etry
and partial coherence interferom etry for intraocu lar len s calcu lation accord ing to H aigis.
Graefe’s A rch Clin Exp Ophthalmol. 2000;238(9):765–773.
4. ULIB, User Group for Laser Interference Biom etry. http :/ / w w w.augen klin ik.u n i-w uerz-
bu rg.de/ u lib/ index.htm (accessed April 1, 2010).
 10

IOLMa s te r
Ex a min a tio n
Wolfgang Haigis, M S, PhD

Ge ne ral Info rm a tio n

1. Take optical m easu rements first, prior to any ocu lar exam ination requ iring eye con-
tact or application of d rops.
2. Check the calibration of the IOLMaster d aily before starting m easu rements on
patients.
3. Adju st the instru ment table, the head rest, and the IOLMaster so that the patient sits
in a relaxed yet stable position.
4. Explain the m easu rem ent procedu re and point out the necessity for the patient’s
head to rem ain in a fixed position w ith no u n necessary eye movements.
5. Let the patient blin k before starting measu rem ents and ask them to focu s stead ily
on the fixation light. Make su re that the patient sees the fixation light.
Wh ile focu sing is not m and atory w ith the IOLMaster for AL measu rem ents, it is
requ ired for keratom etry (K) and ACD measu rements. From software Version 5 onward s,
adju stment aid s are implem ented that signal the quality of focu sing in a traffic-light m an-
ner and autom atically start the measu rement sequence once focu s is optimu m.

67
Hoffer KJ. IOL Power (pp. 67-70).
© 2011 SLACK Incorporated.
68 Ch apter 10

A
B

Figure 10-1. Zeiss IO LMaster in AL mode. (A) Instrument from examiner’s view, (B) AL screen—note
AL readings collected on left side, (C) example of false readings.

Ax ial Le ng th Me a s ure m e nt
1. Start measu rement sequence w ith a focu sed and centered beam (Fig. 10-1A and B).
If signal quality is bad, try decentering and defocu sing.
2. Generally perform at least 5 to 10 measu rements per eye. Th is way, rare artifacts can
easily be identified.
3. If the IOLMaster does not d isplay a mean value, wh ich happ en s when ind ividu al
measu rements d iffer by more than 100 µm (software Versions 4 and older), check
each measu rement trace for consistency.
4. Identify the likely origin of incon sistent read ings (Fig. 10-1C) (for details refer to
the IOLMaster m anu al). The most frequent signal stem s from the RPE. The reason
m ight be a mem brane, retinal detach ment, or mu ltiple signals from the in ner lim it-
ing mem brane (ILM), RPE, choroid, or artifacts (side lobes to m ain signal) from the
instru ment’s laser d iode.
5. If necessary, zoom in and m anu ally sh ift (left or right mou se-click) the m easu re-
ment cu rsor to the correct signal.
 IOLMaster Examin ation 69

A B
Figure 10-2. Zeiss IO LMaster in K reading mode. (A) Examiner adjusting the reflected mires, (B) K read-
ing screen showing green cross-hairs and reflected corneal mires.

Ke ra to m e try
1. Adju st the IOLMaster so that the 6 p eripheral light spots, not the one in the center,
are in focu s (Fig. 10-2A).
2. The eyelid or eyelashes mu st not obscu re these light spots seen in the screen (Fig.
10-2B). If necessary, ask the patient to op en their eye w idely or assist them in m anu-
ally keeping it open.
3. H ave the patient blin k one or more times before triggering the measu rement.
4. Perform 3 con sistent measu rements.
70 Ch apter 10

A B
Figure 10-3. Zeiss IO LMaster in ACD mode. (A) Examiner adjusting the AC box and lens surface, (B)
ACD screen showing anterior slit beam.

Me a s ure m e nt o f Ante rio r Cham be r De pth

1. The im age analysis software hand ling th is measu rement is designed on ly for pha-
kic eyes. A pplying it to pseudophakic eyes would produce erroneous results.
2. Focu s on the reflection of the fixation light (the sm allest dot) (Fig. 10-3A), and m ake
su re it is w ith in the green square. A slit lamp -typ e im age shou ld be visible (Fig. 10-
3B), clearly show ing the anterior su rface of the crystalline lens inside the pupil. The
slit im age of the cornea shou ld be well outlined and clear of external reflexes.
3. After successfu l adju stment, ask the patient to blin k before triggering the measu re-
ment.
As in u ltrasou nd biom etry, the anterior cham ber depth d isplayed is defined from the
anterior corneal vertex to the anterior lenticu lar vertex (ie, includ ing corneal th ickness).
If ind ividu al measu rem ents d iffer by more than 100 µm, no mean value w ill be d is-
played and the measu rement has to be rep eated.
 11

IOLMa s te r in
Difficu lt Eye s
Wolfgang Haigis, M S, PhD

Ax ial Le ng th Me a s ure m e nts

U
n reliable m easu rem ents need not necessarily be w rong. Perform a sequence of
m easu rements and check the repeated occu rrence of “u n reliable values” at the
same position; wh ich, if present, become reliable.
1. Highly hyperopic or myopic eyes: Backscattered laser intensity from these eyes m ay be
too sm all, resu lting in bad signal-to-noise-ratio (SN R) values. Use a trial frame or
have the patient wear h is/ her glasses or contact len ses (CL) du ring m easu rem ent. In
the case of CLs, subtract the center th ickness of the CL from the m easu red AL.
2. Pseudophakic or silicone-filled eyes: Choose the resp ective AL mode for measu rement.
Seldom, u nwanted interferences (artifacts) m ay occu r, especially in pseudophakic
eyes, wh ich show up as add itional peaks some 5 to 8 m m left of the true AL p eak
(Fig. 11-1). Usually, know ing the patient’s refraction and statu s, these p eaks are read-
ily identifiable as artifacts.
3. Cataract eyes, PCO: On-axis opacities m ay render it impossible for the m easu ring
beam to reach the retina. For IOLMaster AL m easu rements, it is not necessary to
have a strictly centered and focu sed beam . A slight decentering of the beam m ay
allow a pathway free of obstructions. Due to the elliptical beam cross-section, the
decentration of the beam m ay be greater in the north and south d irection than the
east or west of the crosshair center.

71
Hoffer KJ. IOL Power (pp. 71-74).
© 2011 SLACK Incorporated.
72 Ch apter 11

Figure 11-1. Additional sig-

nal left of relevant peak:
rare artifact from posterior
IO L surface

4. Detachments, membranes etc. m ay produce signals wh ich can be m istaken for regu lar
ALs.1 Consecutive m easu rements, however, u sually resu lt in signals, wh ich are too
far apart for the IOLMaster’s autom atic pattern recogn ition to allow calcu lation of
an AL average. Also, preferably in you nger ind ividu als, the internal lim iting mem-
brane (ILM) peak m ay som etimes be stronger than the retinal pigment epitheliu m
(RPE) signal. Typically, these signals are som e 0.2 m m apart. If the IOLMaster does
not give a mean AL, check the ind ividu al measu rements for mem branes or RPE/
ILM d iscrepancies.

Ke ra to m e try
l Focu s the in stru ment properly onto the 6 p eripheral light dots. Be aware of eye
lashes or hanging eyelid s, wh ich m ay cau se erroneou s resu lts.
l H ave patient blin k before measu rem ent.
l In case of corneal scars, slightly decenter the instru ment by moving it toward the
eye to measu re beside local scars.
l In a pseudophakic eye, reflection s m ay add itionally be produced at the anterior
su rface of the IOL. If these reflections render a measu rement impossible, slightly
defocu s the IOLMaster (some 1 to 2 m m) to reduce their intensity. The measu rem ent
resu lts w ill not be affected sign ificantly by th is m aneuver.
Be aware of CL wearers, especially among refractive patients. Keratometry resu lts w ill
be w rong if the CL has not been removed long enough before keratometry (hard and rigid
gas-perm eable CLs for 2 weeks; soft CLs for 3 d ays).
Be aware of p ossibly d ifferent keratom eter ind ices when entering external Ks into
the IOLMaster for IOL calcu lation. Make su re that the IOLMaster is set (setup screen on
computer is show n in Fig. 11-2) to the sam e keratometer index as the external in stru ment
provid ing Ks u sed. An exception to th is is if you w ish to rely on the Hoffer Q Formu la,
the IOLMaster mu st be set to an index of 1.3375.

Ante rio r Cham be r De pth Me a s ure m e nts

Focu s the instru ment properly. H ave patient blin k before measu rem ent. The anterior
cham ber depth (ACD) measu rement modu le is not intended for p seudophakic eyes. If
nevertheless applied, resu lts d isplayed for pseudophakic ACDs w ill be erroneou s and
mu st not be u sed.
 IOLMaster in Difficult Eyes 73

Figure 11-2. Setup screen

to adjust keratometer index
in software menu of the
IO LMaster.

Re fe re nce
1. Lege BAM, H aigis W. Laser interference biom etry versu s u ltrasou nd biom etry in certain
clin ical cond ition s. Graefes A rch Clin Exp Ophthalmol. 2004;242(1):8–12.
 12
Ax ia l Le n g t h :
La se r In te rfe ro m e try
Th e LENSTAR LS900 In stru men t
Kenneth J. Hoffer, M D; H. John Shammas, M D;
and Jaime Aramberri, M D

S
ince the introduction of the IOLMaster in 1999, there has been no alternative for
optical biometry u ntil 2009, when H aag-Streit (Kon iz, Sw itzerland) introduced their
ow n optical biometer called the LENSTAR LS900. The question s that need to be
asked are whether th is new instru ment:
1. Can m atch the accu racy and reproducibility of the IOLMaster?
2. Is as easy to u se as the IOLMaster?
3. Is as dep endent upon dense m ed ia and patient fixation ability?
4. Is also sup erior in staphylom atou s eyes and those filled w ith silicone oil?
Though there have been a few previou s stud ies in the EU comparing the biom etry
m easu rements, the authors had access to the first LENSTAR in the US in 2009, and pu b-
lished the report on the IOL power pred iction accu racy comparing the two instru ments.1
The LENSTAR provides the su rgeon w ith all axial parameters of the eye (AL, corneal
th ickness [CT], ACD, and crystalline len s th ickness [LT]) on the visu al axis, measu red
w ith laser interferom etry. At the sam e time, it measu res the corneal cu rvatu re and axis,
wh ite to wh ite d istance (corneal d iam eter [CD]), and pupil d iameter and eccentricity of
the pupil. All measu rem ents are taken in a single align m ent procedu re, thu s improving
the accu racy of the ind ividu al measu rements since it is all done in the sam e sweep.
The measu rem ent d ata is u sed to pred ict an IOL u sing either the internal calcu lator
featu ring stand ard modern formu las (H aigis, Hoffer Q, Hollad ay 1, and SRK/ T) or it can
be sent d irectly to an external IOL calcu lator featu ring 4th generation IOL formu las or
75
Hoffer KJ. IOL Power (pp. 75 -88).
© 2011 SLACK Incorporated.
76 Ch apter 12

Figure 12-1. Measurement setup. LENSTAR LS900

instrument table with head rest and personal com-
puter.

ray tracing based IOL pred iction software. An integrated exp ort interface and a script
langu age tool allow the LENSTAR and its software package (EyeSu ite) to be easily con-
nected to any electron ic med ical record system . DICOM con nectivity is plan ned to be
implemented as soon as the resp ective stand ard s are finalized.

Te chnical Fe a ture s
The instru ment setup con sists of the LENSTAR measu rement u n it and a stand ard
Window s software computer (Fig. 12-1). Th is modu lar concept allow s the u se of the
LENSTAR integrated in varying sizes of its ow n in stru m ent table or as a stand-alone u n it
on an in stru ment table, whereby it fits on a stand ard H aag-Streit slit lamp table after the
slit lamp has been lifted off (Fig. 12-2). One author (Hoffer) has recom mended that they
simply supply the software on an inexpen sive $300 netbook-type computer (Fig. 12-2E),
thu s elim inating the space needed for the computer, mon itor, keyboard, and mou se.
Fu rthermore, it’s p ossible to ru n add itional software like EMR or IOL calcu lation tools on
the same u n it and it can also be part of a network.
The measu rem ent u n it (Fig. 12-2D) consists of an optical m easu rem ent head and a
m echan ical cross slide system to align the device. With in the optical measu rement head
there are comp onents to measu re the AL, CT, ACD, and LT based on laser interferometry
as well as the corneal cu rvatu re, CD, and pupillometry based on d igital im age analysis.

Optical Lo w Co he re nce Re fle cto m e try

The axial measu rements are very reproducible becau se of the optical low coherence
reflectometry (OLCR) tech nology u sed in the LENSTAR. OLCR is a laser interferometric
m easu ring method. It is based on a Michelson interferom eter (Fig. 12-3) and is powered
by a superlu m inescence d iode (SLD), a broad band light sou rce, w ith a spectral w idth of
approxim ately 25 n m, wh ich in tu rn is centered at 820 n m.2 The light is d istributed to the
interferom eter w ith one m easu rement and two reference arm s. Each reference arm con-
sists of an optical single mode fiber, a focu sing len s, a m irror, and a rotation cu be acting
 Axial Len g th : Laser In terferometry: Th e LENSTAR LS900 In strumen t 77

A B

C D

Figure 12-2. Different office setups for the LENSTAR. (A) Full LENSTAR table setup. (B) Slit-lamp setup
with computer components on a side table. (C) LENSTAR unit replacing the slit lamp on its table. (D)
Close-up photo of the LENSTAR measuring unit. (E) LENSTAR mounted on a Haag-Streit stand attached
to a simple laptop computer with the software installed.

Figure 12-3. Schematic of the opti-

cal low coherence reflectometer
measurement setup.

Detection, amplification,
signal processing
78 Ch apter 12

Figure 12-4. A-scan of a

standard cataract patient
measured 5 times. Every
A-scan is derived from
16 single scans, using
advanced digital signal
processing to improve the
signal to noise ratio to bet-
ter penetrate the cataract.

as an optical path length modu lator. The reference beam is retro-reflected from the m ir-
ror back into the fiber. The m easu rement arm consists of a sample arm fiber, polarization
controllers, and several len ses.
The position of the patient’s cornea is mon itored by an in frared camera, and the u ser
adju sts the centering on a live video im age. Becau se of the refractive index changes
at every interface (eg, air to cornea, cornea to aqueou s, etc) the m easu rem ent beam is
reflected back into the sample arm fiber. An interferogram occu rs on ly when the d iffer-
ence between the optical path length s of the reference and the sample arm is less than
the coherence length of the SLD. Mu ltiple scans are collected, d igitally processed, and
d isplayed as a single measu rement on the computer screen con nected to the LENSTAR
measu rement u n it.

Me a s ure m e nts
Axial measu rement of all parts of the hu m an eye is an autom ated objective m easu re-
m ent. All parameters are m easu red along the visual axis as the patient fixates on the mea-
su rement beam and the u n it detects loss of fixation and stops the m easu rement procedu re
u ntil the patient re-fixates on it again. Dep end ing on the age, alertness, and cooperative-
ness of the patient, th is process can vary from very easy to ted iou s or imp ossible, sim ilar
to but more so than the IOLMaster. But in contrast w ith the latter, once the read ing has
been taken successfu lly all the parameters have been measu red at one time and there is
no need to now m easu re the K read ing, ACD, and CD in separate steps.
A u n ique featu re of the LENSTAR is being able to measu re all axial parameters in a
single scan but also to d isplay the d ata in an “A-scan” sim ilar to the d isplays of A-scan
u ltrasou nd biom eters. The abscissa refers to the AL and the ord inate to the signal to noise
ratio (SN R). All measu rem ents are show n in a single graph (Fig. 12-4), allow ing the u ser
to easily assess the qu ality of the repeated m easu rements as well as the autom atic detec-
tion of the ind ividu al structu res w ith in the eye. If the u ser does not agree w ith any of
the autom atic measu rements, they can adju st the gauges in the “A-scan” to measu re the
structu res to their best judgment. User defined measu rements are m arked on the screen
w ith an asterisk.
 Axial Len g th : Laser In terferometry: Th e LENSTAR LS900 In strumen t 79

Parallel to the axial m easu rem ents, the LEN STAR provid es the u ser w ith m easu re-
m ents of the corneal cu rvatu re, CD, and pupil d iam eter, as well as the eccentricity of the
visu al axis w ith resp ect to the pupil and the CD center. The central corneal cu rvatu re
is assessed based on two rings of a total of 32 in frared LED m arkers projected onto
the cornea w ith a 1.65 and 2.3 m m optical zone. Im age an alysis algorith m s are u sed to
d erive the m ain m erid ian s from the reflection s, cap tu red by a h igh-sen sitivity CMOS
cam era. The corneal cu rvatu re is d isplayed either in m m rad iu s of cu rvatu re (r) or D of
corneal p ower, for w h ich the u ser can select h is preferred ind ex of refraction to convert
r valu es into D.
The same im age is u sed to derive the pupil d iameter of the eye at the time of measu re-
m ent. Adju sting the am bient light, th is measu rement featu re m ay be u sed to estim ate the
effectiveness of a mu ltifocal IOL on a sp ecific patient.
The CD can either be analyzed on the in frared im ages taken for the determ ination of
the corneal cu rvatu re or from an optionally taken red-free im age. Th is red-free im age
provides improved contrast as compared to the in frared im age, thu s improving the
repeatability of the measu rement. The h igh contrast red-free im age of the eye m ay also
serve as a base for the preoperative plan n ing of a toric IOL, determ in ing land m arks to
later orient the IOL in the OR.

Me a s ure m e nt Pro ce dure

The concept of having the measu rement u n it and the computer separate allowed the
creation of a graph ical u ser interface. A measu rement w izard gu ides the u ser th rough
the measu rement procedu re and supp orts their interaction w ith the patient by provid ing
clear text messages of how to improve the m easu rem ent resu lts, eg, if the eye lashes of a
patient block m arker points of the keratom etry measu rem ent.
The starting point of every measu rement is the selection of the measu rement mode.
The measu rement modes are phakic, aphakic, pseudophakic, and silicone-filled eye sta-
tu s. In the pseudophakic mode it is possible to select the appropriate IOL m aterial. The
option silicone-filled eye can be com bined w ith any of the other measu rement modes.
If the u ser accidentally forgets to select the correct measu rement mode, it is p ossible to
sw itch again later in the analysis of the “A-scan.”
The next step is the align ment of the device on the patient’s eye. First the u ser coarsely
aligns the m easu rement beam in the center of a cross-hair d isplayed on an overview
live im age of the eye on the computer screen, and then focu ses on the reflections of the
keratom etry m easu rem ent points. When centered properly, the thu m b button is then
pressed on the joystick and a zoom ed live im age is seen and then the u ser p erform s the
fine centering and focu sing—wh ich is supported by on line feed back of the device. As
soon as the measu rem ent position is reached, a green circle is d isplayed (Fig. 12-5). The
size of the circle represents where in the measu rement range the LENSTAR is positioned.
Originally when the circle was large it was red and when med iu m it was yellow. Becau se
th is gave the impression that the exam iner had to keep focu sing u ntil obtain ing the green
(sm allest) circle, one author (Hoffer) recom m ended that they elim inate the red and yellow
colors, wh ich they have done. The sm aller the circle, the better the device is centered in
the m easu rement range. The measu rem ent qu ality is independent of the position of the
device in the measu rem ent range of 3 m m in the d irection away from and closer to the
patient.
80 Ch apter 12

A
Figure 12-5. (A) Measurement wizard and result over-
view at the time measurements are taken. The arrow B
in the measurement screen guides the user to move
the LENSTAR unit away from the patient to center the
device in the measurement range. (B) Close-up of the
screen image of the aiming target.

Figure 12-6. Results screen with cross-

sectional and frontal schematic of the
human eye to educate the patient on the
measurements just taken.

As soon as the measu rement range is reached, one m ay press the joystick button again
to start the m easu rem ent, an autom atic process. Du ring the m easu rem ent, the LENSTAR
detects blin king and loss of fixation. In such cases, the m easu rement procedu re is inter-
rupted and continued after the patient regain s fixation. Du ring the measu rement, a grey
circle starts at the 12:00 position and completes a 360° circle in a clockw ise d irection. Th is
represents the advancem ent of the “A-scan” and biometry d ata acqu isition. The cycle
completes when the 16 scan s and 4 K read ings have been acqu ired. Collecting so much
more in form ation obviou sly m akes th is step slightly longer than w ith the IOLMaster.
Th is was truer w ith the original software (Version 1.04) we u sed, but is much faster w ith
the latest software (Version 1.10). With the latter, a novice u ser took 3 m inutes to obtain a
resu lt wh ile an exp erienced u ser took on ly 90 second s.
The resu lt of every scan is im med iately d isplayed in a su m m ary and also as a sm all
“A-scan.” It is strongly recom m ended to take 5 con secutive measu rements to get a valid
statistical feed back. After all m easu rements on both eyes have been taken, the resu lts
m ay be reviewed on the su m m ary screen (Fig. 12-6). To easily educate the patient on the
 Axial Len g th : Laser In terferometry: Th e LENSTAR LS900 In strumen t 81

measu rements ju st taken, th is screen show s a schem atic cross section and frontal view
of an eye, show ing the ind ividu al measu rements as the operator moves the mou se over
the respective values. Clicking on the values show s the details sim ilar to an A-scan and
enables an easy check of the measu rement qu ality and, if requ ired, m anu al adju stm ent
to the u ser’s best judgment. At th is stage the measu rement procedu re is completed. The
nu m ber of measu rements taken on a single patient per d ay is not lim ited. The light inten-
sity of the SLD em itted to the patient wou ld allow more than eight hou rs of u n interrupted
measu rement w ithout cau sing any harm to the eye.

IOL Calcula tio n

Besides the measu rem ent procedu re, the IOL calcu lation screen is called the EyeSu ite
Biometry (H aag-Streit USA, Mason, Oh io). It provides an integrated IOL calcu lation
modu le. The u ser m ay create their ow n ind ividu al set of IOLs and pred iction formu las.
Per sheet, a com bination of 5 IOL/ formu la com binations is possible. The integrated IOL
calcu lation formu las are H aigis, Hoffer Q, Hollad ay 1, SRK/ T, and (u n fortu nately) SRK II.
The IOL calcu lation modu le m ay be u sed in a network environ ment independently of the
m easu rement u n it. Recalcu lation wh ile in the operating room is also possible.
Apart from the integrated IOL calcu lation, EyeSu ite’s export interface allow s con nec-
tion to external IOL pred iction software such as Dr. Hoffer’s Hoffer Program s® (Su reID,
Mu llica H ill, N J), Dr. Hollad ay’s Hollad ay IOL Con su ltant® (Hollad ay Consu lting Inc,
Bellaire, TX), Dr. Olsen’s PhacoOptics® (IOL In novation s Aps, Aarhu s N, Den m ark), and
Dr. Preu ssner’s Oku lix® (der Leu, H illerse, Germ any). Other software m ight easily be
ad apted to EyeSu ite creating interfaces based on the H aag-Streit script language.

Accura c y Studie s
A clin ical study p erformed at the Un iversity Clin ic Insel, Berne, Sw itzerland by Roh rer
et al3 on 144 eyes fou nd very good rep eatability of the AL measu rement as d isplayed in
Table 12-1, where d ata for patients w ith a complete set of 5 repeat m easu rem ents for both
eyes were included to derive m ean and stand ard deviation (SD) and the coefficient of
variation (CV). For each measu rem ent param eter, the SDs of each set of 5 replicate mea-
su rements were fou nd to be nearly constant over the range of m ean resu lts. Th is find ing
con firm s that the ind ication of one single SD value for each measu rement param eter is a
valid approach. The study not on ly included stand ard cataract patients but also patients
who had previou sly u ndergone lens extraction and/ or vitrectomy. Besides demon strat-
ing excellent rep eatability, the study also showed very good agreem ent of the LENSTAR
m easu rements w ith those of the IOLMaster.
Corneal cu rvatu re was also analyzed in the clin ical study by Roh rer et al. Again, good
rep eatability and agreement w ith the IOLMaster was rep orted. These find ings are in
agreement w ith stud ies by Rabsilber et al,4 Holzer et al,5 Buckherst et al,6 and Cruysberg
et al.7 Even though som e of these stud ies showed statistically sign ificant d ifferences in
some of the parameters measu red between the LENSTAR and the IOLMaster, these d if-
ferences d id not have any sign ificant impact on the IOL calcu lation u sing IOLMaster
optim ized len s con stants from the User Group for Laser Interferom etric Biometry 8 (ULIB)
on both devices.
82 Ch apter 12

Ta b le 1 2 -1 .

D ATA FOR SUBJECTS W ITH A COMPLETE SET OF 5 REPEAT MEASUREMENTS FOR BOTH
EYES TOTAL MEAN, THE STD D EV REPEAT AND THE COEFFICIENT OF VARIATION (CV)
Measurement Mean grand Std Devrepeat CV
AL (mm) 23.973 ± 0.035 0.00145
Corneal thickness (µm) 557.100 ±2.300 0.00407
ACD (mm) 3.190 ± 0.040 0.01220
Lens thickness (mm) 4.560 ± 0.080 0.01784
Corneal curvature (mm) 7.670 ± 0.030 0.00396
Axis of the steep meridian (°) 162.000 ±11.000 0.14191
White to white distance (mm) 12.270 ± 0.040 0.00337

The study by Hoffer, Sham m as & Savin i1 showed excellent comparison between the
LENSTAR and IOLMaster (Table 12-2). Fifty eyes w ith clear lenses and 50 cataractou s
preop erative eyes were measu red p erform ing the exam on on ly one eye of each patient.
Len s th ickness cou ld not be compared becau se it is not m easu red by the IOLMaster so
we compared it to ou r previou s u ltrasou nd stud ies: 600 eyes 9 and 1000 eyes10 show ing
a 0.11 m m th in ner read ing w ith u ltrasou nd. The IOL power pred iction u sing the H aigis
formu la showed practically no clin ical d ifference between the two instru ments.

Co m paring the Te chnical Tim e Invo lve d

Us ing the IOLMa s te r vs the LENSTAR
IOL power calcu lation dem and s the input of d ifferent variables depend ing on wh ich
formu la is u sed. Th ird generation formu las on ly u se AL and K,11-14 wh ile 4th generation
formu las ask for more: ACD, LT, CD, etc.15,16 Measu rement of these parameters has been
simplified by the m arketing of devices that can measu re all or most of them by optical
tech nologies.17
Beyond precision and accu racy, the tim e it takes to perform the measu rem ent is a
relevant issue that m atters in such frequently performed procedu res. Faster devices are
p ositively valued by u sers that tend to consider faster m ach ines more u ser-friend ly. Sp eed
is dep end ant upon software and hardware design and therefore sign ificant d ifferences
can be fou nd among the presently m arketed optical biom eters such as the IOLMaster
(Zeiss), LENSTAR LS900 (H aag-Streit), Biograph (Wavelight, Erlangen, Germ any), OA
1000 (Tomey, Nagoya, Japan), etc.
Aram berri performed a comparative study of the work time of m easu rem ent of the
IOLMaster (Version 5.4.3) and the LENSTAR LS900 (Version 1.3.0).
The IOLMaster m easu res AL by Partial Coherence Interferometry (PCI), the ACD by
slit im age analysis, the K by autom ated keratom etry, and the CD by im age analysis. Each
 Axial Len g th : Laser In terferometry: Th e LENSTAR LS900 In strumen t 83

Ta b le 1 2 -2 .

BIOMETRY FO R 50 EYES W ITH C ATARACT AND 50 EYES

W ITH A C LEAR LENS COMPARING LENSTAR W ITH IO LMASTER
Measurement LENSTAR IOLMaster Range of Diff
Biometry of 50 Eyes with Clear Lenses
AL (mm) 23.72 ±1.21 23.70 ±1.20 -0.05 to +0.11
ACD (mm) 3.10 ± 0.41 2.95 ± 0.39 -0.24 to +0.40
Corneal power K (D) 43.41 ±2.13 43.53 ±2.13 -0.80 to +0.33
Biometry of 50 Eyes with Cataracts (Pre-op)
AL (mm) 23.71 ±1.04 23.68 ±1.04 -0.05 to +0.14
Corneal thickness (µm) 557.100 ±2.3 NA NA
ACD (mm) 3.11 ± 0.47 2.98 ± 0.49 -0.22 to +0.55
4.72 ± 0.47 *US 4.63 ± 0.68
Lens thickness (mm) NA
3.76 to 6.50 **US 4.63 ± 0.46
Corneal power K (D) 43.58 ±1.87 43.69 ±1.92 -0.61 to +0.34

IOL Power Calculation Prediction Error (Haigis Formula) 50 Eyes

Mean Error (ME) -0.002 ± 0.56 +0.003 ± 0.55
Mean Absolute Error (MAE) 0.455 ± 0.32 0.461 ± 0.31
Range of Error -1.06 to +1.38 -1.17 to +1.13
Error ± 0.50 D 58% 56%
Error ±1.00 D 94% 94%
Error ±1.50 D 100% 100%

*Hoffer US study of 600 eyes9

**Shammas US study of 1000 lenses10

parameter is m easu red ind ividu ally and the u ser has to go consecutively th rough d iffer-
ent screen s u sing a laptop -style touchpad, focu sing the target w ith a joystick and clicking
to obtain each measu re. In AL mode each click obtain s 1 m easu rement. Th is software
version dem and s at least 5 measu rements to calcu late an average resu lt w ith a compos-
ite SN R value. After th is the u ser enters into the K mode, focu ses, and w ith one click 3
measu rements are obtained. Then the ACD mode is selected, and the im age is focu sed
and another single click takes 5 m easu rem ents. Then the CD mode can be entered, and,
as in AL mode, each click obtain s 1 m easu rem ent that has to be con firm ed once the u ser
checks it in the screen. It is very easy to u se and the process of focu sing in each mode is
fast, taking ju st a few second s even for inexp erienced u sers.
84 Ch apter 12

The LENSTAR LS900 measu res all optical interfaces from anterior cornea to retinal
pigment epitheliu m by means of Optical Low Coherence Reflectometry (OLCR). The
eye is segmented in the same way u ltrasou nd A-scan does, so the u ser find s a fam iliar
graph ic d isplay w ith p eaks at each optical med iu m bou nd ary. The m easu ring process is
very simple as all measu rements are obtained from ju st one click. The u ser clicks tw ice
to start focu sing on a zoom ed im age of the pupil center. Then another click starts the
m easu ring phase that takes some second s u ntil the resu lts are d isplayed on the screen.
To get the CD, another click is needed to process the d ata, wh ich takes another couple of
second s. The LENSTAR LS900 is considered as u ser-friend ly as the IOLMaster by u sers
and allow s correct m easu ring the first time a new exam iner tries it.

Clinical Co m paris o n
Both devices were compared in a clin ical setting w ith the aim of determ in ing wh ich
u nit is the fastest in the best case scenario. Sixteen eyes of 16 hospital staff and non-cata-
ract patients were m easu red by two exp ert tech n icians u sed to working w ith all types of
ophthalm ic d iagnostic equ ipment. There is a sou rce of bias, since they have had 6 years of
experience w ith the IOLMaster and on ly 1 month w ith the LEN STAR. The patients were
free of any ocu lar d isease and were able to m aintain fixation for the exam ination.
Fou r situ ations were defined for comparison :
1. 1 m easu rem ent of AL+K+ACD
2. 1 measu rement of AL+K+ACD+CD
3. 3 m easu rements of AL+K+ACD
4. 3 m easu rem ents of AL+K+ACD+CD
CD was on ly included in 2 cases, as th is is a param eter on ly u sed by the Hollad ay 2
formu la and m any su rgeons don’t u se it routinely. One m easu rem ent was selected as
reference and 3 m easu rem ents as a com mon strategy u sed by m any su rgeon s to average
the final parameters, although we are aware that 5 m easu rem ents is recom mended to
ensu re precision.

Re s ults
The resu lts of the time measu rements can be seen in Table 12-3 and Fig. 12-7. The
IOLMaster m easu res consecutively AL, K, and ACD in 15.56 second s (±1.36). If CD is
added another 3.5 second s are needed, on average. Scaling the nu m ber of m easu rements
to 3 on ly means a little increase in time as on ly AL and CD measu rem ents shou ld be
repeated w ith th is software version. Th ree AL, K, and ACD m easu rem ents are obtained
in 22.06 second s (±1.69) wh ile 32.38 second s (±1.93) are necessary to add CD to the resu lts
set. The LENSTAR can m easu re faster if on ly 1 measu rem ent is done. It takes 12.19
second s (± 0.98) for AL, K, and ACD and 14.19 second s (± 0.98) if CD is also m easu red.
However, increasing the nu m ber of measu rements m ean s rep eating the entire process,
so th is m ean value is mu ltiplied by the nu m ber of measu rem ents to be p erformed. Th is
p enalizes th is device, m aking it much slower than the IOLMaster in a mu ltiple measu re-
m ent setting. The h igher the nu m ber of measu rem ents, the greater the d ifference between
them. All d ifferences were sign ificant (p <0.001) w ith the Wilcoxon test (Table 12-4).
 Axial Len g th : Laser In terferometry: Th e LENSTAR LS900 In strumen t 85

Ta b le 1 2 -3 .

DESCRIPTIVE STATISTICS OF TIME (IN SECONDS) NEEDED TO MEASURE AL, K, ACD AND CD
Instrument Measures N Mean SD Min. Max.
3 AL+K+ACD 16 22.06 ±1.69 20.00 27.00
3 AL+K+ACD+CD 16 32.38 ±1.93 29.00 38.00
IO LMaster
1 AL+K+ACD 16 15.56 ±1.36 14.00 20.00
1 AL+K+ACD+CD 16 19.06 ±1.61 17.00 24.00
3 AL+K+ACD 16 36.50 ±2.50 32.00 42.00
3 AL+K+ACD+CD 16 42.50 ±2.50 38.00 48.00
LENSTAR
1 AL+K+ACD 16 12.19 ± 0.98 11.00 15.00
1 AL+K+ACD+CD 16 14.19 ± 0.98 13.00 17.00

N = number of eyes, SD = standard deviation

Figure 12-7. Mean time of measure-

ment. IO LM = IO LMaster; 1 or 3
measurements; W = CD parameter
measured. Increasing the number of
measurements makes LENSTAR slower
than IO LMaster.

Ta b le 1 2 -4 .

WILCOXON SIGNED RANK TEST WAS USED FOR PAIRED COMPARISON AMONG VARIABLES

Test Statistics 3 AL+K+ACD 3 AL+K+ACD+CD 1 AL+K+ACD 1 AL+K+ACD+CD

Z -3.526 -3.523 -3.493 -3.457

Asymp. Sig. (2-tailed) 0.000 0.000 0.000 0.001
86 Ch apter 12

It´s interesting to observe that variance is sm aller w ith the LENSTAR when on ly 1
m easu rement is done, and th is is not affected if CD is added to the d ata set. However,
repeating measu rem ents increases d ispersion more than w ith the IOLMaster. Analyzing
(case to case) show s that the m ain contributor to the IOLMaster variance is ACD measu re-
m ent, whose focu sing process can take more time than the one needed for getting AL,
K, or CD.

Co nclus io n
Both devices are equally fast and u ser friend ly. However, some d ifferences can be
show n. The LENSTAR is faster for one m easu rem ent, but clearly becom es slower than the
IOLMaster as more m easu rements are obtained. However, the m agn itude of 10 to 20 sec-
ond s per eye is not clin ically relevant, even in a h igh volu me setting. In order to improve
these nu m bers, m anu factu rers shou ld work toward decreasing the focu sing time as well
as the nu m ber of measu rem ents requ ired when p erform ing mu ltiple measu rements.

Sum m ary
Measu rement of all optical structu res in the hu m an eye:
l AL
l Corneal th ickness (CT)
l ACD (aqueou s depth (AQD) + CT)
l Lens th ickness (LT)
l Corneal cu rvatu re (m ain merid ians and axis p osition)
l Corneal Diameter (CD)
l Pupillom etry
Eccentricity of the visual axis w ith respect to CD and the pupil center:
l Complete “A-scan” of the eye
l All measu rements taken simu ltaneou sly on the visu al axis
l Autom atic m easu rement procedu re w ith fixation and blin king control
l Interactive measu rem ent w izard
l Integrated IOL calcu lation tool
l Interface to EMR system s and th ird party IOL calcu lation software
Networkable w ith other in stru m entation and software (DICOM)

Re fe re nce s
1. Hoffer KJ, Sham m as, Savin i G. Comparison of two optical biom eters. J Cataract Refract
Surg. 2010;36(4):644-648.
2. Ballif J, Gianotti R, Chavan ne P, et al. Rapid and scalable scans at 21 m/ s in optical low-
coherence reflectom etry. Opt Lett. 1997;22:757-759.
3. Roh rer K, Frueh BE, Wälti R, et al. Comparison and evalu ation of ocu lar biom etry u sing
a new noncontact optical low-coherence reflectom eter. Ophthalmol. 2009;116:2087-2092.
 Axial Len g th : Laser In terferometry: Th e LENSTAR LS900 In strumen t 87

4. Rabsilber TM, Jep sen C, Au ffarth GU, Holzer MP. Intraocu lar len s p ower calcu lation :
Clin ical comparison of 2 optical biom etry d evices. J Cataract Refract Surg. 2010;36:230-
234.
5. Holzer MP, Mamu sa M, Au ffarth GU. Accu racy of a new partial coherence interferom etry
analyzer for biom etric m easu rem ents. Br J Ophthalmol. 2009;93:807-810.
6. Buckhu rst PJ, Wolffsoh n JS, Shah S, et al. A new optical low coherence reflectom etry
d evice for ocu lar biom etry in cataract patients. Br J Ophthalmol. 2009;93:949-953.
7. Cruysberg LP, Doors M, Verbakel F, et al. Evalu ation of the LENSTAR all-in-one non-con-
tact biom etry m eter. Br J Ophthalmol. 2010;94:106-110.
8. User Group for Laser Interference Biom etry, Un iversity of Wü rzbu rg. http :/ / w w w.
augen klin ik.u n i-w uerzbu rg.de/ u lib/ (accessed April 1, 2010).
9. Hoffer KJ. Axial d im en sion of the hu m an cataractou s len s. A rch Ophthalmol. 1993;111:914-
918, erratu m 1626.
10. Sham m as H J. A-scan biom etry of 1000 cataractou s eyes. In : Ossoin ig KC, ed . Ophthalmic
Echography. Dord recht, the Netherland s: Ju n k Pu blishers; 1987:57-63.
11. H offer KJ. Th e H offer Q for m u la: A com p arison of theoretic a nd reg ression for m u las.
J Cataract Refract Surg. 1993,19:700-712, errata 1994:20:677 and 2007;33:2-3.
12. Retzlaff JA, Sanders DR, Kraff MC. Developm ent of the SRK/ T intraocu lar lens implant
p ower calcu lation formu la. J Cataract Refract Surg. 1990;16:333-340, erratu m 1990;16:528.
13. Hollad ay JT, Praeger TC, Chand ler TY et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14:17-24.
14. H aigis W. IOL calculation according to Haigis. 1997. http :/ / w w w.augen klin ik.u n iw uerz-
bu rg.de/ u slab/ ioltxt/ haie.htm (accessed on April 1, 2010).
15. Olsen T, Oleson H, Th im K. Pred iction of p ostop erative intraocu lar len s cham ber d epth.
J Cataract Refract Surg. 1990;16:587-590.
16. Hollad ay JR. Ach ieving em m etropia in extrem ely short eyes w ith two piggyback poste-
rior cham ber intraocu lar len ses. Ophthalmol. 1996;103:1118-1123.
17. Find l O, Drexler W, Menapace R, et al. H igh precision biom etry of p seudophakic eyes
u sing partial coherence interferom etry. J Cataract Refract Surg. 1998;24:1087-1093.
 13

Co rn e a l Po w e r:
Dio pte rs Ve rsus Ra d ius
Wolfgang Haigis, M S, PhD

De finitio n o f Co rne al Po w e r in Gaus s ian Optic s

A
bout two-third s of the total refractive power of the eye is provided by the cornea.
In the most basic approxim ation (Equation 1), the cornea can be represented by a
meniscu s lens w ith refractive index n, confined anteriorly and posteriorly by spheri-
cal su rfaces w ith rad ii Ra (anterior) and Rp (posterior) and separated by d (corneal thick-
ness). With n1 = 1.000 being the refractive index of air and n2 = 1.336 the refractive index of
aqueou s, each corneal su rface can be characterized by its su rface powers Da and Dp:
n − n1 n2 − n
(1) Da = and Dp =
Ra Rp
In paraxial (Gau ssian) optics (Fig. 13-1), the refractive power of th is len s can be
expressed either as total (equ ivalent) power De or as back vertex power Dv given by
Equ ation 21:
d De Da
(2) De = Da + Dp − Da Dp and Dv = = Dp +
n d d
1− Da 1− Da
n n
With the resp ective d ata of the Gu llstrand eye, the refractive p owers of the Gu llstrand
cornea in Table 13-1 are obtained.

89
Hoffer KJ. IOL Power (pp. 89 -92).
© 2011 SLACK Incorporated.
90 Ch apter 13

Figure 13-1. Cornea in paraxial approxima-

tion.

Ta b le 1 3 -1 .

D ATA FO R THE CO RNEA O F THE G ULLSTRAND EYE

Ra (mm) Rp (mm) D (mm) n Da (D) Dp (D) De (D) Dv (D)
7.7 6.8 0.5 1.376 48.83 –5.88 43.05 43.83

Me a s ure m e nt o f Co rne al Po w e r
There is no instru m ent to d irectly measu re the d ioptric power of the cornea.
Topograph ic instru ments, keratom eters, and ophthalmom eters do not measu re d iopters.
Topograph ic instru ments determ ine rad ii of cu rvatu re from elevation m aps, and kera-
tometers and ophthalmom eters derive the anterior corneal rad iu s of cu rvatu re from the
size of the fist Pu rkinje reflex of a lu m inant object in the in stru ment.
The translation from the measu red rad iu s Ra in m illimeters into a refractive p ower K
(K-read ing) in d iopters is done by in stru m ent-specific formu las and keratometer ind ices
n9. Among them are the refractive, axial, and in stantaneou s formu las.2 The in stantaneou s
is u sed most and given by Equ ation 3:
n'−1
(3) K=
Ra
R
a
wh ich has the form of a su rface power (Equation 1). Differences exist in the nu meri-
cal value of the keratometer index n9. Javal-Sch iötz-typ e keratometers and all topogra-
phers u se n9 = 1.3375, whereas Zeiss instru ments in Eu rope prefer a value of 1.332. In
 Corn eal Pow er: Diopters Versus Radius 91

com m ercial in stru m ents, n9 values from 1.3315 to 1.338 can be fou nd. The Javal calibration
(1.3375) is close to the refractive index of the tear film, aqueou s, and vitreou s (1.336), for
wh ich Helm holtz had derived a value of 1.337. Th is value was converted to 1.3375 to give
a K = 45.00 D for a corneal rad iu s of 7.50 m m, merely for conven ience. The same rad iu s
produces K = 44.20 D w ith a keratometer index of 1.3315.
Accord ing to Gu llstrand, the true refractive index of the cornea is 1.376 (see Table 13-1).
The K value obtained from Equation 3 w ith th is value is equ ivalent to the anterior su rface
power.

Spe cial Me aning o f

Ke ra to m e te r Indice s 1 .3 31 5 and 1 .3 3 7 5
The cornea’s total (equ ivalent) power (De) as well as its back vertex power (Dv),
accord ing to Equation 2, depend on its front and back su rface p owers. But keratometry
and topography instru ments measu re on ly the anterior rad iu s of cu rvatu re, Ra. Without
knowledge of the posterior rad iu s (Rp), neither corneal power De nor Dv can be calcu-
lated. However, if it can be assu med that anterior and p osterior rad ii have a fixed ratio
Rp/Ra, then it is possible to derive the back su rface properties from the anterior su rface
d ata. If th is ratio is given by the Gu llstrand eye (see Table 13-1), ie, Rp/Ra = 6.8/ 7.7 = 0.883,
then it can be show n 1 that Equ ation 3, w ith an index of 1.3315, gives the total (equ ivalent)
power De, w ith an index of 1.3375, the back vertex power Dv of the cornea.
So w ith rad ii given in m illimeters and refractive powers in d iopters, the follow ing for-
mu las hold for Gu llstrand-like eyes w ith a ratio Rp/Ra = 6.8/ 7.7 (Equations 4 and 5):
337.5
(4) Back vertex power: K =
Ra
331.5
(5) Total equ ivalent power: K =
Ra
An imp ortant consequence of Equ ation s 4 and 5 is in effect in eyes after corneal refrac-
tive su rgery: here, the ratio Rp/Ra has been deliberately altered. Therefore, K-read ings
based on Equ ations 4 and 5 have no physiological or optical mean ing.
The corneal rad ii of a patient measu red w ith d ifferent keratometers or top ographers
shou ld be the same or d iffer on ly slightly, whereas the corneal p ower m ay be exp ected to
vary sign ificantly for the d ifferent instru ments, dep end ing on the respective keratometer
index u sed.

Re fe re nce s
1. H aigis W. Corneal p ower after refractive su rgery w ith myopia: the contact len s m ethod.
J Cataract Refract Surg. 2003;29(10):1397–1411.
2. Roberts C. The accu racy of “power” m aps to d isplay cu rvatu re d ata in corneal top ogra-
phy system s. Invest Ophthalmol Vis Sci. 1994;35(9):3525–3532.
 14
Co rn e a l Po w e r:
Ma n u a l Ke rato me try
a n d In stru m e n t atio n
Kenneth J. Hoffer, M D

M
anual keratometry is a m ethod to determ ine the true central optical p ower
of the cornea by m easu ring the rad iu s of cu rvatu re of the anterior su rface. It
m akes assu mptions regard ing the p osterior su rface (wh ich can not be m easu red
by the instru ment) and then converts the rad iu s into d iopters (D) by a simple equ ation:
D = 337.5/ r
where D = d iopters, r = rad iu s of cu rvatu re.
Argu ments rage as to what the index of refraction shou ld tru ly be, wh ich is not the
su bject of th is chapter.
The original in stru m ent for measu ring corneal power m anu ally is called a keratom-
eter or an ophthalmom eter. The principle by wh ich the in stru m ent measu res the anterior
corneal cu rvatu re is contingent upon accu rately determ in ing the size of a reflected im age
from the front su rface of the cornea or, as it is called, the first Pu rkinje-San son im age.
In m anu al keratometers the exam iner mu st align and focu s the illu m inated m ires (Fig.
14-1), as well as mod ify their position to get the measu rements needed. All keratom eters
m easu re the corneal power of the central approxim ately 3 m m area (from 2.8 to 3.2 m m
in u noperated corneas). A problem arises in post-refractive corneal su rgery eyes becau se
the central 3 m m area is so much flatter that the im age measu red is larger than 3 m m,
cau sing an error in obtain ing the true “central cu rvatu re.”
The keratometer projects an im age (see Fig. 14-1) onto the cornea and then measu res
the size of that im age reflected from the corneal su rface. The device then converts im age

93
Hoffer KJ. IOL Power (pp. 93-96).
© 2011 SLACK Incorporated.
94 Ch apter 14

Figure 14-2. Manual keratometer showing the

measuring ball in place on the magnetic holder for
Figure 14-1. Illuminated mires of the keratom- calibration.
eter that are projected onto the cornea.

size into corneal rad iu s u sing simple vergence relationsh ips of convex m irrors. Becau se
of the sm all but continuou s movem ent of the patient’s eyes, the keratom eter dou bles the
reflected im age and measu res the im age against itself, rather than again st a fixed scale.
There are several com m ercial models, all sharing some com mon featu res:
1. There are two illu m inated m ires (objects) of know n size, located at a set d istance
from the corneal plane.
2. An im age duplicator (dou bling device) to m ake it easier to measu re a moving tar-
get.
3. An optical system wh ich includes a telescope w ith a short focal length. Th is enables
the observer to see the resu ltant im age (wh ich is sm all and virtual).
The variou s models are d ivided into those based on the Helm holtz design and those
based on the Javal-Sch iotz design.
1. In the Helmholtz model, the rad iu s of cu rvatu re is determ ined from a know n object
size (m ires) and the size of the object is fixed (separation between the illu m inated
m ires). The size of the im age is variable and is obtained by a variable dou bling
device, a d isc w ith fou r circu lar ap ertu res. One of the ap ertu res has a prism that
d isplaces the im age in a horizontal plane, wh ile another apertu re has a prism that
d isplaces the im age in the vertical plane. The variabilility of the system is based
upon the possibility of d isplacing the d isc follow ing the axis of the horizontal and
vertical prism s u ntil an adequ ate im age is obtained.
2. The Javal-Schiotz is based on a fixed im age, where the dou bling system is not vari-
able, but the size of the object is variable.

The Ex a m ina tio n

The patient’s ch in is placed in the ch in rest of the in stru m ent (Fig. 14-2) and the fore-
head is pressed against the top for fixation. The keratometer hou sing is then aim ed at
the eye to be m easu red and the ch in rest and the vertical setting on the hou sing is then
 Corn eal Pow er: Man ual Keratometry an d In strumen tation 95

Figure 14-3. Set of calibration balls

for the B+L keratometer.

adju sted, after loosen ing the knobs, so that the reflected im age is cast on the center of the
cornea. It is often helpfu l to sh ine a p en light th rough the ocu lar to help grossly align it
on the eye. The knobs are then tightened to lock it in place. The horizontal and vertical
im ages are then aligned by rotating the gauges on the left and right of the hou sing u ntil
they are lined up precisely. The read ings of corneal power are then taken from the hori-
zontal and vertical gauges. If the values are exactly the same, there is no astigm atism. If
there is astigm atism, it is necessary to rotate the whole hou sing so that the in stru ment is
aligned in the axis of the cylinder and then the gauges are readju sted. The cylinder axis
is read from the gauge.

Im po rtant Re m inde rs
It is prudent to establish a routine of calibrating m anu al keratom eters u sing the set of
steel calibration balls supplied by the m anu factu rer (Fig. 14-3). Find ing where they are
stored m ay be a problem . When record ing corneal p ower in IOL calcu lation it is w ise
to always record the average K, thu s ignoring the cylinder wh ich has no effect in these
calcu lations. Th is elim inates transcription errors. If one can show by past clin ical stud ies
that a con stant change in the corneal power occu rs follow ing you r cataract tech n ique, it
m ight be w ise to add that effect to all you r preoperative K read ings. If not it shou ld be
ignored. I have not been convinced yet that autom ated keratometers are as accu rate on a
consistent basis as m anu al keratom etry.
Lastly, it can be a med icolegal problem if contact len ses have not been left out of the eye
continuou sly for 2 weeks prior to the keratom etry exam for IOL p ower. I obtain patient
coop eration w ith th is fact by removing on ly one contact lens for that p eriod of tim e.

N ote: Sp ecification s on in stru m ents were obtained from the resp ective m anu factu rers.

Sug g e s te d Re a ding
Hollad ay JT. Stand ard izing con stants for u ltrason ic biom etry, keratom etry, and intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1997;23(9):1356–1370.
 15
Au to m a te d Ke ra to m e try
fo r IOL Po w e r
Ca lcu la tio n
Jaime Aramberri, M D

A
utom ated keratometers (Fig. 15-1) measu re anterior corneal cu rvatu re, provid ing
K values (K) that IOL calcu lation formu las w ill u se for IOL p ower calcu lation. In
th is process most formu las first u se the K, w ith other pred ictor variables, to pre-
d ict the IOL plane position estim ation (effective lens position [ELP]). Then the K is u sed
again in the optical calcu lation vergence formu la to solve for IOL p ower.
The m ain advantage of autom ated over m anu al keratometry is simplicity: joystick
focu sing and two clicks allow accu rate and rep eatable measu rem ents in a very short time
(Fig. 15-2).
Moreover, hand held autokeratometers (Fig. 15-3) allow the measu ring of patients in
d ifferent p ostu res—wh ich can be u sefu l for exam ination of ch ild ren u nder general anes-
thesia, d isabled or mentally retarded patients, etc.
Other advantages of autokeratometry are d isplayed in Table 15-1.

97
Hoffer KJ. IOL Power (pp. 97-106).
© 2011 SLACK Incorporated.
98 Ch apter 15

Figure 15-1. A table-mounted autokeratometer.

Figure 15-1. A table-mounted autokeratometer.

Figure 15-3. A handheld autokeratometer.

Ta b le 1 5 -1 .

ADVANTAGES O F AUTO KERATOMETRY

l Examiner-proof method: short learning curve
l Similar accuracy to manual keratometry and videokeratoscopy
l Higher repeatability than manual keratometry or videokeratoscopy
l Fast measuring time: 0.02 to 0.1 seconds
l Eccentricity measurement
l Handheld devices allow measuring non cooperating patients
l Printing of results
l Data interface for a computer network
 Automated Keratometry for IOL Pow er Calculation 99

Figure 15-4. Setup screen to adjust

the corneal index of refraction in the
software menu of the IO LMaster.

Optic s
Cu rvatu re measu rement is based on convex m irror paraxial optics. A know n size
object is projected onto the cornea from a know n d istance and the reflection produced in
the precorneal tear film is measu red. Magn ification is related to rad iu s of cu rvatu re by a
simple paraxial formu la:
r = 2 d h´/ h
where r = rad iu s of cu rvatu re, d = object to im age d istance, h´ = im age size, and h =
object size.
As measu ring speed is h igh (<0.1 sec), im age dou bling system s u sed in m anu al kera-
tometers are u n necessary.1 The object m ires are d ifferent in frared light patterns dep end-
ing on each m ach ine design: rings, dots, etc.

Me a s ure m e nt
Rad iu s of cu rvatu re is u sually expressed in m m. Som e m ach ines describe the cornea
as a spherical toric su rface, measu ring the steep and flat axis rad ii wh ich are forced to
be 90º apart. Others describe the central cornea as an aspherical toric su rface add ing an
asphericity coefficient to the apical rad ii.
Conversion from m m to power in D is perform ed by the paraxial formu la:
P = n1 - n0 / r
where P = power in D, n1 = the index of refraction of cornea, n0 = the index of refrac-
tion of air and r = the anterior corneal rad iu s of cu rvatu re.
Most keratometers u se by defau lt the keratometric stand ard index of refraction of
1.3375 as n1. Th is is to comp en sate for the non measu red power of the posterior cornea.
Some u se other arbitrary indexes, ie, Zeiss u ses 1.332. Modern autokeratometers allow
changing th is param eter by software (Fig. 15-4).
100 Ch apter 15

The 1.3375 value has been u sed in keratometry for h istorical reason s, but it is know n
to induce an overestim ation of central paraxial power of about 0.75 to 1.00 D. However,
the most u sed IOL calcu lation formu las (H aigis, Hoffer Q, Hollad ay, SRK/ T) are designed
to input that nu m ber, correcting it internally. It is essential for the u ser to know wh ich
index is u sed as it w ill affect the calcu lated K read ing. A very simple, and strongly advis-
able, way to avoid conversion errors is getting u sed to inputting K read ings in rad iu s of
cu rvatu re in m m, elim inating the conversion altogether.

Me a s ure d Are a
Although most m anu als declare a 3 to 3.3 m m d iam eter measu red corneal zone, th is
figu re depend s on the cu rvatu re and asphericity of the cornea. In th is way autokeratom-
eters behave like variable doubling m anu al keratometers, where object size is fixed. The
steeper the corneal anterior su rface, the sm aller the area reflecting the keratometer m ires.
An average d iameter range extend s from 2.8 m m (for a 7 m m rad iu s of cu rvatu re) to 3.7
m m (for a 9 m m one).2 Th is has clear implications after corneal refractive su rgery where
steep and prolate or flat and oblate corneas are com mon.

Pe rfo rm ance
The accu racy of autokeratometers has been reported to be at least equ al to m anu al
keratom etry and videokeratoscopy for K value determ ination. Rep eatability is generally
accepted to be better w ith autokeratom etry in norm al eyes. Douthwaite et al3 reported
95% lim its of agreement: ± 0.033 m m for central rad iu s and ± 0.068 for p value measu ring
con icoid al plastic tests w ith the Topcon KR-3500 (Topcon Med ical System s, Oakland, N J).
H am m ack 4 showed sim ilar resu lts measu ring steel calibration spheres w ith the Bau sch +
Lom b (B+L) m anu al keratometer, Nidek ARK-2000 (Nidek Inc, Fremont, CA), Hu mph rey
410 (Carl Zeiss, Dublin, CA), and Alcon Renaissance (Alcon Laboratories, Fort Worth,
TX) autokeratometers. Average and stand ard deviation of errors were lower than 0.10
D in most cases. Edward s et al5 proved better rep eatability w ith the Nidek KM-500 and
Top con RK-3000A than w ith the B+L m anu al keratometer. Coefficients up to 0.05 m m for
the autom ated devices and up to 0.12 m m for the m anu al were rep orted. Giraldez et al6
fou nd no sign ificant d ifference in central power among the EyeSys 2000 (EyeSys Vision
Inc, Hou ston, TX) , Javal, and Nidek ARK-700 (m aximu m mean d ifference 0.25 D). Ninety-
five p ercent lim its of agreem ent range were clin ically sign ificant: ± 0.5 D. Su nderraj et al7
rep orted a 0.11 D mean d ifference between the Canon RK-1 (Canon USE, Lake Success,
N Y) and the Top con OM-4 m anu al keratometer. Davies et al8 compared the Javal kera-
tometer w ith the N Vision-K 5001 (Sh in Nippon, Ajinomoto Inc, Tokyo, Japan), find ing no
bias in central rad iu s and ± 0.22 m m 95% lim its of agreement for the vertical merid ian.
There was a sign ificant d ifference in astigm atism axis; on ly 45% were w ith in ±10º of the
Javal read ing. Lam et al9 rep orted sim ilar K read ings between erect and supine measu re-
m ents, as long as the hand held keratometer (Nidek KM-500) was correctly aligned w ith
the eye. In th is study, d ifference in central power between topography and autokeratom-
etry was not clin ically sign ificant: 0.24 ± 0.78 D. Santodom ingo-Ru bido et al10 compared
the Zeiss IOLMaster keratom etry w ith the Javal show ing a mean d ifference of -0.03 m m
w ith 0.13 m m 95% con fidence lim its. They showed w ith the EyeSys videokeratoscop e a
mean d ifference of -0.06 m m w ith 0.07 m m 95% con fidence lim its. In th is work, repeat-
ability of autokeratometry after an interval of 1 to 10 d ays was excellent: m ean d ifference
was 0.00 m m w ith 0.04 m m 95% con fidence lim its.
 Automated Keratometry for IOL Pow er Calculation 101

Autokeratom etry has been described to be reliable and not affected by applanation
tonom etry 11 or gon ioscopy. In comparison, videokeratoscopy Sim K needed 20 m inutes to
retu rn to baseline values after gon ioscopy both w ith Gold m an n and Su ssm an lenses.12

S o urce s o f Erro r
The exam iner can not evalu ate m ire reflection irregu larity as in m anu al keratom etry.
Moderate irregu larity is u ndetected and can induce m easu rement error. If irregu larity is
h igher, the device simply doesn’t take the read ing. Whenever an irregu lar corneal su rface
is su sp ected, videokeratoscopy shou ld be p erform ed to con firm corneal shap e.
After corneal refractive su rgery K read ings w ill be erroneou s, as the corneal ante-
rior/ posterior ratio becomes a fu nction of the induced shap e change. Th is mu st be com-
p ensated for if keratom etry (m anu al or autom ated) is u sed for K determ ination for IOL
calcu lation in these eyes.

Co nclus io n
Autom ated keratometry is a fast, easy-to-u se, accu rate, and precise m ethod for corneal
central rad iu s m easu rement. It is advised to u se K values in m m instead of D to avoid
conversion errors. A list of available autokeratometers is provided in Table 15-2.

Re fe re nce s
1. Rabbetts RB, Mallen EAH. Measu rem ent of ocu lar d im en sion s. In : Rabbetts RB, ed.
Bennett and Rabbett’s Clinical Visual Optics, 4th ed. Boston : Butterworth-Heinem an n; 2007.
2. Stone J. Keratom etry and sp ecialist optical in stru m entation. In: Ru ben M, Gu illon M, ed s.
Contact Lens Practice. London: Chapm an and H all Med ical; 1994.
3. Douthwaite W, Pard han S. Accu racy and rep eatability of the Top con K-3500 autokeratom-
eter on calibrated convex su rfaces. Cornea. 1995;14(3):253–257.
4. H am m ack GG. Evalu ation of the Alcon Renaissance hand held autom ated keratom eter. Int
Contact Lens Clin. 1997;24(2):59–65.
5. Edward s MH, Cho P. A new hand-held keratom eter: Com parison of the Nidek KM-500
auto keratom eter w ith the B&L keratom eter and the Topcon RK-3000A keratom eter. J Br
Contact Lens A ssoc. 1996;19(2):45–48.
6. Giraldez MJ, Yebra-Pimentel E, Parafita A, et al. Comparison of keratom etric values of
healthy eyes m easu red by Javal keratom eter, Nid ek autokeratom eter and corneal analysis
system (EyeSys). Int Contact Lens Clin. 2000;27(2):33–40.
7. Su nd erraj P. Clin ical comparison of autom ated and m anu al keratom etry in pre-op erative
ocu lar biom etry. Eye. 1992;6(1):60–62.
8. Davies LN, Mallen EAH, Wolffsoh n JS, et al. Clin ical evalu ation of the Sh in-Nippon
N Vision-K 5001/ Grand Seiko WR-5100K autorefractor. Optom Vis Sci. 2003;80(4):320–324.
9. Lam AKC, Chan R, Chiu R. Effect of postu re and artificial tears on corneal power measu re-
ments w ith a hand held autom ated keratometer. J Cataract Refract Surg. 2004;30(3):645–652.
10. Santodom ingo-Ru bido J, Mallen EAH, Gilm artin B, Wolffsoh n JS. A new non-contact opti-
cal device for ocu lar biom etry. Br J Ophthalmol. 2002;86(4):458–462.
11. Beatty S, Nischal KK, Jones H, Eagling EM. Effect of applanation tonom etry on m ean
corneal cu rvatu re. J Cataract Refract Surg. 1996;22(7):970–971.
12. George MK, Ku riakose T, DeBroff BM, Em erson JW. The effect of gon ioscopy and kera-
tom etry and corneal su rface top ography. BM C Ophthalmol. 2006;6:26.
102 Ch apter 15

Ta b le 1 5 -2 .

COMPILATIO N O F THE VARIO US AUTOMATED KERATOMETERS

AVAILABLE O N THE MARKET
Manufacturer Model Description
Canon, Inc RK- F1 Auto-Refractor-Keratometer Choose between refraction
and R-F10 Full Auto-refractor only or with keratometry.
O ne touch does it all: from
automatic alignment to
precision output. Trackball
replaces joystick. Motorized
chinrest. Ergonomic controls.

Haag-Streit O M-900 Uses a classical distance-

independent measuring
principle, which meets the
requirements of Helmholtz.
Two different test patterns
(Javal or Cross mark) can be
used. Can be user-specifically
calibrated, thus eliminating
subjective systematic measur-
ing errors.

Nikon Speedy-K Auto Refract Keratometry and refrac-

Instruments, Inc Keratometer tion readings in 0.35 sec
each. Peripheral readings
in 0.45 sec each. Medical
Retro Illumination Mode. A
100 patient memory. Wider
reading range. Easy posi-
tion adjustment. Video jack.
Stopper lever. A 2.5 mm
Pupil Diameter Plus Light
Intensity Control. Auto shut
off.

(continued)
 Automated Keratometry for IOL Pow er Calculation 103

Ta b le 1 5 -2 co n t in u e d .

COMPILATIO N O F THE VARIO US AUTOMATED KERATOMETERS

AVAILABLE O N THE MARKET
Manufacturer Model Description
Nikon Retinomax K-plus 2 Handheld Auto-measurement and auto-
Instruments, Inc Auto Refract Keratometer finish. Selectable reading
modes. Fast Ref-Keratometric
readings. Shorter reach-
ing distance. Wider reading
range. Melody function. Mire
ring. Super quick mode. Retro
mode key. Peripheral corneal
curvature reading. A 2.5 mm
pupil size and selectable
fixation intensity. Anti-theft
design. Flexible configuration.

Marco KM-500 The unique viewing and

O phthalmic, Inc alignment window lets you
observe the targeted eye
from any distance. View the
eye through the window and
move the instrument towards
the eye. When the mire ring
appears on the cornea, slow-
ly move the unit closer. The
KM-500 will fire automati-
cally when properly focused.
Readings appear instantly on
the LCD display. Both left and
right readings display simul-
taneously. 1.5 lbs. O ptional
printer. Battery good for one-
hour continuous use.

Marco ARK 700A and ARK 760A Refractor/keratometer with

O phthalmic, Inc high-speed mode for read-
ings. Takes 180 separate reti-
noscopy readings. Auto-fog-
ging to prevent accommoda-
tion. IO L button. Automatic
alignment, focusing, reading,
printing. User-friendly on-
screen prompts.

(continued)
104 Ch apter 15

Ta b le 1 5 -2 co n t in u e d .

COMPILATIO N O F THE VARIO US AUTOMATED KERATOMETERS

AVAILABLE O N THE MARKET
Manufacturer Model Description
Marco PALM- AR and PALM – ARK Portable. Easy to align.
O phthalmic, Inc Automatic measurement
and fogging. Accurate and
dependable readings. Infrared
printer communication.
Rechargeable battery.

Reichert KR460 Auto Keratometer/Refractor Hands-free alignment tech-

O phthalmic nology. No joystick or chin-
Instruments rest adjustments needed.
Provides consistent, fast,
accurate readings.

Tomey RC4000 Auto Refractor- Multiple functions:

Corporation Keratometer Refraction, keratometry,
Contact lens base curves, PD.
Data output: Built-in printer.
Fast and accurate measure-
ments. Power saving feature.

(continued)
 Automated Keratometry for IOL Pow er Calculation 105

Ta b le 1 5 -2 co n t in u e d .

COMPILATIO N O F THE VARIO US AUTOMATED KERATOMETERS

AVAILABLE O N THE MARKET
Manufacturer Model Description
Topcon Medical KR-8000PA SUPRA Auto-Kerato Combines an auto-refractor,
Systems, Inc Refractometer an auto-keratometer and a
computerized color corneal
mapping system in a single
compact unit. Rotary prism
technology. Auto-track-
ing and auto measurement
enhance ease of operation.
Incorporated placido rings for
enhanced testing precision,
providing corneal mapping
measurements out of 10 mm.

Zeiss Meditec KR-7000S Auto Kerato- Provides accurate objective

Refractometer refraction, thorough yet sim-
ple subjective refinement and
keratometry measurements all
in one. O nline capability to
compare refraction to actual
spectacle, using computer-
ized lensometer.

Zeiss Meditec HARK Model 599 Full featured glare testing.

Automatic eye tracking.
Immediate patient verifica-
tion of prescription. Two dif-
ferent configurations fit any
space (operator control panel
can be positioned at either
90 or 180 degrees from the
patient). Complete testing.
Automatic keratometric read-
ings. Communicom Interface.
(continued)
106 Ch apter 15

Ta b le 1 5 -2 co n t in u e d .

COMPILATIO N O F THE VARIO US AUTOMATED KERATOMETERS

AVAILABLE O N THE MARKET
Manufacturer Model Description
Zeiss Meditec Humphrey Acuitus Flat screen color display.
Models 5000 and 5015 Thirty measurements per
second. Auto XYZ. Expanded
refracting capabilities. Auto-
acquisition. Full screen eye
image. Illuminated icon con-
trol buttons. Chinrest and
built-in printer. Keratometry
only available on model
5015.
Zeiss Meditec IO L Master Automatic K readings taken
optically. Setup screen
allows adjustment of index
1999 of refraction used. If not set
to 1.3375, the Hoffer Q for-
mula will not report accurate
results.

Haag-Sreit LENSTAR LS900 Automatic K readings taken

Koeniz, Switzer- by laser. Setup screen allows
land adjustment of index of refrac-
tion used. It is not as fast, but
2009 in one pass it measures:
AL, K, CT, AQ D, LT, CD, RT

Som e of the instrum ents listed in this table may no longer be available for new purchase,
but may be available on the secondary/ used/ refurbished market.
 16
Co rn e a l Po w e r:
Co rn e a l To p o g rap h y fo r
IOL Po w e r Ca lcu la t io n
Jaime Aramberri, M D

C
orneal topography is the m easu rement of corneal shap e and optical fu nction. It
can be u sefu l for IOL power calcu lation in d ifferent ways:

l Determ ination of corneal central cu rvatu re and power for the regu lar IOL calcu la-
tion formu las
l More complex IOL calcu lation u sing exact ray tracing method s
l Measu rem ent of corneal spherical aberration to calcu late aspheric IOLs
l Corneal astigm atism m easu rement for toric IOL calcu lation
Other u ses in IOL su rgery:
l Astigm atism corneal su rgery plan n ing and postoperative mon itoring
l Corneal d iagnosis for postoperative excimer su rgery ind ication

Te chno lo g ie s
Th ree d ifferent tech nologies can be fou nd in com mercial devices (Fig. 16-1).

107
Hoffer KJ. IOL Power (pp. 107-114).
© 2011 SLACK Incorporated.
108 Ch apter 16

Figure 16 -1. Reflexion topography

(Placido videokeratoscopy) measures
corneal curvature and calculates
elevation. Projection tomography
measures elevation and calculates
curvature.

PLACIDO TOPOGRAPHY
Convex m irror optics are u sed to measu re the size of the im age reflected by the cor-
nea of a know n object at a know n d istance. The object is u su ally a set of concentric rings
called a Placido d isc (Fig. 16-2) so that the flatter the cornea the w ider the rings. Cu rvatu re
recon struction depend s on the algorith m program med in each top ographer. The first
models u sed a keratom etric analysis method, calcu lating shape the way a keratometer
does; that is, measu ring ring m agn ification.1 Modern topographers u se d ifferent arc step
algorith m s where a continuou s sequence of arcs is traced from the vertex to the periph-
ery of each sem i-merid ian. Th is method allow s simu ltaneou s calcu lation of cu rvatu re
and height.2 The advantage of the latter is that peripheral slope and cu rvatu re, as well as
height, are more accu rate than w ith keratometric-like algorith m s that incorrectly assu me
a spherical geom etry (spherically biased). Central cornea measu rement (Sim K) is sim ilar
w ith both reconstruction m ethod s as long as the shap e is sphero-cylind rical.3

SCANNING SLIT TOMOGRAPHY

The Orbscan IIz (B+L, Rochester, N Y) com bines Placido d isk and scan n ing slit tech nol-
ogy for corneal topography. Forty light slits are projected at 45º angles in 1.5 second s scan-
n ing tim e. Im aging is perform ed by a camera aligned w ith the corneal vertex. Su rface
elevation is measu red, and by d ifferentiation, cu rvatu re can be calcu lated. Anterior cor-
neal topography can be calcu lated both from scan n ing slit or from Placido in form ation
(Fig. 16-3).
 Corn eal Pow er: Corn eal Topog raph y for IOL Pow er Calculation 109

Figure 16 -2. Placido disk image showing the

reflected rings.

Figure 16-3. Axial maps of the same eye with O rbscan IIz. Left map is generated from
Placido data; superior and inferior information are lost due to Placido disk configuration.
Right map is generated from scanning slit data.

Figure 16 -4. Example of a Pentacam

Scheimpflug image of the anterior segment.

SCHEIMPFLUG TOMOGRAPHY
The Scheimpflug camera u ses a laterally im aging light projected at 0º angle (Fig. 16-4).
1. The Pentacam (Ocu lu s, Wetzlar, Germ any) is a rotating camera that scan s the ante-
rior segment in 2 second s taking 50 slits (up to 100 w ith the H R model.) It m easu res
elevation and calcu lates cu rvatu re.
110 Ch apter 16

Figure 16-5. Instantaneous (left) and axial (right) maps of the same eye using the Keraton.
Central curvatures are similar. Beyond 1.5 to 2 mm radial distance, instantaneous map
shows flatter and more variable values, making the map noisier.

2. The Galilei (Ziem er, Port, Sw itzerland) has two rotating cameras scan n ing the ante-
rior segment in 2 second s taking 60 slits (Fig. 16-5). The advantage of the second
camera is comp en sation of any corneal decentration th rough the im age acqu isition
process. As does the Orbscan IIz, the Galilei has a bu ilt-in Placido d isk. Corneal
cu rvatu re is calcu lated by merging the Placido and elevation d ata by a proprietary
algorith m .
Both Scan n ing Slit and Scheimpflug instru ments measu re anterior and posterior cor-
neal cu rvatu re. Th is has the potential advantage of calcu lating central corneal p ower w ith
paraxial or exact formu las instead of assu m ing p osterior corneal cu rvatu re like Placido
topographers do. Although most of the modern IOL calcu lation formu las have not been
ad apted to th is possibility yet, som e new formu las have been pu blished in th is way4 and
there is some software that calcu lates IOL power u sing exact ray tracing (Oku lix5).

Curvature and Po w e r Maps

Cu rvatu re can be expressed as axial rad iu s of cu rvatu re, where the center of cu rvatu re
of each point is forced to be at the optical axis. It can also be expressed as the instanta-
neou s rad iu s of cu rvatu re, where th is con straint is not applied. The m anu al keratometer
m easu res the axial rad iu s of cu rvatu re at approxim ately a 1.6 m m rad ial d istance.6
The d ifference between axial and in stantaneou s rad iu s of cu rvatu re is not sign ificant
in the paraxial zone (Fig. 16-6C). Beyond the 2 m m rad ial d istance, instantaneou s rad iu s
of cu rvatu re describes corneal shap e more accu rately, giving a flatter read ing in a nor-
m ally prolate cornea.
From axial and in stantaneou s cu rvatu re (in m m), keratometric power (in D) is cal-
cu lated by means of a paraxial refractive formu la and a keratometric stand ard index of
refraction of 1.3375 (see Chapter 14 for a fu ll explanation). It is important to u nderstand
that th is is a m easu re of shape and not of power. As a paraxial formu la is u sed, a sphere
w ill show the same power in the center as in the periphery (same color in the m ap), wh ile
due to Snell’s law, power shou ld increase in the periphery (hotter color than in the center)
to express spherical aberration. Most videokeratoscop es calcu late a refractive or Snell
m ap where optical power of the anterior cornea is d isplayed.
 Corn eal Pow er: Corn eal Topog raph y for IOL Pow er Calculation 111

A B

Figure 16 -6. Ziemer Galilei instrument: (A) from

examiner view, (B) from patient view, (C) elevation
map, and (D) Scheimpfug photos of anterior seg- D
ment.

Ce ntral Co rne al Po w e r/ Ra dius

As modern IOL calcu lation formu las were designed to input keratometric K read ings,
most top ographers calcu late a simu lated K measu rem ent they call the Sim K, wh ich is
calcu lated from the 3 m m ring analysis, d isregard ing more central cu rvatu re d ata. Some
devices calcu late other central ind ices, averaging central corneal cu rvatu re: Effective
Refractive Power 7 (EyeSys), Average Corneal Power 8 (Tomey), etc.
112 Ch apter 16

Ta b le 1 6 -1 .

NO RMAL EYES TO PO GRAPHIC D ATA

Author’s database: SimK and asphericity measured with EyeTop (CSO ) topographer;
Anterior/posterior ratio measured with Pentacam.

SimK 43.45 ±1.37

Asphericity (Q 4.5 mm) -0.12 ± 0.13
Asphericity (Q 8 mm) -0.27 ± 0.12
Anterior/Posterior 1.21 ± 0.02

Accura c y and Pre cis io n

Placido videokeratoscopy has accu racy and precision w ith in ± 0.25 D measu ring test
spheres.9 Douthwaite10 fou nd that the EyeSys overestim ated apical rad iu s 0.11 m m w ith
± 0.02 m m 95% con fidence lim its when m easu ring ellipsoid al convex su rfaces. Spherical
su rfaces were more accu rately measu red than aspherical. Repeatability of vertex rad iu s
was 0.014 m m. Lam et al11 compared the Med mont E300 w ith a hand held autokeratom-
eter rep orting 0.24 D m ean d ifference in K value w ith ± 0.78 D and 95% lim its of agree-
m ent. Potvin et al12 reported inter-observer rep eatability between 0.25 and 0.50 D both
for TMS1 (Tom ey) and EyeSys topographers. Giraldez et al13 compared EyeSys, Nidek
ARK700, and Javal and reported ± 0.12 m m w ith 95% lim its of agreem ent between EyeSys
and Nidek and 0.17 m m between EyeSys and Javal. Dave et al14 rep orted sim ilar repeat-
ability for EyeSys; ± 0.072 D and for B+L m anu al keratometer, ± 0.103 D. The 95% lim its of
agreement between both m ethod s was ± 0.34 D.

To po g raphy fo r IOL Calcula tio n

If the cornea is w ith in norm al lim its of cu rvatu re, asphericity and anterior/ posterior
ratio (Table 16-1), m anu al or autom ated keratometry and videokeratoscopy w ill perform
sim ilarly in term s of accu racy and precision for IOL power calcu lation.15
As mentioned previou sly, 95% agreement lim its show a w ide range (>1 D in some stud-
ies) so d ifferent tech nologies can’t be considered exchangeable: Bias mu st be calcu lated
and corrected in the calcu lations (ie, th rough A con stant adju stment.)
If the cornea has an abnorm al cu rvatu re or asphericity, or an increased level of irregu-
larity, corneal topography allow s a more accu rate evalu ation of cu rvatu re than keratom-
etry becom ing probably more adequ ate for IOL power calcu lation. However, repeatability
can be affected d ifferently in each case depend ing on the typ e of irregu lar astigm atism.
Thu s, m easu ring repeated times w ith each method and averaging the K value is a rea-
sonable approach. Th is is the case in corneal scarring, tear film irregu larity, keratoconu s,
keratoplasty, etc.
If the anterior/ posterior ratio is altered (as happ ens after corneal refractive su rgery,
where th is ratio becomes a fu nction of the induced change in anterior corneal shap e)
u sing a fixed corneal index of refraction to estim ate the contribution of the u n know n pos-
terior su rface is incorrect by defin ition. It lead s to an overestim ation of the Sim K: 14% to
 Corn eal Pow er: Corn eal Topog raph y for IOL Pow er Calculation 113

Figure 16-7. Keratometric power maps calculated from anterior and posterior radii. Left: Total
Corneal Power by Galilei. Center: Mean total power by O rbscan. Right: True Net Power by
Pentacam.

25% of the corrected d iopters.16,17 Many m ethod s have been described to correct th is fact
and w ill be explained in a later chapter. A very simple way is to subtract (0.15*corrected
d iopters) from the topograph ic Sim K.
The anterior/ posterior ratio problem is overcome if both rad ii are accu rately measu red
w ith corneal tomographers.
1. Srivan naboon et al18 fou nd a good correlation (r2= 0.853) between su rgery induced
refractive change and Orbscan Total Optical Power for 4 m m w ith an average d if-
ference of +0.17 ±1.16 D.
2. Sónego-Krone et al19 also reported a slight d ifference between refractive change and
Orbscan Total Optical Power for 4 m m (+0.08 ± 0.53 D) and Mean Total Power for 2
m m (+0.07 ± 0.62 D).
3. Qazi et al20 fou nd the best correlation between the ideal Sim K and the Orbscan
Total Axial Power for 4 m m, Total Optical Power for 4 m m, and Mean Total Power
for 3 m m w ith -0.16 ± 0.64 D, +0.04 ± 0.72 D and -0.24 ± 0.68 D average d ifferences
respectively.
Pentacam calcu lates central corneal paraxial power from anterior and p osterior rad ii
in the True Net Power m ap. In order to u se th is value w ith the cu rrent IOL theoretical
formu las (designed to input the Sim K, calcu lated w ith n =1.3375) th is value is converted
to the Equ ivalent K Read ing in the Hollad ay Rep ort modu le of the Pentacam software.
Galilei calcu lates sim ilarly in the so-called Total Corneal Power m ap (Fig. 16-7).
New IOL calcu lation software w ill be m arketed in the near futu re where real calcu la-
tions, both paraxial and exact (p erformed w ith actu al anterior and posterior corneal d ata)
w ill avoid the need for the assu mptions and conversions described in th is chapter—m ak-
ing everyth ing simpler (and hopefu lly more accu rate).

Re fe re nce s
1. Cohen KL, Trip oli N K, Holm gren DE, et al. Assessm ent of the p ower and height of rad ial
aspheres reported by a computer-assisted keratoscope. A m J Ophthalmol. 1995;119(6):723–
732.
2. Mattioli R, Carones F, Cantera E. New algorith m s to improve the reconstruction
of corneal geom etry on the Keratron videokeratographer. Invest Ophthalmol Vis Sci.
1995;36(sup pl):1400.
3. Mattioli R, Carones F. How accu rately can corneal profiles heights be m easu red by placi-
d o-based videokeratography? Invest Ophthalmol Vis Sci. 1996;37(suppl):4273.
114 Ch apter 16

4. Norrby S. Using the haptic plane concept and th ick-len s ray tracing to calcu late intraocu -
lar len s p ower. J Cataract Refract Surg. 2004;30(5):1000–1005.
5. Preu ssner PR, Wah l J, Lado H. Ray tracing for intraocu lar len s calcu lation. J Cataract
Refract Surg. 2002;28(8):1412–1419.
6. Douthwaite WA, Bu rek H. The Bau sch and Lom b keratom eter does not m easu re the tan-
gential rad iu s of cu rvatu re. Ophthalmic Physiol Opt. 1995;15(3):187–193.
7. Hollad ay JT. Corneal top ography u sing the Hollad ay Diagnostic Su m m ary. J Cataract
Refract Surg. 1997;23(2):209–221.
8. Maed a N, Klyce SD, Smolek MK. Comparison of m ethod s for detecting keratoconu s u sing
vid eokeratography. A rch Ophthalmol. 1995;113(7):870–874.
9. H an nu sh SB, Craw ford SL, Waring GO, et al. Accu racy and precision of keratom -
etry, photokeratoscopy, and corneal mod eling on calibrated steel balls. A rch Opthalmol.
1989;107(8):1235–1239.
10. Douthwaite WA. EyeSys corneal topography m easu rem ents applied to calibrated ellipsoi-
d al su rfaces. Br J Ophthalmol. 1995;79(9):797–801.
11. Lam AKC, Chan R, Ch iu R. Effect of p ostu re and artificial tears on corneal p ower m ea-
su rem ents w ith a hand held autom ated keratom eter. J Cataract Refract Surg. 2004;30(3):645–
652.
12. Potvin R, Fon n D, Sorbara L. In vivo comparison of corneal top ography and keratom etry
system s. Contact Lens A nterior Eye. 1996;23(1):20–25.
13. Giraldez MJ, Yebra-Pim entel E, Parafita A, et al. Comparison of keratom etric values of
healthy eyes m easu red by Javal keratom eter, Nid ek autokeratom eter and corneal analysis
system (EyeSys). ICLC. 2000;27:33-38.
14. Dave T, Ru ston D, Fowler C. Evalu ation of the EyeSys model II com puterized videokera-
toscop e. Part I: Clin ical assessm ent. Optom Vis Sci. 1998;75(9):647–655.
15. Cu aycong MJ, Gay CA, Em ery J, H aft EA, Koch DD. Comparison of the accu racy of com -
puterized videokeratoscopy and keratom etry for u se in intraocu lar len s calcu lation s.
J Cataract Refract Surg. 1993;19(suppl):178–181.
16. Hollad ay JT. Cataract su rgery in patients w ith previou s keratorefractive su rgery (RK, PRK
and LASIK). Ophthalmic Practice. 1997;15:238–244.
17. Seitz B, Langenbucher A, Nguyen N X, Ku s MM, Ku ch le M. Und erestim ation of intraocu lar
len s p ower for cataract su rgery after myopic photorefractive keratectomy. Ophthalmology.
1999;106:693–702.
18. Srivan naboon S, Rein stein DZ, Sutton HCS, Holland SP. Accu racy of Orbscan total opti-
cal power m aps in detecting refractive change after myopic laser in situ keratom ileu sis.
J Cataract Refract Surg. 1999;25(12):1596–1599.
19. Sonego-Krone S, Lop ez-Moreno G, Beaujon-Balbi OV, et al. A d irect m ethod to m easu re
the power of the central cornea after myopic laser in situ keratom ileu sis. A rch Ophthalmol.
2004;122:159–166.
20. Qazi MA, Cu a IY, Roberts CJ, Pepose JS. Determ in ing corneal power u sing Orbscan II
vid eokeratography for intraocu lar lens calcu lation after excim er laser su rgery for myopia.
J Cataract Refract Surg. 2007;33(1):21–30.
 17
Co rn e a l Po w e r:
Me a s u rin g Co rn e a l Po w e r
Wit h t h e Pe n t a ca m
Giacomo Savini, M D

T
he Pentacam (Ocu lu s Inc, Wetzlar, Germ any) is a rotating Scheimpflug cam era that
has been developed to im age the anterior segment of the eye and has increased
the range of available tech nologies for the estim ation of corneal power. In contrast
to videokeratography (VKG), where cu rvatu re d ata are derived from the measu red d is-
tances between the rings projected onto the cornea, the Scheimpflug camera m easu res
the corneal rad iu s on the basis of acqu ired im ages of the cornea, via triangle calcu lation.
Th is tech nology can provide the follow ing corneal m easu rements: simu lated keratometry
(Sim K), True Net Power, Equ ivalent K Read ing (EKR), and Total Refractive Power; in
add ition it has been u sed to develop the BESSt formu la.

Sim ulate d Ke ra to m e try

Th is is calcu lated by entering the anterior corneal cu rvatu re rad iu s (in m eters (m)) into
the th in lens formu la for paraxial im agery, wh ich con siders the cornea a single refractive
sphere and read s as:
corneal power = (n – 1)/ r
where r = corneal rad iu s, n = stand ard keratom etric index of refraction (the assu m ed
value of the refractive index of the cornea and the aqueou s hu mor com bined, wh ich is
1.3375 in the case of Pentacam), and 1.000 is the refractive index of air.

115
Hoffer KJ. IOL Power (pp. 115 -118).
© 2011 SLACK Incorporated.
116 Ch apter 17

Although the m ean Sim K measu red by the Pentacam does not show sign ificant d iffer-
ences in comparison to stand ard VKG, the m easu rements of the two tech nologies can not
be u sed interchangeably as the agreement between the Pentacam and VKG is good, but
not perfect: the 95% lim its of agreement are approxim ately ±1.00 D. In other word s, a d if-
ference of up to 1 D can be expected in 95% of eyes; such a range can cau se relevant d if-
ferences in the pred iction of IOL power. Fu rther study is requ ired in order to determ ine
which of the previou sly-mentioned instru ments can assu re the h ighest pred ictability in
IOL p ower calcu lation.

True Ne t Po w e r
This is the resu lt of the Gau ssian optics formu la (GOF) for thick lenses, which read s as:
corneal power = (n 1-n 0)/ r 1+(n 2-n 1)/ r 2 – [(d/ n 1) x (n 1-n 0)/ r 1 x (n 2-n 1)/ r 2]
where n 0 = refractive index of air (= 1.000), n 1 = refractive index of the cornea (= 1.376),
n 2 = refractive index of the aqueou s hu mor (= 1.336), r 1 = rad iu s of cu rvatu re of the ante-
rior corneal su rface (in m eters), r 2 = rad iu s of cu rvatu re of the posterior corneal su rface
(in m eters) and d = corneal th ickness (in m eters).
In virgin corneas, the corneal power measu red by th is m ethod is lower than Sim K
(about 1.2 D) and can not be entered into 3rd generation formu las for IOL power calcu la-
tion. Accord ingly, d ifferent authors have already suggested correcting and re-optim izing
the cu rrent formu las so that the corneal power calcu lated by GOF u sing the Pentacam
d ata can be entered.1

Equivale nt K Re ading
Th is value, wh ich is available in the Hollad ay report of the Pentacam (Fig. 17-1), is an
adju stment of the True Net Power and aim s to provide ophthalmologists w ith a corneal
p ower measu rement that can be u sed for IOL power calcu lation in eyes that have u nder-
gone excimer laser corneal refractive su rgery. Measu rements at the 1, 2, 3, 4, 4.5, 5, 6, and 7
m m zones are generated by the internal software. Even if the m anu factu rer suggests that
the 4.5 m m measu rem ent shou ld be the preferred value, it has been rep orted 2,3 that the 2
and 3 m m measu rem ents show the closest agreement w ith respect to the Clin ical H istory
Method. However, caution is warranted, as the 95% lim its of agreement are still w ide.

To tal Re fra c tive Po w e r

The latest software version of the Pentacam (Version 1.17) provides an add itional mea-
su rem ent, called Total Refractive Power. Corneal power is calcu lated accord ing to Snell’s
law and u sing ray-tracing: parallel light is sent th rough the cornea and each light beam
is refracted accord ing to the d ifferent refractive index of air, cornea, and aqueou s hu mor;
the slope of the su rfaces; and the exact location of the refraction. Th is measu rem ent is
prom ising, as it wou ld not su ffer from the keratom etric index problem typical of eyes that
have had excimer laser su rgery, and awaits valid ation from clin ical stud ies.
 Corn eal Pow er: Measurin g Corn eal Pow er With th e Pen tacam 117

Figure 17-1. The Holladay report of the Pentacam of a patient that had undergone myopic PRK. The
Equivalent K Reading at 4.5 mm is shown at the center top.

BES St Fo rm ula
In order to ach ieve the best agreem ent w ith the Clin ical H istory Method, Borasio 4
and coauthors mod ified the GOF and develop ed the BESSt formu la, wh ich is based
on a variable rather than a static refractive index of the cornea. Th is formu la has been
implemented in a Microsoft Window s (Microsoft, Red mond, WA) software program and
is also available in the freely dow n load able Hoffer/ Savin i Tool from w w w.EyeLab.com.
In prelim inary stud ies, the values thu s obtained lead to a good accu racy in IOL power
calcu lation after myopic and hyperopic LASIK. Fu rther stud ies, however, are necessary
to con firm these resu lts.

Re fe re nce s
1. Norrby S. Pentacam keratom etry and IOL p ower calcu lation. J Cataract Refract Surg.
2008;34(1):3.
2. Savin i G, Barbon i P, Profazio V, Zan in i M, Hoffer KJ. Corneal p ower m easu rem ents w ith
the Pentacam Scheimpflug cam era after myopic excim er laser su rgery. J Cataract Refract
Surg. 2008;34(5):809–813.
3. Savini G, Barboni P, Carbonelli M, Hoffer KJ. Accu racy of Scheimpflug corneal power mea-
su rements for intraocu lar lens power calcu lation. J Cataract Refract Surg. 2009;35(7):1193–1197.
4. Borasio E, Steven s J, Sith GT. Estim ation of tru e corneal p ower after keratorefrac-
tive su rgery in eyes requ iring cataract su rgery: BESSt formu la. J Cataract Refract Surg.
2006;32(12):2004–2014.
 18

IOL Po sitio n :
ACD a n d ELP
Kenneth J. Hoffer, M D

Axial Po s itio n o f the IOL

T
he axial position for the IOL as it relates to the anterior su rface of the cornea has
h istorically been referred to as the a nt er ior cha mber dept h (or ACD) becau se the
optic of all IOLs in the early era was positioned in front of the iris, in the anterior
cham ber, or in the iris plane. Becau se almost all IOLs tod ay are p ositioned beh ind the
iris, new term inology has been offered such as effect ive lens posit ion (ELP) by Hollad ay 1
and actu al len s position (ALP) by the FDA. ELP is also u sed when referring to anterior
cham ber (AC) len ses, wh ich do sit in the ACD.
Tod ay, ACD is defined as the axial d istance from the central front su rface (anterior
vertex) of the cornea to the central front su rface of the crystalline len s. ELP is defined as
the axial d istance between the front su rfaces of the two len ses (cornea and IOL); or, more
exactly, the d istance from the central front su rface (anterior vertex) of the cornea to the
effective principle plane of the IOL. For biconvex IOLs, the principle plane is located about
in the m idd le of the IOL (50% of the th ickness).
The ELP is requ ired for all formu las, but is u sed in d ifferent form s. In the Bin khorst,
Colenbrander, and Hoffer formu las, it is u sed d irectly and called the ACD. In the Hoffer
Q formu la, it is referred to as the pACD (p ersonal ACD) and in the Hollad ay 1 formu la it
is calcu lated u sing a su rgeon factor (SF) specific to each IOL style. In the SRK I and II and
the SRK/ T, it is incorporated into the A con stant specific to each IOL style.

119
Hoffer KJ. IOL Power (pp. 119 -122).
© 2011 SLACK Incorporated.
120 Ch apter 18

The formu las for converting these values are:

ACD = ((0.5663 * A) – 62.005)/ 0.9704
ACD = (SF + 3.595)/ 0.9704
A = (SF +65.6)/ 0.5663
A = ((ACD * 0.9704) + 62.005)0.5663
SF = 0.9704 * ACD – 3.595
SF = 0.5663 * a – 65.6
where SF = su rgeon factor, A = SRK A con stant, and ACD = pACD.
These formu las resu lt in rough approxim ations to begin w ith before developing ind i-
vidu al p ersonalized constants.
As an example, ACDs of 2.5, 4.0, and 5.5 m m wou ld be converted to A constants of
113.78, 116.35, and 118.92, respectively. From these A constants you can see that the d iffer-
ence between an AC len s sitting at 2.5 m m (113.78) and a PC lens sitting at 5.5 m m (118.92)
is about 5 D. Thu s you can see that the ELP is contained w ith in the A con stant.
Some have prop osed that it wou ld be u sefu l to measu re the preop erative anatom ic
ACD (corneal epitheliu m to anterior capsu le) either w ith an A-scan u n it or by optical
pachymetry. I performed such a comparison study 2 on 44 eyes and showed that the opti-
cal pachymetry method resu lted in a m ean 0.20 (± 0.35) m m deep er ACD than obtained
by u ltrasou nd u sing 1548 m/ sec (3.14 vs. 2.93 m m) and recom mended the optical pachym-
etry for accu rate measu rement of ACD.
The IOL position has been con sidered the least important of the th ree variables as a
cau se of IOL p ower error, but in 1998 I exam ined a one-d ay postop erative (PO) IOL patient
w ith a shallowed ACD and myopia of -2.50 D. After 3 d ays, the cham ber deep ened by 2.0
m m and the refractive error changed to plano. So it is fou nd to be clin ically important.
IOL position has received the most attention from formu la w riters over the past 15 years.
The m ajor effort has been toward better pred iction of where the IOL w ill u ltim ately rest.
In the 1970s and early 1980s, 3.5 m m was u sed as an ACD for all IOLs and all eyes. In
1982, I rep orted a study 3 of the postoperative position of PMMA posterior cham ber (PC)
lenses and noted that the PO ACD (or ELP) increased as the AL increased. From th is
d ata a formu la was proposed to calcu late the pred icted ACD to u se in the IOL power
formu la:
ACD = 2.92 * AL – 2.93
where AL = axial length.
Of cou rse th is regression formu la was specific for th is one lens style and wou ld requ ire
adju stment for each IOL style. In 1988, Hollad ay referred to th is AL-dep end ant method
as the 2nd generation of formu las when he proposed the Hollad ay 1 formu la as the 3rd
generation, wh ich introduced ACD dep endence on the corneal cu rvatu re as well as the
AL. He also introduced h is lens-specific SF. He u sed a Fyodorov corneal height formu la
to calcu late the d istance from the front su rface of the cornea to the iris plane and to that
added the IOL-sp ecific SF. The SF wou ld be d eterm ined for each IOL style by back cal-
cu lation know ing the PO resu lts from a series of cases. Th is same process was u sed by
Retzlaff when w riting the SRK/ T formu la.
A study4 I p erformed on a series of 270 eyes receiving a silicone plate haptic lens
showed that the IOL sh ifted a m ean of 0.06 m m posteriorly (ACD deep ened) at 3 months
PO, compared to its position on the first d ay after su rgery. Th is correlated w ith a m ean
0.21 D sh ift toward hyperopia seen in these patients.
 IOL Position : ACD an d ELP 121

The Olsen formu la u ses add itional parameters, such as the preoperative ACD, corneal
d iameter, and len s th ickness to pred ict the ELP. Hollad ay followed sim ilar lines when
devising the Hollad ay 2 formu la. The H aigis formu la d ropped u sing the K and replaced
it w ith the preoperative ACD.

N OTE: A n IOL intended for capsular bag placement should be decreased by 0.75 to 1.25 D
(depending upon the IOL power) when placed instead in the ciliary sulcus. Of course it would be
wise to calculate a power for sulcus placement in advance by decreasing the ELP for that lens.

Re fe re nce s
1. Hollad ay JT, Prager TC, Chand ler TY, Mu sgrove KH, Lew is JW, Ru iz RS. A th ree-part sys-
tem for refin ing intraocu lar len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–
24.
2. Postop erative Measu rem ent of Anterior Cham ber Depth s. Second U.S. Intraocu lar Len s
Symp osiu m, Am erican Intra-Ocu lar Im plant Society (AIOIS). Los Angeles, CA, April
1979.
3. Hoffer KJ. Biom etry of the p osterior cap su le: A new formu la for anterior cham ber d epth
of posterior cham ber len ses. In: Em ery JC, Jacobsen AC, ed s. Current Concepts in Cataract
Surgery: Selected Proceedings of the Eighth Biennial Cataract Surgical Congress. Norwalk, CT:
Appleton-Centu ry Crofts; 1983;56–62.
4. PO Pachym etry ACD of 230 Staar AA4203 Silicone Plate IOL’s. AAO Cou rse #256, Modern
Implant Su rgery: XVI (Hoffer). Am erican Academy of Ophthalmology; October 31, 1995;
Atlanta, GA.
 19
IOL Po sitio n :
Me a su rin g t h e ACD by
Optica l Pa ch ym e try
Kenneth J. Hoffer, M D

D
istances in the anterior cham ber can be measu red in a variety of ways. The earliest
method was by optical pachymetry u sing a device developed by Gold m an (H aag-
Streit USA) for the slit lamp. I am most fam iliar w ith the device available for the
H aag-Streit u nit (Fig. 19-1). There is also one available for the Zeiss slit lamp (Fig. 19-2). I
was first influenced to pu rchase this instru ment after read ing the stud ies by Corneliu s
Binkhorst,1 who u sed it to measu re the position of prepupillary IOLs in the supine and
prone positions in the early 1970s.1 It was qu ite a feat perform ing the exam ination in these
two positions and his technician, Leo Loones, had qu ite a time. After receiving the u nit, I
taught myself how to u se it and began u sing it on all my cataract patients.
Tod ay, most ophthalmologists are u n fam iliar w ith these method s since the introduc-
tion of and popu larity of u ltrasou nd pachym eters in the 1980s. The latter instru ments
tend to measu re the depth of the anterior cham ber 0.20 to 1.0 m m shorter than by opti-
cal pachymetry. I have long questioned the accu racy of measu ring the anterior cham ber
depth (ACD) preoperatively or p ostop eratively u sing A-scan u ltrasou nd, especially by
contact applanation but even w ith im mersion. Koranyi et al2 compared contact appla-
nation A-scan (u sing 1532 m/ sec average sou nd velocity) w ith Scheimpflug im aging, a
H aag-Streit optical pachymetry, and Orbscan analysis. Their resu lts (Table 19-1) show the
26% to 28% increase in AC depth obtained by the optical mod alities versu s u ltrasou nd.
These authors opined that the 1.0 m m shorter ACD m ay have been due to the indenta-
tion of the cornea in the contact method, but indentation of that m agn itude by experi-
enced tech n ician s seem s h igh ly u n likely. Since the axial ACD is m ade up of the aqueou s
(at 1532 m/ sec) and the cornea (at 1641 m/ sec), the authors shou ld have u sed an average
123
Hoffer KJ. IOL Power (pp. 123-130).
© 2011 SLACK Incorporated.
124 Ch apter 19

A B C

Figure 19-1. O ptical pachymeters for measuring ACD

using a slit lamp (A) Haag-Streit O ptical Pachymeter II
device (B) in its storage box with the split ocular, (C)
viewing end of split ocular, (D) slit image as seen
through split ocular.

Figure 19-2. (A) Zeiss O ptical

Pachymeter in its storage A
box with the split ocular (B)
attached to the top of the
Zeiss slit lamp.

B
 IOL Position : Measurin g th e ACD by Optical Pach ymetry 125

Ta b le 1 9 -1 .

INCREASE IN AC D EPTH O BTAINED BY O PTICAL MO DALITIES VS ULTRASO UND

Instrument Pre-op ACD PO ACD Increase %
Over US
A scan US 3.05 3.69
Scheimpflug 3.37 4.65 0.96 26%
H-S O ptical 4.69 1.00 27%
O rbscan 4.71 1.02 28%

where US = ultrasound, H-S = Haag-Streit, Pre-op = preoperative, PO = postoperative

velocity of 1534 m/ sec. However, if th is d iscrepancy was due to u sing the w rong u ltra-
sou nd velocity, the correct velocity wou ld have had to be 1947 m/ sec, wh ich is not pos-
sible. I believe there is someth ing in herent to the u ltrasou nd m easu rement of the ACD
that prevents getting an accu rate read ing by im mersion or applanation. It app ears that
it wou ld be more prudent to u se other method s than u ltrasou nd to measu re the ACD,
esp ecially in scientific stud ies for publication.
The Zeiss IOLMaster also incorporates an optical measu rement of ACD but I have
fou nd it to be qu ite variable compared to carefu l m anu al optical pachym etry u sing the
H aag-Streit. It appears the optical pachymeter is presently the most accu rate method to
measu re these AC d istances.
Lastly, Suto and associates 3 p erformed a retrospective study of 30 norm al eyes (22.0 to
24.5 m m) in wh ich the sam e IOL (Alcon MA60BM) was implanted in the ciliary su lcu s
of one eye rather than the capsu lar bag due to capsu lar ruptu re. The physician s noted an
average overcorrection in refractive error of -0.78 D at 3 months and a +1.11 D overcor-
rection in IOL power over and above the correspond ing in-the-bag IOL. The d ifference
between the mean measu red ACD in these eyes (3.47 ± 0.25 m m) and that of the contra-
lateral eye (4.21 ± 0.29) receiving the same len s in the bag was a statistically sign ificant
0.74 m m (p <.0001). Th is translates into a 1.50 D/ m m change in IOL position in norm al
AL range eyes. In order to lower the pred iction error, the authors prospectively deducted
1.00 D from the plan ned IOL p ower when the len s was placed in the su lcu s in 16 eyes. The
pred iction error then fell to -0.06 D ± 0.19.
Th is chapter is intended to be a short cou rse on how to take measu rements w ith the
H aag-Streit Optical Pachym eter II. The steps m ay seem complex and cu m bersom e but
once one gets u sed to it, it is very easy to u se.
126 Ch apter 19

Figure 19-3. Placing the pachymeter onto the

Goldman tonometer post. Figure 19-4. Slit lamp microscope housing turned
45° to the right of the slit beam housing.

A B

Figure 19-5. Replacement of the regular microscope right ocular (A) with the special split
beam ocular (B).

Optical Pa chym e try Ex a m ina tio n

First remove the pachymeter from its storage box and slip it onto the Gold m an tonom-
eter post at the top of the slit lamp, m aking su re it is firm ly seated (Fig. 19-3).
The slit beam hou sing shou ld be aimed d irectly straight ahead at the patient’s eye
while you rotate the m icroscope hou sing 45° to the right of the slit beam hou sing. The
exam iner mu st now move the stool so that they can sit to the right side and com fortably
take the read ing (Fig. 19-4).
The regu lar m icroscope ocu lar for the right eye mu st now be removed (Fig. 19-5) and
replaced w ith the sp ecial split beam ocu lar from the storage box (see Fig. 19-1).
Several adju stm ents have to be m ade. First, rotate the right ocu lar’s m agn ification
cou nterclockw ise to +6 on the scale (Fig. 19-6).
Then you mu st rotate the split ocu lar so that the line in the open ing is perfectly hori-
zontal (Fig. 19-7).
 IOL Position : Measurin g th e ACD by Optical Pach ymetry 127

Figure 19-6. Rotation of the right ocular’s magni- Figure 19-7. Rotation of the split ocular to hori-
fication to +6. zontal.

Figure 19-9. Magnification turned to maximum.

Figure 19-8. Voltage switch; maximum is to the

right.

Now the voltage sw itch of the slit beam, wh ich is located at the bottom left of the slit
lamp table, is tu rned to the right to its m aximu m level of illu m ination (Fig. 19-8).
Finally you mu st flip the objective m agn ification hand le (below the binocu lars) to
m aximu m m agn ification (Fig. 19-9).
The patient’s head can now be placed in the ch in rest (Fig. 19-10) and the slit beam
brought close to the patient and centered in the pupil by adju sting the slit u sing the other
hand on the joystick.
Now the room door shou ld be closed and all lights tu rned out so it is pitch-black.
While looking th rough the right (split) ocu lar, u se you r right hand to rotate the gauge at
the top of the instru ment (Fig. 19-11).
You w ill see a single im age (at read ing zero) split into two (Fig. 19-12A) w ith one im age
slid ing past the other as you rotate the gauge to the left (Fig. 19-12B).
Now rotate the gauge at the top of the pachymeter u ntil the vertical slit im age of the cor-
neal front su rface is perfectly aligned w ith the vertical slit im age of the anterior lenticu lar
su rface (see Fig. 19-12B)—or whatever you desire to measu re—m aking su re that the im ages
are positioned in the center of the pupil. This latter requ irement does take some constant
128 Ch apter 19

Figure 19-11. Taking the measurement.

Figure 19-10. Patient placed in the slit lamp.

B
A

Figure 19-12. (A) Diagram of image of the anterior chamber through the split ocular; left at zero, right
at ACD measurement. (B) Sliding the gauge.

adju sting of the slit lamp joystick to keep the im age dead center. It helps to u se a penlight
and a hand m agnifier to read the gauge scale and guesstim ate to tenths of a m illimeter.
The cornea m agn ifies or m in im izes the im age you are m easu ring, so the read ing
obtained mu st be corrected u sing the chart supplied by the m anu factu rer (Fig. 19-13). The
chart u ses rad iu s of cu rvatu re rather than d iopters, and the K read ings and scale read ings
are so far apart that it requ ires a lot of mental interp olation on both scales.
To m ake th is step easier I created a chart u sing a spread sheet that first converted the K
read ings from rad iu s of cu rvatu re to d iopters and then expanded the ranges of both the
X and Y axis to m ake the interpolation much less intense. The expansion was such that
it requ ired two charts wh ich can be printed and lam inated back to back (Fig. 19-14). The
final read ing is obtained by add ing (the black nu m bers) or su btracting (the red nu m bers)
the correction to the original pachymetry read ing obtained. These charts can be obtained
by requesting them by em ail ([email protected]).
Once the process is learned, it becomes a very simple procedu re to perform.
 IOL Position : Measurin g th e ACD by Optical Pach ymetry 129

Figure 19-13. The Haag-

Streit correction chart.

Re fe re nce s
1. Nord loh ne ME. The effect of supine and prone position of the patient on the position of the
Binkhorst lens in the eye after intracapsular and extracapsular operation respectively. The H ague,
Netherland s: Dr. W. Ju n k B.V. Pu blishers; 1975:231-237.
2. Koranyi G, Lyd ah l E, Norrby S, Tau be M. Anterior cham ber depth m easu rem ent: A-scan
versu s optical m ethod s. J Cataract Refract Surg. 2002;28(2):243–247.
3. Suto C, Hori S, Fu kuyam a E, Aku ra J. Ad ju sting intraocu lar len s p ower for su lcu s fixation.
J Cataract Refract Surg. 2003;29(10):1913–1917.
130 Ch apter 19

Figure 19-14. Correction charts to make the process easier.

 20
IOL Po sitio n :
Do u b le -AL Me th o d fo r
Scle ra l Bu ckle Eye s
Kenneth J. Hoffer, M D

M
odern 3rd generation formu las (except the H aigis) u se both the AL and the K
read ing to pred ict the position the IOL w ill sit in the eye. When the central
cornea is flattened by refractive laser su rgery, the relationsh ips in the anterior
cham ber do not change. However, now the formu las w ill be u sing a very flat K read ing
to pred ict the ELP. The Aram berri Double-K1 method solves th is problem by u sing two
K read ings; the preop erative K read ing (or 43.5 D if u n know n) to pred ict the ELP, and
the p ostoperative flat K read ing to calcu late the IOL power. The Hoffer Program s® IOL
p ower software allow s the Double-K to be calcu lated for the Hoffer Q, the Hollad ay 1 and
the SRK/ T formu las and the Hollad ay IOL Con su ltant® allow s it on ly for the Hollad ay
2 formu la.
In the same vein, most post-encircling band retinal detach m ent (RD) eyes have an
approxim ate 1.0 m m increase in AL, but the ACD is not affected by the encircling band
(Fig. 20-1). Therefore, it wou ld be best to u se the m easu red AL m inu s 1 m m in the part
of the formu la that calcu lates the pred icted ELP and u se the measu red AL in the part of
the formu la that calcu lates the IOL power; ie, Dou ble-AL. Th is basically amou nts to m ak-
ing the IOL p ower calcu lated a little weaker than wou ld be pred icted u sing the modern
formu las. Since no IOL p ower program s autom atically allow you to enter two ALs, alter-
natively one cou ld ju st lower the power of the recom mended IOL power in such RD eyes.
I proposed th is method in 2000.2

131
Hoffer KJ. IOL Power (pp. 131-132).
© 2011 SLACK Incorporated.
132 Ch apter 20

Figure 20-1. Rational for using a Double-AL in

eyes with a scleral buckle.

Re fe re nce s
1. Aram berri J. Intraocu lar len s p ower calcu lation after corneal refractive su rgery: Dou ble-K
m ethod . J Cataract Refract Surg. 2003;29(11):2063–2068.
2. Mod ern izing IOL Power. IOL Power Calcu lation : Striving for Accu racy. ASCRS Cou rse
#2102, Symp osiu m on Cataract, IOL and Refractive Su rgery. Am erican Society of Cataract
and Refractive Su rgery; May 21, 2000; Boston, MA.
 Fo rmu las
a n d Sp e cia l
Circu msta n ce s
II
 21

Se ctio n II:
In tro d u ct io n
Kenneth J. Hoffer, M D

I
n the second half of th is book we w ill be covering all the subjects dealing w ith IOL
power calcu lation. It w ill be u sefu l to you to have covered the first half of the book
dealing w ith biometry, but it is not necessary. Here we w ill cover h istorical back-
grou nd on the formu las we u se and the problem s cau sed by the p opu larity of regression
formu las. The newest theoretic formu las, includ ing the Olsen and H aigis formu las w ill be
covered ; along w ith IOL power program s, formu la personalization, and formu la u sage.
We w ill then cover most of the d ifficu lt problem s that can face an ophthalmologist in
trying to obtain an accu rate IOL power, such as eyes w ith staphylom a (Fig. 21-1), eyes
filled w ith silicone oil, u n ilateral h igh myop es and hyp erop es, eyes also need ing a p en-
etrating keratoplasty, or eyes that have a scarred cornea.
Important tod ay are the problem s associated w ith eyes that have had refractive corneal
su rgery such as rad ial keratotomy (RK) as well as laser refractive su rgery (PRK, LASIK
and LASEK). We w ill cover the Dou ble-K Method, problem s w ith spherical aberration
and how to calcu late power for mu ltifocal and toric IOLs as well as IOLs for p ed iatric
eyes, piggyback IOL powers, an isometropia, and an iseikon ia.
Finally we w ill give an overview as to how to prevent power errors and how to treat
them when they occu r. After going th rough all th is m aterial, it w ill become obviou s that
attention to detail can go a long way to provid ing the desired p ostoperative refractive
error for a h igh percentage of you r patients.

135
Hoffer KJ. IOL Power (pp. 135 -136).
© 2011 SLACK Incorporated.
136 Ch apter 21

Figure 21-1. B-scan of an eye with staphyloma

showing the bulging area above the optic nerve
shadow.
 22
Fo rmu la s a n d Pro g ra ms:
Fo rmu la Histo ry
a n d Basics
Kenneth J. Hoffer, M D

Fo rm ula His to ry

I
t is important to u nderstand the h istory of IOL power formu las so that one gain s an
u nderstand ing of the vagaries of tod ay’s modern formu las (Fig. 22-1).

1 s t Ge ne ratio n
The first IOL power formu la was a theoretic one published by Fyodorov and Kolon ko1
in 1967. It was based on Gau ssian optics and the Gu llstrand eye. Gernet d id extensive
work in th is regard in the late 1960s and 1970s also studying the effects of aneisikon ia.
Colenbrand er 2 w rote h is formu la in 1972, followed by the Hoffer 3 formu la in 1974 (pub-
lished in 1981). Van der Heijde4 pu blished h is formu la and nomogram in 1975, the sam e
year Bin khorst 5 published h is formu la. The latter became w idely u sed in America in
conju nction w ith the popu larity of the Sonometrics 1st A-scan for IOL p ower. Becau se of
d issatisfaction w ith the resu lts of the Bin khorst formu la, in 1978, first Lloyd and Gills,6,7
followed by Retzlaff7 and later Sanders and Kraff,9 each develop ed a regression formu la
based on analysis of their previou s IOL cases. Later Sanders, Retzlaff, and Kraff collabo-
rated to produce the SRK I regression formu la 10 in 1980. Since regression formu las were
simple to calcu late (P = A – 2.5 * AL – 0.9 * K) they were rapid ly adopted as the stand ard
th roughout the world. The essence of th is formu la generation, theoretic or regression,

137
Hoffer KJ. IOL Power (pp. 137-142).
© 2011 SLACK Incorporated.
138 Ch apter 22

A B

Figure 22-1. Formula history.

was that they all u sed a single con stant for each len s that represented the pred icted IOL
position (ACD/ ELP).

2 nd Ge ne ratio n
The 2nd generation was issued by me11,12 at the Welsh Cataract Congress in Hou ston in
1982. The resu lts of a study of a large series of eyes showed a d irect relation sh ip between
the AL of the eye and the position of the PMMA posterior cham ber IOL. A simple regres-
sion formu la was presented to better pred ict ACD:
ACD = 2.92 * AL – 2.93
Others followed by developing d ifferent m echan ism s to apply th is AL-related pre-
d ictive relation sh ip (Bin khorst,13 SRK II14), wh ich Hollad ay later defined as the second
generation.

3 rd Ge ne ratio n
In 1988, Hollad ay 15 proposed a d irect relationsh ip between the steepness of the cornea
and the position of the IOL. He mod ified the Bin khorst formu la to incorp orate th is as well
as the AL relation sh ip. Instead of ACD input, the formu la wou ld calcu late the pred icted
d istance from the cornea to the iris plane (u sing a corneal height formu la by Fyodorov)
and add to it the d istance from the iris plane to the IOL. The latter he called the surgeon
factor (SF) and it is specific to each len s. Since the SF is impossible to determ ine u ntil
after the IOL has been implanted, he calcu lated the SF for each len s style by u sing the
PO refractive error, the IOL power implanted, along w ith the AL and K to back calcu late
th rough h is formu la what the ideal SF shou ld have been. Then he wou ld take the aver-
age of all the SFs in a large series to arrive at the SF to u se for that len s in the futu re. The
m athem atics requ ired a qu ad ratic equ ation that necessitated a m athem atician to solve.
In 1990, Retzlaff16 followed su it and mod ified the Hollad ay 1 formu la to allow u se of
an A con stant instead of a SF, calling it the SRK/ T theoretic formu la. He d id not personal-
ize the A con stant by back-calcu lating th rough the formu la but instead relied on the older
 Formulas an d Prog rams: Formula History an d Basics 139

method of personalizing an A con stant. The SRK/ T was intended to replace the previou s
SRK regression formu las, but over the next decade 50% of American su rgeon s were still
u sing the old regression formu las.
In 1992, at the u rging of Hollad ay (p ersonal com mu n ication), I develop ed the Q for-
mu la 17 u sing a tangent fu nction of the K read ing to accomplish a sim ilar effect. The
Fyodorov corneal height formu la was NOT u sed. The base Hoffer formu la (1974) was not
changed. The Q part of the formu la is ju st a separate calcu lation to pred ict the ELP u sed
in the base formu la. The Q formu la starts w ith a personalized ACD (pACD) and adju sts it
based on the AL and K. The personalized pACD is calcu lated for each len s style by back-
calcu lating th rough the Hoffer Q formu la, and the average of a series of ideal pACDs is
u sed. Th is also requ ired a quad ratic equation and I also needed the aid of a m athem atics
professor to solve it for me.

Figure 22-2. Holladay 2 formula.

4 th Ge ne ratio n
In 1990, Olsen 18 proposed u sing other anterior segment measu rements (such as the
preoperative ACD) to better estim ate the postoperative IOL p osition, and published algo-
rith m s for th is. After several stud ies19 showed the Hollad ay 1 formu la not as accu rate as
the Hoffer Q in eyes shorter than 22 m m, Hollad ay, in fluenced by Olsen’s concept, u sed
the preoperative ACD measu rement as well as the corneal d iameter, the crystalline lens
th ickness, the preop erative refractive error, and age to calcu late an estimated scaling factor
(ESF). The IOL-specific ACD is then mu ltiplied by the ESF to arrive at the ELP u sed in
the formu la. Th is he called the Hollad ay 2 formu la (Fig. 22-2) wh ich he has promu lgated
since 1996, but has yet to be published. Not being published, it is not possible for others
to know if the formu la has been tweaked since 1996.
140 Ch apter 22

Figure 22-3. Wolfgang Haigis, PhD.

5 th Ge ne ratio n
In 1999, Wolfgang H aigis20 (Fig. 22-3) proposed u sing th ree constants to pred ict the
position of the IOL based on the characteristics of the eye and the IOL to be u sed. The
formu la replaces the u se of the K read ing w ith u sing the Olsen concept of u sing the pre-
operative ACD m easu rement. It calcu lates the pred icted PO ELP by:
ELP = a 0 + a 1 * ACD + a 2 * AL
where ELP = pred icted IOL position, a 0 = an IOL specific constant, a 1 = a lens specific
con stant to be effected by the measu red preop erative ACD, a 2 = a lens specific constant
to be effected by the m easu red preop erative axial length, ACD = the m easu red axial d is-
tance from the corneal ap ex to the front su rface of the lens, and AL = axial length.
As in all formu las, the con stants mu st be “optim ized” (p ersonalized) to each IOL style
and su rgeon. Single optim ization on ly optim izes the a 0 and creates accu racy equ al to the
Hoffer Q and Hollad ay 1, but triple optim ization of all th ree constants creates add itional
accu racy. The problem is that triple optim ization requ ires a series of 500 to 1000 eyes of
one len s style and the eyes in the series mu st statistically cover all axial lengths from very
short to very long. Th is m ay be qu ite d ifficu lt to ach ieve for the average su rgeon.

Re fra c tio n Fo rm ula

In 1993, Hollad ay 21 pu blished a formu la to calcu late the power of an IOL to be implant-
ed into an aphakic eye or an ametropic pseudophakic eye (piggyback IOL), or a phakic eye
having a phakic IOL implanted. It does not need the AL but requ ires the corneal power,
preop erative refractive error, and desired postop erative refractive error, as well as the
vertex d istance of both. I do not recom mend its u se in aphakic eyes becau se the vertex
d istance is d ifficu lt to m easu re accu rately and, due to the h igh power of their refractive
error, greater errors can resu lt. It is, however, a good check against the AL formu la cal-
cu lation.
 Formulas an d Prog rams: Formula History an d Basics 141

Sum m ary
Tod ay, it is pretty well accepted that regression formu las such as the SRK I and SRK
II shou ld not be depended upon for patient IOL power calcu lation except in em ergency
situations and on ly in eyes in the norm al AL range. They are esp ecially a problem in eyes
that have had refractive corneal su rgery and in these cases shou ld N EVER be u sed.

Re fe re nce s
1. Fyodorov SN, Kolon ko AI. Estim ation of optical p ower of the intraocu lar len s. Vestnik
Oftalmologic (Moscow). 1967;4:27.
2. Colenbrand er MC. Calcu lation of the p ower of an iris clip lens for d istant vision. Br J
Ophthalmol. 1973;57(10):735–740.
3. Hoffer KJ. Intraocu lar len s calcu lation : The problem of the short eye. Ophthalmic Surg.
1981;12(4):269–272.
4. Van d er Heijd e GL. A nomogram for calcu lating the p ower of the prepupillary len s in the
aphakic eye. Bibl Ophthalmol. 1975;83:273-275.
5. Bin khorst RD. The optical design of intraocu lar lens implants. Ophthalmic Surg. 1975;
6(3):17–31.
6. Gills JP. Regression formu la. J A m Intraocul Implant Soc. 1978;4(4):163–164.
7. Gills JP. Min im izing p ostop erative refractive error. Contact and Intraocular Lens M ed J.
1980;6:56–59.
8. Retzlaff J. A new intraocu lar len s calcu lation formu la. J A m Intraocul Implant Soc.
1980;6(2):148–152.
9. Sand ers DR, Kraff MC. Im provem ent of intraocu lar len s p ower calcu lation u sing empiri-
cal d ata. J A m Intraocul Implant Soc. 1980;6(3):263–267.
10. Sanders D, Retzlaff J, Kraff M et al. Comparison of the accu racy of the Bin khorst,
Colenbrander and SRK im plant p ower pred iction formu las. J A m Intraocul Implant Soc.
1981;7(4):337–340.
11. Hoffer KJ. Biom etry of the p osterior cap su le. In : Em ery JC, Jacobson AC, ed s. Current
Concepts in Cataract Surgery (Eighth Congress). New York, N Y: Appleton-Centu ry Crofts;
1983:56–62.
12. Hoffer KJ. The effect of axial length on posterior cham ber lenses and posterior capsu le
p osition. Current Concepts in Ophthalmic Surg. 1984;1:20–22.
13. Bin khorst RD. Biom etric A-scan u ltrasonography and intraocu lar len s power calcu la-
tion. In : Em ery JE, ed. Current Concepts in Cataract Surgery: Selected Proceedings of the Fifth
Biennial Cataract Surgical Congress. St. Lou is, MO: Mosby CV; 1987:175–182.
14. Sanders DR, Retzlaff J, Kraff MC. Comparison of the SRK II formu la and other second
generation formu las. J Cataract Refract Surg. 1988;14(2):136–141.
15. Hollad ay JT, Prager TC, Chand ler TY, et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–24.
16. Retzlaff J, Sand ers DR, Kraff MC. Developm ent of the SRK/ T intraocu lar len s implant p ower
calcu lation formu la. J Cataract Refract Surg. 1990;16(3):333–340. Erratu m : 1990;16(4):528.
17. Hoffer KJ. The Hoffer Q formu la: A comparison of theoretic and regression formu las.
J Cataract Refract Surg. 1993;19(6):700–712. Errata: 1994;20(6):677 and 2007;33(1):2–3.
18. Olsen T, Oleson H, Th im K, Corydon L. Pred iction of p ostop erative intraocu lar len s
cham ber d epth. J Cataract Refract Surg. 1990;16(5):587–590.
19. Hoffer KJ. Clin ical resu lts u sing the Hollad ay 2 intraocu lar lens power formu la. J Cataract
Refract Surg. 2000;26(8):1233–1237.
142 Ch apter 22

20. H aigis W. The H aigis formu la. In: H J Sham m as, ed. Intaocular Lens Power Calculations.
Thorofare, N J: SLACK Incorp orated ; 2003:41–57.
21. Hollad ay JT. Refractive power calcu lations for intraocu lar len ses in the phakic eye. A m J
Ophthalmol. 1993;116(1):63–66.
 23
Fo rmu la s a n d Pro g ra ms:
Re g re ssio n a n d
Th e o re tic Fo rmu la s
Kenneth J. Hoffer, M D

Accura c y Re po rting

I
n reporting resu lts of IOL power calcu lation or instru ment perform ance it has been
long established that the follow ing d ata shou ld be rep orted :

1. The m ean error (ME) and stand ard deviation (SD) in pred iction.
2. The m ean absolute error (MAE) and stand ard deviation (SD) in pred iction.
3. The p ercentage of eyes ± 0.50 D from pred icted target refraction.
4. The p ercentage of eyes ±1.00 D from pred icted target refraction.
5. The p ercentage of eyes >2.00 D from pred icted target refraction.
6. The range of errors from m aximu m plu s to m aximu m m inu s.
Eg: -0.06 D/ 0.46 D/ 67%/ 90%/ 1%/ +1.76 to -1.01 D
It shou ld be the comparison of the actu al p ostop erative refractive error w ith that pre-
d icted by the p ower calcu lation and NOT the comparison of the IOL power implanted
w ith the IOL power pred icted by the calcu lation to produce the actu al PO refractive error.
The d ifference between these method s of rep orting is a factor of 1.25/ 1. Thu s, comparison
of d ifferent stud ies u sing the two m ethod s can not be m ade easily.
Th is reporting schem a allow s one to see the clin ical effects in one’s practice from
the resu lts. Not on ly is the percentage of eyes w ith in ±1.00 D imp ortant, but so is the

143
Hoffer KJ. IOL Power (pp. 143-14 8).
© 2011 SLACK Incorporated.
144 Ch apter 23

Ta b le 2 3 -1 .

EYES W ITHIN ±0.50 D O F PREDICTIO N D EPENDING O N AL O F THE EYE

(HO FFER STUDY)
Best Results
Size Formula % ± 0.50 D # Eyes
Short Hoffer Q 67% 24
Medium Hoffer Q 67% 219
Medium long Holladay 71% 47
Very long SRK/T 57% 13
=303
303/450 = 67%

m aximu m range of errors as regard s to IOL removals and med icolegal situ ations. It is
important to rem em ber that a study of 2 eyes, one w ith a -10 D error and the other w ith a
+10 D error, resu lts in a ME of 0.00 D and a MAE of 10.0 D.
Many stud ies on formu la or instru ment accu racy fail to report all the above in form a-
tion. They also rep ort error in IOL power rather than in postoperative refractive error, and
then convert the IOL power to a refractive error u sing an erroneou s conversion factor not
valid for all AL ranges. Still others u se bilateral eyes in the same patient, wh ich detracts
from the statistical quality of resu lts. All these d iscrepancies m ake comparing stud ies
qu ite d ifficu lt.

Fo rm ula Accura c y
After the SRK1 regression formu la became very popu lar, there were rep orts of large
errors clin ically u sing the formu la. I performed a sm all study comparing the SRK to the
Hoffer (1974) theoretic formu la 2 and noted a d ram atic increase in errors u sing SRK. If one
plots the resu lts w ith any regression formu la, it can be seen that it form s a straight line
from the longest eye to the shortest. The same plot u sing a theoretic formu la based on the
actu al optics of the eye show d ram atic changes in the slope of the line as the eye becom es
shorter or longer. Th is is the in herent problem w ith regression formu las in eyes outside
the norm al AL range of 22 to 24.5 m m .
However it wasn’t u ntil Hollad ay pu blished the Hollad ay 1 formu la 3 in 1988, that atten-
tion was paid to th is problem . My large study4 of 450 eyes (by one su rgeon u sing one IOL
style) compared the new Hoffer Q formu la w ith the Hollad ay 1, the SRK/ T5, and the SRK
I and II. The resu lts (Table 23-1) showed that in the norm al range (72% of the eyes) of AL
(22.0 to 24.5 m m) almost all formu las fu nction adequ ately, but that the SRK I formu la
is the lead ing cau se of poor refractive resu lts in eyes outside th is range. The d ifference
between the Hoffer Q and the SRK I and SRK II was statistically sign ificant to a p value
of 0.004 (Fig. 23-1).
The study also showed that the Holladay 1 formu la was the most accu rate in med iu m
long eyes (24.5 to 26.0 m m)—which was 15% of the eyes—and the SRK/ T was more accu rate
 Formulas an d Prog rams: Reg ression an d Th eoretic Formulas 145

Figure 23-1. Error ranges of 450 eyes comparing 2 Figure 23-2. Prediction errors within ± 0.50 D
regression and 3 theoretic formulas. error for the four modern formulas.

Ta b le 2 3 -2 .

HO FFER RECOMMENDED FO RMULA USAGE D EPENDING UPO N THE AL O F THE EYE

Recommended Formula Usage
• Hoffer Q <24.5 mm
• Holladay 1 24.5 to 26.0 mm
• SRK/T >26 mm
• Holladay 2 and Haigis O K, but require more data collection
• Never SRK I or II

in very long eyes (>26.0 m m)—which was 5% of the eyes (Fig. 23-2). In short eyes (<22.0 m m)
(8% of eyes) the Hoffer Q formu la was most accu rate, and this was confirmed (p>0.0001)
in an add itional large study by myself6 (u npublished) of 830 short eyes (supplied by James
Gills, MD) as well as in a mu ltiple-su rgeon study by Holladay. Holladay has postu lated that
the other formu las overestimate the shallow ing of the effective lens position (ELP) in these
very short eyes. The recom mendations made since that study are show n in Table 23-2.
A more recent study 7 was p erformed on 317 eyes, again u sing one style of IOL and
one su rgeon. Since the Hollad ay 2 formu la is not published, the study had to u se the
Hollad ay IOL Consu ltant® computer program to analyze the resu lts w ith each of the for-
mu las. We have recently learned that there were program m ing errors in the Hollad ay IOL
Consu ltant® for the SRK/ T formu la. The resu lts are d isplayed in Table 23-3. The study
showed that the Hollad ay 2 formu la equaled the Hoffer Q in short eyes (<22 m m) but was
not superior to it (Fig. 23-3).
The Hollad ay 2 was not as accu rate as the Hollad ay 1 or Hoffer Q in average AL eyes
(22 to 24.5 m m). What was most d isapp ointing is that the Hollad ay 1 was more accu rate
than the Hollad ay 2 in med iu m long eyes (24.5 to 26 m m). It is in th is range that the
Hollad ay 1 excels.
It app ears that in attempting to improve the accu racy of the Hollad ay 1 formu la in
AL extremes, the add ition of more biometric d ata input has improved the Hollad ay 2
146 Ch apter 23

Ta b le 2 3 -3 .

RESULTS O F ACCURACY O F FO UR THEO RETICAL FO RMULAS O N 317 EYES

USING THE HO LLADAY IO L C O NSULTANT® FO R ANALYSIS
Mean Absolute Error All 317 Eyes
Very All
Short Normal Med Long Max >±2 D
Formula Long Long All Eyes
<22.0 22.0–24.5 24.5–26.0 Error Error
>26.0 <24.5
Holladay 2 0.72 0.56 0.51 0.49 0.50 0.55 –1.60 0%
Holladay 1 0.85 0.42 0.37 0.56 0.43 0.43 –1.44 0%
Hoffer Q 0.72 0.43 0.47 0.58 0.50 0.45 –1.61 0%
SRK/T 0.83 0.46 0.35 0.44 0.36 0.44 –1.45 0%
Average 0.78 0.47 0.42 0.52 0.45 0.47
H-Q
Best H-Q H-1 S/T H-1 S/T S/T
H-2

Shaded = recommended formulas6

Figure 23-3. Mean Absolute Errors (MAE) comparing the Hoffer Q , Holladay 1 & 2, and SRK/T formu-
las. (A) Various AL ranges, and (B) medium long eyes showing the Holladay 2 having the worst MAE
and the Holladay 1 the best.

formu la in the extremes of AL, but deteriorated its excellent perform ance in the norm al
and med iu m long range of eyes (22.0 to 26.0 m m)—wh ich is 82% of the popu lation. It
therefore does not appear to be necessary to collect all the d ata requ ired for the Hollad ay
2 when the Hoffer Q, Hollad ay 1, and SRK/ T on ly requ ire AL and K read ings.
On the other hand, there have been a few stud ies that report there is no formu la AL
dependence, and that the Hoffer Q, Hollad ay 1 and 2 and SRK/ T formu las produce the
same resu lts regard less of the length of the eye.8 There are also stud ies show ing that the
Hoffer Q formu la is the most accu rate in myopic eyes. However, most of these stud ies are
 Formulas an d Prog rams: Reg ression an d Th eoretic Formulas 147

Figure 23-4. Recommended formula usage

depending upon the AL of the eye.

based on aggregates of patients operated on by mu ltiple su rgeons, u sing a variety of IOLs

and wh ile m ixing applanation, im m ersion and IOLMaster biom etric measu rements.
Finally, Hollad ay has stated that “every formu la author is able to prove that h is/ her
formu la is the overall best” (p ersonal com mu n ication). I broke that ru le in 1993, by recom-
mend ing the u se of 3 formu las and proved that my ow n formu la (Hoffer Q) was not the
best in all eyes. Otherw ise, since then, Hollad ay’s d ictu m has held true. It is up to each
ophthalmologist to critically look at the scientific d ata that has been reported and come
to their ow n conclu sion as to what w ill provide their patients w ith the greatest accu racy.
A positive note is that the u se of SRK regression formu las has decreased d ram atically
th roughout the world.
Overall the recom m end ation s for IOL formu la u sage based on AL still stand (Fig.
23-4) and have been statistically proven by a large 8000 eye UK study pu blished in
Janu ary 2011.9

Re fe re nce s
1. Sand ers D, Retzlaff J, Kraff M, et al. Comparison of the accu racy of the Bin khorst,
Colenbrander and SRK im plant power pred iction formu las. J A m Intraocul Implant Soc.
1981;7(4):337–340.
2. Hoffer KJ. Intraocu lar lens calcu lation : The problem of the short eye. Ophthalmic Surg.
1981;12(4):269–272.
3. Hollad ay JT, Prager TC, Chand ler TY, et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–24.
4. Hoffer KJ. The Hoffer Q formu la: A comparison of theoretic and regression formu las
[pu blished correction app ears in : J Cataract Refract Surg. 1994;20(6):677 and J Cataract
Refract Surg. 2007;33(1):2–3]. J Cataract Refract Surg. 1993;19(6):700–712.
5. Retzlaff J, Sanders DR, Kraff MC. Developm ent of the SRK/ T intraocu lar lens implant
p ower calcu lation formu la [pu blished correction ap p ears in : J Cataract Refract Surg.
1990;16(4):528]. J Cataract Refract Surg. 1990;16(3):333–340.
6. Accu racy of Hoffer Q Formu la in 830 Short Eyes. 13th Bien n ial Welsh Cataract Congress,
Cu llen Eye Institute, Gallaria Hotel; Septem ber 9, 1994; Hou ston, TX.
7. Hoffer KJ. Clin ical resu lts u sing the Hollad ay 2 intraocu lar len s power formu la. J Cataract
Refract Surg. 2000;26(8):1233–1237.
148 Ch apter 23

8. Narvaez J, Zim m erm an G, Stu lting RD, Chang DH. Accu racy of intraocu lar len s power
pred iction u sing the Hoffer Q, Hollad ay 1, Hollad ay 2, and SRK/ T formu las. J Cataract
Refract Surg. 2006;32:2050–2053.
9. Aristod emou P, Cartw right N EK, Sparrow JM, Joh n ston RL. Formu la choice: Hoffer Q,
Hollad ay 1, or SRK/ T and refractive outcom es in 8108 eyes after cataract su rgery w ith
biom etry by partial coherence interferom etry. J Cataract Refract Surg. 2011; 37:63–71.
 24

Fo rmu la s a n d Pro g ra ms:

Olse n Fo rmu la
Thomas Olsen, M D

Sig nifica nce o f IOL Po s itio n Pre dic tio n

A
s described in Chapter 17, the position of the IOL after su rgery is an imp ortant fac-
tor in the pred iction of the refractive outcom e of IOL implantation. If the IOL end s
up deep er in the eye than expected, the effect is a hyperopic sh ift in refractive
error. If it end s up more shallow, the effect is a myopic sh ift. The m agn itude of the effect
is such that about +0.7 m m sh ift in ACD correspond s to about +1.0 D sh ift in refraction
w ith a strong dep endence of the AL producing relatively h igher errors in the short eyes.
Although it m ay be h idden to the u ser, all cu rrent IOL power calcu lation formu las
contain their ow n method to arrive at an ELP (Table 24-1). At the time of the early theo-
retical formu las (eg, Colenbrander,1 Hoffer,2 Bin khorst 3), very little was know n regard ing
the actu al position of the implant after su rgery and a fixed value was therefore assu m ed.
It soon became obviou s that the fixed-ACD model was a bad assu mption. The modern
progress in IOL power calcu lation formu las has largely been m ade in the method s to
pred ict the position of the implant after su rgery based on preoperative measu res.
Tod ay there is ind isputable evidence that the postoperative ACD or ELP is positively
correlated w ith the axial length.4 Therefore, to avoid a bias w ith the AL, the pred iction of
the p ostoperative ELP shou ld in som e way be corrected w ith the AL. The “way,” however,
is not easily defined, as it depend s on the natu re of the IOL calcu lation formu la and the
level of am bition.

149
Hoffer KJ. IOL Power (pp. 149 -154).
© 2011 SLACK Incorporated.
150 Ch apter 24

Ta b le 2 4 -1 .

VARIABLES USED TO PREDICT ELP BY VARIO US FO RMULAS

ACD/ELP Predictor Formula/Author
AL Hoffer2 (original), Binkhorst II,7 SRK/T,9 Holladay 1,10 Haigis,11
Hoffer Q ,8 O lsen 6
Tangent of K Hoffer Q 8
Corneal height Fyodorov,12 Holladay 1,10 SRK/T,9 O lsen,6 Holladay 2*
Preoperative ACD Haigis,11 O lsen,6 Holladay 2*
Lens thickness Olsen,6 Holladay 2*
Age Olsen,6 Holladay 2*
Refraction Olsen,6 Holladay 2*

*N ot Published

If the IOL calcu lation formu la is a “th in len s” formu la (and most cu rrent formu las
are), the IOL position is defined as the effective lens position (ELP), ie, the d istance that
correctly pred icts the observed refractive effect when the corneal power and the AL are
know n. One way to proceed is to analyze a large series of cases to deduce the ELP and its
dependence on the axial length, on the corneal power, and other preoperatively defined
m easu res that m ight be con sidered.
If the IOL calcu lation formu la is a “th ick lens” formu la (like the Olsen formu la 5,6), the
IOL position is defined as the actu al lens position (ALP), ie, the d istance from the anterior
central corneal su rface to the anterior central su rface of the IOL. Th is method requ ires a
detailed knowledge of the optic con figu ration of the IOL in order to calcu late the position
of the anterior and the posterior principal planes of the IOL. The empirical method s to
study the ALP dep endence on preoperative m easu res do not d iffer from the “th in lens”
approach, as it takes a large series to derive the exact dependency. Th is is often stud ied
u sing statistical regression method s. It shou ld be realized, however, that becau se of the
d ifferences between a “th in lens” and a “th ick len s” model, the resu lting regression coef-
ficients are not interchangeable between the two approaches.
I believe that a “th ick lens” approach gives a better and more realistic model, as it refers
to the physical p osition and not the virtu al position of the IOL implant. Th is m ay be a
better path for the study of the anatom ical dep endence. More research seem s requ ired to
resolve the issue.

ACD Mo de ls

CORNEAL CURVATURE AND CORNEAL H EIGHT

One of the earliest models for the pred iction of the postoperative ACD was pu blished
by Fyodorov,10 who u sed the base of the anterior spherical segm ent as the reference
 Formulas an d Prog rams: Olsen Formula 151

plane. He prop osed that th is plane cou ld be calcu lated from the corneal cu rvatu re and
the corneal d iam eter, the latter by taking an average value or by u sing the wh ite-to-wh ite
measu rement of the cornea. Th is Fyodorov formu la was intended for iris-clip len ses and
was adopted by myself for anterior cham ber len ses13 and later for posterior cham ber
lenses.14
The Fyodorov “corneal height” formu la was reintroduced by Hollad ay 10 for the calcu-
lation of the so-called “su rgeon factor”, defined as the d istance from the corneal height to
the effective optical plane of the IOL and later adopted by the SRK/ T formu la.9 Both of
these formu las (Hollad ay 1 and SRK/ T) were, in reality, a mod ified Bin khorst formu la.
However, recent work by th is author seem s to ind icate that there is no sign ificant in for-
m ation in the corneal height based on the corneal d iameter, as compared w ith the corneal
cu rvatu re itself. Other pred ictors like the AL, the preoperative ACD, and the lens th ick-
ness (LT) have been fou nd to be more imp ortant.15

PREOPERATIVE ACD
Tod ay, m ost of the new er generation IOL p ow er calcu lation form u las recog n ize the
im p ortance of factors other th an the AL to p red ict the ACD. O ne su ch p red ictor is
the p reop erative ACD w h ich h as been u sed in for m u las su ch as the H aigis for m u la 11
a nd the Olsen for m u la.5 From my stu d ies, the im p ortance of the p reop erative ACD is
ran ked second to the AL in statistical sig n ifica nce as show n by m u ltiple regression
a n alysis.15

LENS THICKNESS
If one accepts the importance of the preoperative ACD in pred icting the postoperative
ACD, it seem s logical to assu m e som e in fluence of the preoperative LT as well. Th is is
due to the th icken ing of the lens w ith age and the statistical negative correlation between
ACD and LT in the norm al eye.20 Despite th is logical assu mption and the fact that most
u ltrasou nd equ ipment is capable of measu ring the LT, it is su rprising how little the LT has
been u sed in ACD pred iction algorith m s. One exception to th is ru le is the Olsen formu la
(wh ich has u sed th is pred ictor since 199516), and more recently it has also been considered
by Norrby.17 Recent stud ies on large series have con firm ed that the LT is important for an
accu rate ACD pred iction, especially in com bination w ith the preoperative ACD.5

Co m m e nts
It shou ld be noted that all stud ies to pred ict the postoperative ACD or the ELP from
preoperatively defined measu res requ ire a norm al anatomy of the eye. If th is is not the
case, such as eyes that have had keratorefractive su rgery or if the AL is changed as a resu lt
of scleral buckling procedu re (see Chapter 19), the statistical model beh ind the pred iction
of the ACD m ay no longer be valid and it m ay be necessary to “norm alize” the anatomy.
Th is is the rationale beh ind the Dou ble-AL method of Hoffer and the Dou ble K-method
of Aram berri.18
Assu m ing the total pred iction error in IOL p ower calcu lation to be the su m of the
error associated w ith the m ain variables, nam ely m easu rem ent of AL, m easu rem ent
of corneal p ower, and estim ation of the p ostop erative ELP; it is p ossible to calcu late
the relative m agn itud e of each of these errors as previou sly show n by the author
u sing u ltrasou nd biom etry.19 The conclu sion d raw n from th is study w as that the AL
152 Ch apter 24

con stituted the largest sou rce of error for the IOL p ower pred iction, outran king the
ACD error. Th is estim ation is based on op tim ized IOL con stants keeping the m ean
nu m erical error zero.
Assu m ing optical biom etry is now the stand ard w ith its mu ch h igher reprodu cibility,
the conclu sion can be m ad e th at the statistical error arising from the estim ation of the
ELP is now the prim ary sou rce of error (>40% of total statistical variance) am ong the
total sou rces of error in IOL p ower calcu lation. The error from the m easu rem ent of the
AL is second (>30%) and the corneal p ower error is th ird (rem ain ing at 10% to 20%).5
Therefore, assu m ing optical biometry and optim ized IOL con stants, the accu racy by
which the postoperative position of the IOL can be pred icted is the m ajor lim iting factor
of modern IOL power calcu lation formu las tod ay.

Re fe re nce s
1. Colenbrand er MC. Calcu lation of the p ower of an iris clip len s for d istant vision. Br J
Ophthalmol. 1973;57(10):735–740.
2. Hoffer KJ. Intraocu lar len s calcu lation: the problem of the short eye. Ophthalmic Surg.
1981;12(4):269–272.
3. Bin khorst RD. The optical d esign of intraocu lar len s implants. Ophthalmic Surg. 1975;
6(3):17–31.
4. Hoffer KJ. Biom etry of the p osterior cap su le. In : Em ery JC, Jacobson AC, ed s. Current
Concepts in Cataract Surgery (Eighth Congress). New York, N Y: Appleton-Centu ry Crofts;
1983:56–62.
5. Olsen T. In: Sham m as H J, ed. Intraocular Lens Calculations. Thorofare, N J: SLACK
Incorp orated ; 2004:27–40.
6. Olsen T. Calcu lation of intraocu lar len s p ower: A review. A cta Ophthalmol Scand.
2007;85(5):472–485.
7. Bin khorst RD. Intraocu lar len s p ower calcu lation. Int Ophthalmol Clin. 1979;19(4):237–
252.
8. Hoffer KJ. The Hoffer Q formu la: A comparison of theoretic and regression formu las
[pu blished correction app ears in : J Cataract Refract Surg. 1994;20(6):677 and J Cataract
Refract Surg. 2007;33(1):2–3]. J Cataract Refract Surg. 1993;19(6):700–712.
9. Retzlaff J, Sand ers DR, Kraff MC. Developm ent of the SRK/ T intraocu lar len s im plant
p ower calcu lation formu la [pu blished correction app ears in: J Cataract Refract Surg.
1990;16(4):528]. J Cataract Refract Surg. 1990;16(3):333–340.
10. Hollad ay JT, Prager TC, Chand ler TY, et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–24.
11. H aigis W. The H aigis formu la. In : Sham m as H J, ed. Intraocular lens power calculations.
Thorofare, N J: SLACK Incorporated. 2004:41–57.
12. Fyodorov SN, Galin MA, Lin ksz A. Calcu lation of the optical power of intraocu lar len ses.
Invest Ophthalmol. 1975;14(8):625–628.
13. Olsen T. Pred iction of intraocu lar len s position after cataract extraction. J Cataract Refract
Surg. 1986;12(4):376–379.
14. Olsen T, Oleson H, Th im K, Corydon L. Pred iction of postop erative intraocu lar len s
cham ber d epth. J Cataract Refract Surg. 1990;16(5):587–590.
15. Olsen T. Pred iction of the effective p ostop erative (intraocu lar len s) anterior cham ber
d epth. J Cataract Refract Surg. 2006;32(3):419–424.
16. Olsen T, Coryd on L, Gim bel H . Intraocu lar len s p ower calcu lation w ith an im p roved
anterior ch am ber d ep th pred iction algorith m . J Cataract Refract Surg. 1995;21(3):313–
319.
 Formulas an d Prog rams: Olsen Formula 153

17. Norrby S, Lyd ah l E, Koranyi G, Tau be M. Clin ical ap plication of the len s h ap tic plane
concep t w ith tran sform ed axial length s. J Cataract Refract Surg. 2005;31(7):1338–1344.
18. Aram berri J. Intraocu lar len s p ower calcu lation after corneal refractive su rgery: Dou ble-K
m ethod . J Cataract Refract Surg. 2003;29(11):2063–2068.
19. Olsen T. Sou rces of error in intraocu lar len s p ower calcu lation. J Cataract Refract Surg.
1992;18(2):125–129.
20. Hoffer KJ. Axial d im en sion of the hu m an cataractou s lens [published correction app ears
in : J Cataract Refract Surg. 1993;111(12):1626]. A rch Ophthalmol. 1993;111(7):914–918.
 25
Fo rmu la s a n d Pro g ra ms:
Acce ssin g Mo d e rn IOL
Po w e r Fo rmu la s
Kenneth J. Hoffer, M D

T
o be able to u se the modern theoretic IOL power formu las, it is easiest to u se a com-
puterized method, rather than doing it by hand as cou ld be done w ith the outd ated
regression formu las. There are several means by which the formulas may be used:
1. Program them you rself from the pu blished formu las.
a. Th is is most fraught w ith d ifficu lties becau se of the errata pu blished for the
Hoffer Q 1 and SRK/ T2 formu las, wh ich are d ifficu lt to find and often ignored.
It wou ld be prudent to request the formu la author to con firm that the program-
m ing has been done correctly prior to u sing them in a clin ical situ ation.
b. The author of the program m ing w ill be legally liable for any error in program-
m ing when u sed in a clin ical setting.
c. The Hollad ay 2 has never been published.
d. Autom atic personalization of IOL con stants is more complicated to program.
e. The Dou ble-K m ethod of formu la calcu lation is more d ifficu lt to program.
2. Software in stalled on A-scan instru ments.
a. Be su re the formu las are du ly licen sed by their resp ective authors to in su re that
they have been program m ed properly. The Hoffer Q and SRK/ T formu las are
the most prone to program m ing errors becau se of the pu blished errata that are
often ignored. Ask the m anu factu rer if the formu la author has approved their
program m ing of their formu la.

155
Hoffer KJ. IOL Power (pp. 155 -162).
© 2011 SLACK Incorporated.
156 Ch apter 25

b. One d raw back is that personalization of IOL constants is u sually not available
and few offer d atabase storage of patient d ata.
3. Software in stalled on the Zeiss IOLMaster.
a. Formu las: H aigis,3 H aigis-L, Hoffer Q, Hollad ay 1,4 SRK II, 5 SRK/ T.
b. Be su re the index of refraction in the Setup screen is set to 1.3375 when input-
ting external K read ings to assu re the Hoffer Q formu la produces the proper
resu lt.
c. Personalization of IOL constants is included along w ith d atabase storage of
patient d ata.
4. Software in stalled on the H aag-Streit LENSTAR LS900.
a. Formu las: Hoffer Q, Hollad ay 1,4 SRK II,5 SRK/ T.
b. A d atabase stores all the patient d ata.
5. H and held program m able calcu lators.
a. It is possible to program these u n its you rself, being carefu l to take into accou nt
the errata associated w ith the Hoffer Q and SRK/ T formu las.
b. Hoffer Program s® is available on a Casio calcu lator. It includes p ersonalization
of IOL con stants.
6. Computer program s sp ecific for IOL p ower calcu lation
a. Hoffer Program s® is available for Windows and MAC.
i. Formu las: H aigis, Hoffer Q, Hollad ay 1, SRK/ T.
ii. All IOL lens constants stored, possible to add new ones.
iii. Autom atic personalization of IOL con stants.
iv. Dou ble-K available for Hoffer Q, Hollad ay 1 and SRK/ T.
v. Includes several method s to calcu late K in LASIK eyes.
vi. No an nu al m aintenance fee.
b. Hollad ay IOL Con su ltant® is available for Window s.
i. Formu las: Hoffer Q, Hollad ay 1, Hollad ay 2, SRK/ T.
ii. All IOL lens constants stored, possible to add new ones.
iii. Autom atic personalization of IOL con stants.
iv. Dou ble-K available only for Hollad ay 2.
v. Excellent piggyback IOL calcu lation.
vi. An nu al m aintenance fee involved.
7. Palm PDA program s.
a. Hoffer Program s® is available for all Palm hand held s, includ ing telephones ru n-
n ing Palm OS Ver 6.0 (Fig. 25-1).
i. Includes autom atic p ersonalization of IOL constants.
ii. Includes clin ical h istory and contact lens method s.
b. DO NOT u se free Palm program dow n load able from a Ru ssian website. All the
formu las were checked by their respective authors and they are all program med
incorrectly.
8. iPhone, iTouch, iPad.
a. Hoffer Program s® available at Apple app store.
 Formulas an d Prog rams: Accessin g Modern IOL Pow er Formulas 157

A B

C D

Figure 25-1. Palm O S screens for Hoffer Programs®. (A) Data entry screen, (B) IO L calculation
screen, (C) Formula personalization screen, (D) Resultant personalization factors.

6
His to ry
The first attempt to program IOL formu las was by Herm an n Gernet in the early 1970s
in Germ any. He d id th is on a m ain-frame computer. I program med my original Hoffer
formu la on the first program m able calcu lator, the Hewlett-Packard H P-65 in 1974; and
later on a lineage of Casio program m able calcu lators. Later, Bin khorst had h is formu la
program med on a Texas Instru ments calcu lator wh ich cam e free w ith the pu rchase of the
new Sonometrics A-scan u n it. After pu blish ing the Hollad ay 1 formu la, Hollad ay m ade
it com mercially available on a Dioptron System calcu lator.
In 1988, I added the Hollad ay 1 to the Casio program, and later also added the SRK/ T
formu la. Th is becam e com mercially available as the “Hoffer Program s®” in 1990. The
Hoffer formu la was replaced w ith the Hoffer Q in 1993. In the same year, the first DOS
and Window s computer program for IOL power was m ade available as Hoffer Program s®.
158 Ch apter 25

Ta b le 2 5 -1 .

FO RMULA RESULTS PUBLISHED IN O RIGINAL ARTICLE7

(Note Hoffer Q the worst.)
Formula Mean SD(D) Range (D)
SRK II +11.94 ±7.07 +4.22 to +21.60
SRK/T +4.40 ±4.34 +0.40 to +11.17
Holladay 1 +2.74 ±4.47 -0.56 to +10.20
Hoffer Q +11.44 ±7.49 +4.08 to +21.70

Ta b le 2 5 -2 .

FO RMULA RESULTS PUBLISHED IN JO URNAL ERRATUM8

Corrections by authors in bold. (Note Hoffer Q the best.)
Formula Mean SD(D) Range (D)
SRK II +11.94 ±7.07 +4.22 to +21.60
SRK/T +4.40 ±4.34 +0.40 to +11.17
Holladay 1 +3.03 ±4.23 -0.56 to +10.20
Hoffer Q +2.80 ±1.83 -4.02 to +5.00

Th is was followed several years later by the Hollad ay IOL Consu ltant® program avail-
able for Window s. Over the years variou s A-scan m anu factu rers added software to their
u ltrasou nd u n its to allow program m ing of variou s formu las. Several times they were
program med incorrectly, lead ing to clin ical errors. Th is stimu lated the formu la authors to
requ ire licen sing of the com mercial u se of their names so they cou ld assu re the software
was program med correctly.

Fo rm ula Pro prie ty

The concept of formu la propriety is often m isu nderstood. First, once it is pu blished, a
formu la is free for anyone to u se. It is not really possible to patent a formu la. However,
to be recogn ized as such, the formu la u sually has the name of its author. The author’s
name can be tradem arked (™) or registered tradem arked (®), thu s requ iring p erm ission
to u se the nam e in com mercial ventu res. Th is allow s the author to authenticate the for-
mu la program m ing and protect patients as well as the reputation of the formu la and its
author. There is no financial gain in licen sing formu las as m ay be su spected. It is for the
protection of the pu blic.
The worst example of th is problem was the m ajor error by the Tomey u ltrasou nd u n it,
which led to a published rep ort by Osh ika 7 and associates in 2001 show ing that in a series
of m icrophthalm ic ped iatric eyes, the Hoffer Q had a mean pred iction error of 11.44 D
compared w ith an error of on ly 2.74 D for the Hollad ay 1 and 4.40 D for the SRK/ T (Table
25-1). The actu al resu lt, when program med correctly, showed that the Hoffer Q formu la
was actu ally more accu rate than the other 4 formu las in th is series (Table 25-2), w ith a
 Formulas an d Prog rams: Accessin g Modern IOL Pow er Formulas 159

Figure 25-2. Holladay IO L Consultant ® screen

emphasizing the data required to calculate the
Holladay 2 formula.

Figure 25-3. Hoffer Programs® screen

showing calculation of the SRK/T formula
in a myopic LASIK eye (23.51 D).

mean error of 2.78 D; a fou r-fold error of almost 9.0 D. The authors pu blished an erratu m 8
wh ich states that ‘‘The error was cau sed by the incorrect power calcu lation program
incorporated into the A-scan in stru m ent (UD-7000, Tomey Corp.), wh ich the authors
failed to notice. Recalcu lation u sing the corrected program ind icated that the Hoffer Q
formu la offered the most accu rate pred iction s for the 5 m icrophthalm ic eyes.’’ It is note-
worthy that their resu lts for the Hollad ay 1 formu la were also incorrect (see the Tables).
Tom ey im m ed iately corrected their error in all their instru ments and also issued an ap ol-
ogy letter to ophthalmologists arou nd the world. Licensing prevents such problem s.

Po pular Pro g ram s

The most popu lar com mercial program s are the IOLMaster, the Hoffer Program s®
System, and the Holladay IOL Consu ltant®, which include several formu las and the ability
to personalize them as well as routines to deal w ith odd clinical situations. The Holladay IOL
Consu ltant® is the only place the Holladay 2 formu la is available. As illu strated in Fig. 25-2,
the Holladay 2 requ ires the collection of other data than the AL and K read ing (green boxes).
The age of the patient, the preoperative refractive error (blue box), horizontal corneal d iam-
eter, phakic preoperative ACD, and phakic lens thickness (red boxes) must also be measu red
and entered to perform the calcu lation. The Aramberri Double-K method is available in the
“Alternate K ” box, but this is not available for the Hoffer Q or SRK/ T formu las.
The most beneficial developm ent has been the availability of the Aram berri Dou ble-K
calcu lation for p ost-refractive eyes. Fig. 25-3 is an example of a Hoffer Program s® screen
of a norm al calcu lation for a myopic AL eye that has had LASIK, yield ing a 23.51 D IOL
160 Ch apter 25

Figure 25-4. Hoffer Programs® screen

showing calculation of the SRK/T formula
in a myopic LASIK eye using the Double-K
(25.54 D).

Figure 25-5. Hoffer Programs screen ®

showing calculation of the Hoffer Q for-
mula in a hyperopic LASIK eye (21.95 D).

for em metropia u sing the SRK/ T formu la. By clicking the “Dou ble-K” box (Fig. 25-4)
and entering the pre-LASIK K read ing, the em m etropic power is now calcu lated as 25.54
D—thu s preventing a 2 D error in IOL power. Th is wou ld have resu lted in more than 2.50
D of hyperopia.
A sim ilar situ ation for a PO hyperopic LASIK eye is show n in Fig. 25-5. By clicking the
“Dou ble-K” box (Fig. 25-6) and entering the pre-LASIK K read ing, the em metropic power
is now calcu lated as 20.59 D—thu s preventing a 1.36 D error in IOL power. Th is wou ld
have resu lted in 1.75 D of myopia.

Re fe re nce s
1. Hoffer KJ. The Hoffer Q formu la: A comparison of theoretic and regression formu las
[published correction app ears in : J Cataract Refract Surg. 1994;20(6):677 and J Cataract
Refract Surg. 2007;33(1):2–3]. J Cataract Refract Surg. 1993;19(6):700–712.
2. Retzlaff J, Sand ers DR, Kraff MC. Developm ent of the SRK/ T intraocu lar len s implant
p ower calcu lation formu la [pu blished correction app ears in : J Cataract Refract Surg.
1990;16(4):528]. J Cataract Refract Surg. 1990;16(3):333–340.
 Formulas an d Prog rams: Accessin g Modern IOL Pow er Formulas 161

Figure 25-6. Hoffer Programs® screen

showing calculation of the Hoffer Q for-
mula in a hyperopic LASIK eye using the
Double-K (20.59 D).

3. H aigis W. The H aigis formu la. In : H J Sham m as, ed . Intraocular Lens Power Calculations.
Thorofare, N J: SLACK Incorporated. 2003:41–57.
4. Hollad ay JT, Prager TC, Chand ler TY, et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–24.
5. Sanders DR, Retzlaff J, Kraff MC. Comparison of the SRK II formu la and the other second
generation formu las. J Cataract Refract Surg. 1988;14(2):136–141.
6. Hoffer KJ. The h istory of IOL p ower calcu lation in North America. In: Kw itko ML,
Kelm an CD, ed s. The History of M odern Cataract Surgery. The H ague, Netherland s: Kuglen
Pu blication s; 1998:193–208.
7. Osh ika T, Im amu ra A, Am ano S, et al. Piggyback fold able intraocu lar len s implantation
in patients w ith m icrophthalmos. J Cataract Refract Surg. 2001;27(6):841–844.
8. Osh ika T, Im amu ra A, Am ano S, et al. Erratu m . J Cataract Refract Surg. 2001;27:1536.
 26

Fo rmu la s a n d Pro g ra ms:

Fo rm u la Pe rso n a liza t io n
Kenneth J. Hoffer, M D

T
he concept of personalizing a formu la based on a su rgeon’s past experience and
d ata was introduced by Retzlaff1 after the introduction of regression formu las
u sing an A constant in place of an anterior cham ber depth (ACD). He proposed
that the accu racy of regression formu las cou ld be improved u sing th is m ethod. Here is
how it works.
After an eye has had an IOL implanted, the follow ing d ata is collected :
1. Preop erative axial length (AL)
2. Preoperative corneal p ower (K)
3. Implanted IOL power (P)
4. Stable postoperative refractive error (R)
Of the 5 parameters in all IOL power formu las, the on ly u n know n factor now is the
A constant. Using the regression formu la, the A constant can be back-calcu lated. Th is is
the ideal A con stant that th is eye wou ld have needed to produce a perfect error-free IOL
p ower pred iction resu lt. Th is can then be performed on a series of eyes and the average
of all the ideal A constants can be calcu lated, producing the su rgeon’s ideal personalized
A con stant. But, as Retzlaff recom m ended, the series of eyes mu st be all the same, eg:
1. Same IOL model and m anu factu rer
2. Sam e su rgeon
3. Same cataract procedu re and IOL placement (eg, in or out of the bag)
4. Same AL measu ring equ ipment
5. Same keratom eter typ e
163
Hoffer KJ. IOL Power (pp. 163-166).
© 2011 SLACK Incorporated.
164 Ch apter 26

Eyes w ith postoperative su rprises or acu ities worse than 20/ 40 shou ld not be included
in the analysis, due to the poor accu racy in obtain ing a precise refractive error. The
m in imu m series size shou ld be 20 eyes and, as more eyes are added, the benefit of the
p ersonalization increases.
Th is concept took on more importance in 1988 when Hollad ay 2 was w riting h is
Hollad ay 1 formu la. In attempting to better pred ict the postoperative IOL position (ELP),
he u sed the Fyodorov corneal height formu la to pred ict the d istance from the anterior
vertex of the cornea to the iris plane by u sing the AL and K. Since the IOL sat farther
beh ind the iris plane, there was an add itional d istance that was m issing; the d istance
from the iris plane to the principle plane of the IOL. He gave the value the name surgeon
factor (SF). Since there was no way to calcu late th is nu m ber preop eratively, he back-calcu-
lated th is value from a series of eyes previou sly op erated on, took the average, and that
becam e the SF to u se on futu re cases. He also pu blished a series of formu las to convert an
ACD or A con stant into a SF for a given lens style.
l When the SRK/ T3 theoretic formu la was developed, they d id not u se th is same
m ethod to personalize the new A constant to be u sed w ith it. Instead they u sed the
old regression formu la for personalization.
l When I4 developed the Q formu la for the Hoffer Q, I back-calcu lated for a p ersonal-
ized ACD and called it the pACD.
l When H aigis 5 develop ed h is formu la, he replaced corneal power w ith the preopera-
tive ACD to pred ict ELP. He created th ree constants; a 0 (constant), a 1 (con stant for
AL), and a 2 (con stant for ACD). Single optim ization (p ersonalization) of the H aigis
formu la lead s to a personalized a 0 and its resu lts are com m ensu rate w ith the Hoffer
Q, Hollad ay, and SRK/ T formu las. Triple optim ization resu lts in personalized a 0, a 1,
and a 2, and greater accu racy than the other formu las, but th is requ ires a series of
500 to 1000 eyes of one lens style—wh ich m ight be d ifficu lt for the average su rgeon
to acqu ire.
Several stud ies have proven that formu la personalization defin itely improves formu la
accu racy in a clin ically sign ificant way. Since it involves solving qu ad ratic equation s, the
m athem atics involved in the back-calcu lation s is qu ite cu m bersom e. Both Hollad ay and
I needed to resort to u n iversity m ath professors to accomplish it. Since computers w ill
be needed, some in stru m ents such as the IOLMaster w ill p erform personalization. It has
also been a prom inent featu re of both available IOL power calcu lation program s: Hoffer
Program s® and Hollad ay IOL Con su ltant® (Fig. 26-1).
 Formulas an d Prog rams: Formula Person alization 165

Figure 26 -1. Personalization screen for Hoffer Programs® (left) and Holladay IO L Consultant ®
(right).

Re fe re nce s
1. Retzlaff J. Calcu lating the su rgeon’s p ersonal A-con stant. In : Retzlaff J, Sand ers DR,
Kraff MC, ed s. Lens Implant Power Calculation M anual, 3rd ed . Thorofare, N J: SLACK
Incorp orated. 1990:12–13.
2. Hollad ay JT, Prager TC, Chand ler TY, et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–24.
3. Retzlaff J, Sand ers DR, Kraff MC. Developm ent of the SRK/ T intraocu lar len s implant p ower
calcu lation formu la. J Cataract Refract Surg. 1990;16(3):333–340. Erratu m : 1990;16(4):528.
4. Hoffer KJ. The Hoffer Q formu la: a com parison of theoretic and regression formu las
[pu blished correction app ears in : J Cataract Refract Surg. 1994;20(6):677 and J Cataract
Refract Surg. 2007;33(1):2–3]. J Cataract Refract Surg. 1993;19(6):700–712.
5. H aigis W. The H aigis formu la. In: H J Sham m as, ed. Intaocular Lens Power Calculations.
Thorofare, N J: SLACK Incorp orated. 2003:41–57.
 27
Sp e cia l Circu msta n ce s:
AL Me a s u re m e n t in
Sta p h ylo ma Eye s
H. John Shammas, M D

T
he presence of a p osterior pole staphylom a m ay be the most frequent cond ition in
wh ich a precise AL m easu rement m ay not be obtained. A p osterior p ole staphy-
lom a is u su ally present in severe axial myopia, and the obliqu ity of the m acu lar
plane to the visual axis is responsible for sign ificant AL d ifferences w ith in a sm all area.
Occasionally it can be u ndetected if associated w ith a m atu re cataract in an eye w ith
u n ilateral axial myopia.1
In A-scan biometry, the spatial orientation of the staphylom atou s posterior pole su r-
face cau ses an oblique rather than orthogonal interception of the u ltrasou nd beam by
the vitreoretinal interface.2 Th is cau ses a saw-toothed asp ect of the p eak of the vitreo-
retinal interface (Fig. 27-1), wh ich prevents precise localization of the foveolar area, and
it becomes extrem ely d ifficu lt to select the right echogram (supposed to coincide w ith
the visual axis). Measu ring the AL w ith an IOLMaster can yield more accu rate resu lts,
especially if the patient m aintain s good fixation.
In d ifficu lt cases, a B-mode gu ided biometry is the preferred tech nique, provided that
the B-scan u nit allows for such a measu rement. A simplified im mersion bath is created
u sing the m anually opened eyelid fissu re filled w ith methylcellu lose gel. The probe’s tip is
held in su spension w ithin the gel layer, w ithout touching the corneal su rface. An optim al
axial section of the eye is obtained and a control vector (seen on the screen as a superim-
posed dotted line) is aligned w ith the visual axis on the frozen im age. An A-scan biom-
etry is reconstructed along the control vector line (Fig. 27-2). It is important to visualize
all echospikes correspond ing to the anterior corneal su rface, anterior and posterior lens
su rfaces, and the vitreoretinal interface in a location temporal to the optic nerve head. This
tech nique is more popu lar in Eu rope and is routinely u sed in some centers.3
167
Hoffer KJ. IOL Power (pp. 167-168).
© 2011 SLACK Incorporated.
168 Ch apter 27

Figure 27-1. Immersion ultrasound

measurement of a cataractous eye
with posterior pole staphyloma.
Note the low amplitude of the
retinal spike, which in advanced
cases might be difficult to iden-
tify.

Figure 27-2. B-scan examination

shows the increased curvature of
the posterior pole in the catarac-
tous eye with a central staphy-
loma. The A-scan is taken from
the central vector axis. (Reprinted
from O. Berges.)

In add ition, the B-scan offers a good evalu ation of the vitreou s cavity and of the retina,
which is esp ecially important in presence of a den se cataract and where the fu ndu s can-
not be evalu ated.

Re fe re nce s
1. Sham m as H J, Milkie CF. Matu re cataracts in eyes w ith u n ilateral axial myopia. J Cataract
Refract Surg. 1989;15(3):308–311.
2. Fernandez-Vigo J, Castro J, Diaz J, Cid MR. Ultrason ic form s of posterior staphylom a. A nn
Ophthalmol. 1990;22(10):391–394.
3. Berges O, Siah m ed K, Pu ech M, Perrenoud F. B-mod e gu id ed biom etry. In : H J Sham m as,
ed . Intraocular Lens Power Calculations. Thorofare, N J: SLACK Incorporated. 2004:159–170.
 28

Sp e cia l Circu msta n ce s:

Silico n e Oil-Fille d Eye s
Wolfgang Haigis, M S, PhD

D
ifferent m ed ia are characterized by d ifferent propagation velocities for u ltra-
sou nd as well as light waves (eg, the speed of an u ltrasou nd pu lse th rough
silicone oil is considerably slower—≈1/ 3—than th rough vitreou s). Accord ingly,
it takes longer for the sou nd to cross the eye and it w ill retu rn later than norm al; ju st as
if it had traveled a longer d istance. Th is is why a norm al eye w ith its vitreou s replaced by
silicone oil app ears to be some 33 m m long when measu red as a norm al phakic eye. The
necessary correction to the AL is nearly 10 m m .
The sam e physics ap ply to the propagation of laser light th rough silicone oil. The
effect, however, is m ore th an an ord er of m agn itud e sm aller th an w ith u ltrasou nd .
Th is is du e to the d ifferences in sp eed s of light for vitreou s and silicone oil being on ly
arou nd 4%; the (group) refractive ind ex of silicone oil is on ly slightly h igher than th at of
vitreou s. Com pared to an IOLMaster m easu rem ent in ph akic m od e, a correction factor
of som e 0.7 m m has to be su btracted . Th is correction is autom atically ap plied w hen the
in stru m ent’s “AL Settings” for silicone oil-filled eyes are chosen.
There is a fu rther d ifference between u ltrasou nd and optical biometry: w ith u ltra-
sou nd, segmental measu rements can be perform ed —NOT so w ith the IOLMaster.
Consequently, it is principally possible w ith u ltrasou nd to determ ine and correct the
vitreou s d istance in silicone oil for each eye ind ividu ally, wh ile the correction factor for
IOLMaster biometry had to be derived for an average eye. Thu s, a short eye m ay be falsely
too short (≈50 µm) and a long eye m ay be falsely too long (≈100 µm). Clin ically, these
effects don’t play a role.1-3

169
Hoffer KJ. IOL Power (pp. 169 -170).
© 2011 SLACK Incorporated.
170 Ch apter 28

Re fe re nce s
1. Parravano M, Oddone F, Sampalm ieri M, Gazzan iga D. Reliability of the IOLMaster in
axial length evalu ation in silicone oil-filled eyes. Eye. 2007;21(7):909–911.
2. H abibabad i H F, H ashem i H, Jalali KH, Am in i A, Esfahan i, MR. Refractive outcom e of
silicone oil removal and intraocu lar len s implantation u sing laser interferom etry. Retina.
2005;25(2):162–166.
3. Dietlein TS, Roessler G, Lu ke C, et al. Signal qu ality of biom etry in silicone oil-filled eyes
u sing partial coherence laser interferom etry. J Cataract Refract Surg. 2005;31(5):1006–1010.
 29
Sp e cia l Circu msta n ce s:
Un ila te ra l Hig h Myo p e s
a n d Hyp e ro p e s
Kenneth J. Hoffer, M D

A
vexing clin ical problem faces the cataract su rgeon when dealing w ith a relatively
u n ilateral cataract in a patient w ith bilateral h igh ametropia. The d ilem m a is
whether to m ake the su rgical eye em metropic or attempt to m atch the large
ametropia of the other eye, wh ich m ay never need su rgery. If the su rgical eye is m ade
em metropic (or relatively so), the resu ltant an isometropia w ill be intolerable. The on ly
solution wou ld be to wear a contact len s on the non-su rgical eye. If the su rgical eye is
m ad e refractively equ ivalent to the other eye, the patient m isses the opportu n ity to be
em metropic for the first tim e in their life. Th is is a m issed opportu n ity, especially if they
do w ind up developing a cataract in the 2nd eye. Of cou rse it wou ld be p ossible to p erform
a lens exchange or a piggyback IOL in the operated eye at a later d ate, but that wou ld
cau se add itional trau m a to the eye.
I have always referred to the em metropia option as “going for the brass ring.” Th is
option is possible if:
1. The patient is successfu l in a trial of monocu lar contact len s wear on the non-su rgi-
cal eye.
2. The patient desires a phakic IOL in the other eye.
3. The patient desires a clear lens extraction on the other eye for refractive pu rposes.
4. The patient wou ld prefer ignoring the am etropic eye and u se the monocu lar vision
of the newly em metropic eye.
It has been my exp erience to convince most patients to accept a monocu lar CL,
or ignore the other eye and go for the “brass ring” of em metropia. Tod ay it is also

171
Hoffer KJ. IOL Power (pp. 171-172).
© 2011 SLACK Incorporated.
172 Ch apter 29

conceivable to place a piggyback lens over the em metropic IOL to provide a refractive
error that m atches the non-su rgical eye; th is cou ld easily be removed at a later d ate if the
other eye u ltim ately needed su rgery. Most stud ies have show n that the SRK/ T formu la 1
is more accu rate for IOL p ower in extrem e axial myopes and it is recom m ended that it be
u sed for these eyes. In some cases a plano or negative-p owered IOL m ay be requ ired.
H aigis has recom m ended to m e (p ersonal com mu n ication) the u se of special pACD
for very low and negative IOL powers when u sing the Hoffer Q formu la. Using an Alcon
MA60, for example, he recom mend s a pACD of 15.94 for powers from +5.0 to 0.00 D and
a pACD of -5.25 for powers from 0.00 to -5.0 D.
Since h igh ly myopic eyes are more prone to retinal detach ment follow ing cataract su r-
gery, it is w ise to be more conservative in th is regard and defray clear len s extraction in
the non-cataractou s eye. On the other hand, th is is not the case w ith extrem e hyperop es
and it m ay be a better choice to perform clear lens extraction, wh ich w ill resu lt in a deep er
anterior cham ber and less chance for angle closu re glaucom a. Most stud ies have show n
that the Hoffer Q formu la 2 and the Hollad ay 2 formu la 3 are more accu rate for IOL power
in axial hyperopes (<22 m m) and it is recom m ended that it be u sed for these eyes.
Obtain ing em metropia in these eyes m ay requ ire very h igh powers that are not com-
m ercially available in you r favorite IOL style. It is w ise to check w ith all IOL m anu factu r-
ers to see if they have the power you need or are w illing to sp ecial order it for you. If that
is not possible, it w ill be necessary to piggyback two IOLs and attention mu st be paid to
the calcu lation s needed.
The anterior IOL w ill force the p osterior IOL more posterior. Th is w ill decrease its
effective power by moving its focal point beh ind the retina, cau sing hyp eropia. The
p osterior movement is estim ated to be about 50% of the th ickness of the IOL, wh ich can
be calcu lated. Therefore, the total power needed has to be increased to m ake up for th is
and it shou ld be split between the two IOLs, such that the m ajority of the p ower is in
the posterior IOL. If there is an error, th is w ill also m ake it easier to remove the th in ner
anterior IOL.
For som e patients w ith m ilder am etropia, it m ay be advisable to plan a target refraction
that is a comprom ise between em metropia and the ametropia of the other eye such that
an isometropia does not resu lt.
In su m m ary, it is extremely important to d iscu ss these options w ith the patient and
fu lly explain the advantages and d isadvantages of each plan and allow them to m ake
the decision as to wh ich option to u se. It is also prudent to m ake extensive notes of these
conversations in the patient’s chart.

Re fe re nce s
1. Retzlaff J, Sand ers DR, Kraff MC. Developm ent of the SRK/ T intraocu lar len s implant
p ower calcu lation formu la [pu blished correction app ears in : J Cataract Refract Surg.
1990;16(4):528]. J Cataract Refract Surg. 1990;16(3):333–340.
2. Hoffer KJ. The Hoffer Q formu la: A comparison of theoretic and regression formu las
[published correction app ears in : J Cataract Refract Surg. 1994;20(6):677 and J Cataract
Refract Surg. 2007;33(1):2–3]. J Cataract Refract Surg. 1993;19(6):700–712.
3. Hollad ay JT, Prager TC, Chand ler TY, et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–24.
 30
Sp e cia l Circu msta n ce s:
Pe n e tratin g Ke rato p last y
a n d Sca rre d Co rn e a s
Kenneth J. Hoffer, M D

Co rne al Trans plant Eye s

T
he problem w ith IOL power calcu lation in eyes schedu led for a com bined proce-
du re of cataract/ IOL and penetrating keratoplasty (triple procedu re) is attempting
to pred ict what the corneal p ower w ill be after the corneal tran splant. Some have
suggested u sing either:
1. The corneal power of the other eye (if it is obtainable), or
2. Using an average of the su rgeon’s post-tran splant corneal powers
The problem w ith these two options is that published reports show a very large range
of pred iction and refractive errors. Th is is especially u n fortu nate for patients that have
already su ffered enough from the p oor vision due to corneal d isease.
In 1986, I1 pu blished a suggestion that refractive resu lts follow ing a com bined proce-
du re wou ld be better if the IOL implantation were performed as a second ary procedu re
after the corneal transplant has settled dow n. Thu s one cou ld either perform the corneal
transplant alone and later p erform the cataract/ IOL su rgery, or the corneal transplant
can be com bined w ith cataract removal followed by a second ary lens implant. In 1990,
Geggel2 reported excellent refractive resu lts u sing th is two-step approach. More than
two-th ird s of h is eyes attained 20/ 40 or better u ncorrected visu al acu ity. Ninety-five
percent of the eyes were w ith in ±2.00 D of the target postoperative refractive error (Fig.
30-1). What is most clin ically sign ificant, is that the total range of postop erative refractive
error was on ly from +1.75 to -3.87 D instead of from +5.08 to -4.75 D if a triple procedu re

173
Hoffer KJ. IOL Power (pp. 173-174).
© 2011 SLACK Incorporated.
174 Ch apter 30

Figure 30-1. Visual results of Geggel study delay- Figure 30-2. Range of error of Geggel study delay-
ing lens implant until after PK. ing lens implant until after PK.

had been perform ed u sing the su rgeon’s cu stom ary K read ing selection. The total range
of error d ropped from 9.83 D to 5.62 D, a 57% decrease (Fig. 30-2). Though some m ay feel
that two operations is a d isadvantage, the refractive resu lts are certain ly an improvem ent
for the patients.
An alternative to th is concept wou ld be to correct any residu al ametropia after the
triple procedu re by implanting a second ary piggyback toric IOL or toric phakic IOL.

Co rne al Scar Eye s

The problem of getting an accu rate corneal p ower measu rement in eyes w ith corneal
scarring and irregu lar astigm atism has not received much attention. Cu a et al3 stud ied
th is in 2 eyes need ing IOL exchange due to large postoperative IOL “su rprises” of +5 D
and -7.50 D each. They compared 6 method s to ascertain the corneal power and fou nd
the hard contact len s over refraction method to be the most accu rate; decreasing the error
they wou ld have obtained w ith the m anu al keratometer of +4 to +5 D to -0.4 to -1.6 D. Th is
m ay be a u sefu l clin ical option in such cases.

Re fe re nce s
1. Hoffer KJ. Triple procedu re for intraocu lar len s exchange. A rch Ophthalmol. 1987;105(5):609–
610.
2. Geggel H S. Intraocu lar len s implantation after p enetrating keratoplasty: Improved u naid-
ed visu al acu ity, astigm atism , and safety in patients w ith com bined corneal d isease and
cataract. Ophthalmol. 1990;97(11):1460–1467.
3. Cu a IY, Qazi MA, Lee SF, Pep ose JS. Intraocu lar len s calcu lation s in patients w ith corneal
scarring and irregu lar astigm atism . J Cataract Refract Surg. 2003;29(7):1352–1357.
 31

Sp e cia l Circu msta n ce s:

Ra d ia l Ke ra to to my Eye s
Giacomo Savini, M D

I
n 1985, 6 years after rad ial keratotomy (RK) was introduced in the Un ited States, the
problem of power calcu lation in RK eyes was first rep orted by Koch and associates,1
and later by others.2,3
Correctly measu ring the corneal power in eyes that have u ndergone RK is a d ifficu lt
task due to the size of the optical zone (wh ich is u su ally sm aller than the area evaluated
by keratometers or topography system s) so that measu rem ents are likely to be p erformed
in the region of the knee between the treated portion of the cornea and the flattened
optical zone (Fig. 31-1). Th is m ay produce an overestim ation of the corneal power, lead-
ing to an u nderestim ation of the IOL power and resu lting in u nexpected postop erative
hyperopia.
Inaccu racy of corneal power measu rem ents m ay also be related, as in the case of
excimer laser su rgery, to the u n reliability of the keratometric index (eg, 1.3375 or 1.3315)
—wh ich is know n to assu me a stable ratio between the anterior and posterior corneal
cu rvatu re. Although several authors have previou sly stated that in incisional tech n iques
like RK there is no loss of corneal tissue and both anterior and p osterior corneal cu rva-
tu res deform in parallel, so that the ratio between them is m aintained. Actu ally there
are no stud ies to con firm th is statem ent. Using a Scheimpflug cam era, we have recently
observed that such a ratio is not m aintained (d ata u npu blished) and that the keratom etric
index shou ld be adju sted in relation to the nu m ber of rad ial incisions.

175
Hoffer KJ. IOL Power (pp. 175 -178).
© 2011 SLACK Incorporated.
176 Ch apter 31

Figure 31-1. The optical zone following RK (left) is considerably smaller than after myopic
LASIK (right).

It shou ld also be observed that the u npred ictability of IOL power calcu lation after RK
can also be cau sed by the mechan ical in stability of the cornea follow ing incisional su r-
gery. Once phacoemu lsification is perform ed, the RK incisions m ay temporarily reopen
as if the RK incision procedu re had ju st been carried out. Th is instability m ay temporarily
exacerbate central flatten ing and p eripheral bu lging, wh ich sometim es m ay persist.
However, the d iscrepancy between corneal topography m easu rements and true cor-
neal power (as determ ined by the Clin ical H istory Method 4,5 [CH M]) seem s less than in
the case of PRK and LASIK.
When phacoemu lsification and IOL implantation are p erformed in eyes that have
previou sly u ndergone RK, the choice of available m ethod s to calcu late the keratom etric
p ower (d iopters) is con siderably narrower w ith respect to the cases that received excim er
laser su rgery. If all preoperative d ata and the postoperative refraction are know n, the
CH M has been considered the stand ard for qu ite a long time; its reliability, however,
m ay be lim ited by the frequent cases that experienced hyperopic sh ift after RK. For
th is reason, corneal topography probably represents a better option than CH M in these
eyes. In a theoretical study by Stakheev and Balashevich,6 the Sim-K generated by VKG
seemed the most accu rate among measu red keratom etric values. Other stud ies have also
been rep orted.7-10 Due to the sm all RK optical zones (u sually <3.5 m m), it is now rec-
om m ended to d iscard the Sim-K values and con sider the more central area, such as the
th ird ring or the Average Central Corneal Power (ACCP) of the TMS top ographer (Tom ey,
Erlangen, Germ any) or the Effective Refractive Power (EffRP) in the Hollad ay Diagnostic
Su m m ary 9 of the EyeSys Corneal Analysis System (EyeSys Vision, Hou ston, TX). A recent
study by Aw wad 11 and coauthors has show n that entering these values into Aram berri
Dou ble-K formu las lead s to accu rate resu lts in IOL p ower calcu lation in these RK eyes.

Re fe re nce s
1. Koch DD, Liu JF, Hyd e LL, Rock RL, Em ery JM. Refractive complication s of cataract su r-
gery after rad ial keratotomy. A m J Ophthalmol. 1989;108(6):676–682.
2. Cellikol L, Pavlop ou los G, Wein stein B, Cellikol G, Feld m an ST. Calcu lation of intra-
ocu lar len s p ower after rad ial keratotomy w ith com puterized vid eokeratography. A m J
Ophthalmol. 1995;120(6):739–750.
 Special Circumstan ces: Radial Keratotomy Eyes 177

3. Chen L, Man n is MJ, Salz JJ, Garcia-Ferrer FJ, Ge J. Analysis of intraocu lar len s p ower
calcu lation in p ost-rad ial keratotomy eyes. J Cataract Refract Surg. 2003;29(1):65–70.
4. Hollad ay JT. IOL calcu lation s follow ing rad ial keratotomy su rgery. Refract Corneal Surg.
1989;5:36A.
5. Hoffer KJ. Calcu lation of intraocu lar len s p ower in p ost-rad ial keratotomy eyes. Ophthalmic
Practice (Canad a). 1994;12(5):242–243.
6. Stakheev AA, Balashevich LJ. Corneal power determ ination after previou s corneal refrac-
tive su rgery for intraocu lar len s calcu lation. Cornea. 2003;22(3):214–220.
7. Kim SH, Lee JH. Vid eokeratography to calcu late intraocu lar len s p ower after rad ial kera-
totomy. J Refract Surg. 2004;20(3):284–286.
8. Packer M, Brow n LK, Hoffm an RS, Fine IH. Intraocu lar lens power calcu lation after inci-
sional and therm al keratorefractive su rgery. J Cataract Refract Surg. 2004;30(7):1430–1434.
9. Hollad ay JT. Corneal top ography u sing the Hollad ay Diagnostic Su m m ary. J Cataract
Refract Surg. 1997;23(2):209–221.
10. Maed a N, Klyce SD, Smolek MK, McDonald MB. Disparity between keratom etry-style
read ings and corneal p ower w ith in the pupil after refractive su rgery for myopia. Cornea.
1997;16(5):517–524.
11. Aw wad ST, Dwarakanathan S, Bow m an W et al. Intraocu lar len s p ower calcu lation
after rad ial keratotomy: estim ating the refracting corneal p ower. J Cataract Refract Surg.
2007;33(6):1045–1050.
 32
Sp e cia l Circu msta n ce s:
Po st -Lase r Re fra ct ive
Su rg e ry Eye s
Kenneth J. Hoffer, M D

T
here are 3 m ain cau ses for the errors we see in IOL p ower calcu lation for eyes that
have had corneal refractive su rgery to correct ametropia.

1. Instrument Error. The problem of IOL power calcu lation errors in corneal refractive
su rgery eyes was first described by Koch et al1 in 1989. The first problem that arises
is that the instru ments we u se can not accu rately measu re the corneal p ower needed
in the IOL p ower formu la in eyes that have had rad ial keratotomy (RK), photore-
fractive keratectomy (PRK), laser-assisted intrastrom al keratom ileu sis (LASIK) and
laser-assisted epithelial keratom ileu sis (LASEK). Th is m ajor cau se of error is due to
the fact that most m anu al keratometers measu re at the 3.2 m m zone of the central
cornea, wh ich often m isses the central flatter zone of effective corneal power; the
flatter the cornea, the larger the zone of measu rem ent and the greater the error. The
instru ments u su ally overestim ate the corneal power, lead ing to a hyperopic refrac-
tive error p ostoperatively.
2. Index of Refraction Error. The assu m ed index of refraction of the norm al cornea
is based on the relationsh ip between the anterior and posterior corneal cu rvatu res.
Th is relationsh ip is changed in PRK, LASIK, and LASEK but not in RK eyes. RK
cau ses a relatively prop ortional equal flatten ing of both the front and back su rface
of the cornea, leaving the index of refraction relationsh ip relatively the same. The
other refractive procedu res flatten the anterior su rface but not the posterior su rface
thu s changing the refractive index calcu lation, wh ich creates an overestim ation of

179
Hoffer KJ. IOL Power (pp. 179 -194).
© 2011 SLACK Incorporated.
180 Ch apter 32

the corneal power by approxim ately 1 D for every 7 D of refractive su rgery correc-
tion obtained. A m anu al keratom eter m easu res on ly the front su rface cu rvatu re of
the cornea and converts the rad iu s (r) of cu rvatu re obtained to d iopters (D) u sing
an index of refraction (IR) of u su ally 1.3375. The formu la to change from d iopters to
rad iu s is [r = 337.5/ D] and from rad iu s to d iopters is [D = 337.5/ r].
3. Formula Error. Most of the modern IOL p ower formu las (Hoffer Q,2 Hollad ay 1,3
and SRK/ T4 —but not the H aigis 5) u se the AL and corneal power (K) read ing to
pred ict the position of the IOL postoperatively. The flatter than norm al K in RK,
PRK, LASIK, and LASEK eyes cau ses an error in th is pred iction becau se the anterior
cham ber d imen sion s do not really change in these eyes.

His to ry o f S o lutio ns
In 1989, Hollad ay 6 was the first to pu blish and popu larize two method s to attempt
to pred ict the true corneal power in refractive su rgery eyes. I referred to them as the
Clin ical H istory Method and the Contact Lens Method.7,8 The latter was first described
by Frederick Rid ley 9 in the Un ited Kingdom in 1948 and introduced in the Un ited States
by Soper and Goffm an 10 in 1974. Over the years m any researchers and authors have pro-
p osed mu ltiple m ethod s to solve th is problem. No one procedu re has yet to be proven to
be the most accu rate in all cases.
In th is regard Giacomo Savin i of Bologna, Italy and I collaborated over a 2-year period
to create an Excel spread sheet tool that wou ld autom atically calcu late most all the pro-
p osed method s and also provide a place to store all the d ata collected and entered. All
the in form ation cou ld be stored in one place and it cou ld be printed out on one sheet and
stored in the patient’s chart. The Hoffer/ Savin i LASIK IOL Power Tool (Fig. 32-1) was
fin ished on Ju ly 4, 2007 and can be dow n loaded at no cost from w w w.EyeLab.com by
clicking on the IOL Power button and then the Hoffer/Savini button.
In the creation of the Tool, we d ivided all the pu blished method s into those that
attempt to pred ict the true power of the cornea and those that fudge the target IOL power
calcu lated w ith the stand ard d ata. We then d ivided each group into those method s that
need h istorical d ata regard ing the statu s of the patient’s eye prior to refractive su rgery
and those that do not need any h istorical d ata.
Before fin ish ing the Tool, we asked each formu la author to beta test it to m ake su re
they agreed w ith ou r calcu lation s and assu mptions. We have converted formu la abbre-
viations to m aintain con sistency. The legend for these abbreviations is listed on Sheet #3
in the Tool and at the end of th is chapter.
ASCRS has also developed a lim ited ad aptation of th is concept on their website at
w w w.ascrs.org. Several method s are u sed to calcu late an IOL power, but it has the short-
com ing of not u sing the Hoffer Q or Hollad ay 2 formu las for short AL hyperopic LASIK
eyes—wh ich have been show n to be the most accu rate in short eyes.
 Special Circumstan ces: Post-Laser Refractive Surg ery Eyes 181

Figure 32-1A. Hoffer/Savini LASIK Tool for refractive surgery IO L power (empty of
data).

Figure 32-1B. Hoffer/Savini LASIK Tool for refractive surgery IO L power (all data
entered).
182 Ch apter 32

Me tho ds to Es tim a te
True Po s to pe ra tive Co rne al Po w e r
THOSE NEEDING CLINICAL H ISTORY
Clinical History Method
K = KPRE + RPRE – RPO or [K = KPRE + RCC]
This m ethod 1-9 is based on the fact that the final change in refractive error the eye
obtains from su rgery was due on ly to a change in the effective corneal power. If th is
refractive change the patient experienced is algebraically added to the presu rgical cor-
neal power, we w ill obtain the effective corneal power the eye has now. Obviou sly th is
requ ires knowledge of the K read ing and refractive error prior to refractive su rgery.
Originally it was recom mended to vertex-correct the refractive errors to the corneal
plane. Odenthal et al11 showed that clin ical resu lts were better if they were not cor-
rected. We have decided to u se vertex correction in the Hoffer/ Savin i Tool becau se th is is
more scientifically accu rate. Several IOL power calcu lation computer program s calcu late
the Clin ical H istory method autom atically when needed (eg, Hoffer Program s ® and
Hollad ay IOL Consu ltant ®).

Hamed-Wang-Koch Method12
K = TKPO – (0.15 * RC) – 0.05
Th is method requ ires knowledge of the refractive change from the su rgery and the
p ostoperative Sim-K from the top ography u n it.

Speicher13 (Seitz14,15) Method

K = 1.114 * TKPO – 0.114 * TKPRE
Th is m ethod requ ires obtain ing the pre- and postoperative top ograph ic Sim-Ks.

Jarade Formula16
K = TKPRE – (0.376 * (TKPOr – TKPREr )/ (TKPOr * TKPREr )
Th is m ethod requ ires obtain ing the pre- and p ostoperative topograph ic Sim-Ks in
rad iu s of cu rvatu re, not d iopters.

Ronje Method17
K = KPOFLAT + 0.25 * RC
Th is method requ ires knowledge of the refractive change from the su rgery and the
p ostop erative flattest K read ing measu red now.

Adjusted Refractive Index Methods

These method s attempt to “correct” the index of refraction to better pred ict the corneal
p ower. The first two m ethod s requ ire know ing the su rgically induced refractive change
at the spectacle plane and the average rad iu s of cu rvatu re of the cornea now. The th ird
m ethod requ ires know ing the su rgically induced refractive change at the corneal plane
and the average rad iu s of cu rvatu re of the cornea now.
1. Savin i18 Method : K = ((1.338 + 0.0009856 * RC S) – 1)/ (KPOr/ 1000)
2. Cam ellin 19 Method : K = ((1.3319 + 0.00113 * RC S) – 1)/ (KPOr/ 1000)
3. Jarade20 Method : K = ((1.3375 + 0.0014 * RCC) – 1)/ (KPOr/ 1000)
 Special Circumstan ces: Post-Laser Refractive Surg ery Eyes 183

Figure 32-2. Example calculation of the Contact

Lens Method.

Figure 32-3. Rigid PMMA plano

contact lens kit.

THOSE NOT NEEDING CLINICAL H ISTORY

Contact Lens Method9,10
K = BCL + PCL + RCL – RNoCL
The Contact Len s Method was first described by Frederick Rid ley 9 of England (the
inventor of NaOH IOL sterilization) in 1948 and taught by Joseph Sop er 10 in 1974. Th is
method is based on the principle that if a hard PMMA (not rigid gas perm eable) contact
lens (CL) of plano p ower (PCL) and a base cu rve (BCL) equ al to the effective power of the
cornea is placed on the eye it w ill not change the refractive error of the eye. Therefore,
the d ifference between the m an ifest refraction w ith the contact lens (RCL) and w ithout
it (R N oCL) is zero. The formu la above computes the effective corneal p ower if there is a
d ifference in any of these parameters (Fig. 32-2).
Originally it was recom mended to vertex-correct the refractive errors to the corneal
plane. Odenthal et al11 showed that clin ical resu lts were better if they were not cor-
rected. We have decided to u se vertex correction in the Hoffer/ Savin i Tool becau se th is is
more scientifically accu rate. Several IOL power calcu lation computer program s calcu late
th is method and the Clin ical H istory Method autom atically when needed (eg, Hoffer
Program s ® and Hollad ay IOL Con su ltant ®). Plano contact lens sets for perform ing th is
procedu re are com mercially available (Fig. 32-3).
184 Ch apter 32

Figure 32-4. Humphrey

Topography axial map.

Obviou sly, th is method is impossible if the cataract precludes perform ing an accu rate
refraction whereby the visu al acu ity is worse than 20/ 80.

Maloney Central Topography Method21

K = 1.1141 * TKPO-CTR – 5.5
Based on h is analysis of post-LASIK corneal topography (Fig. 32-4) central Ks (TK) on
LASIK eyes, Maloney developed a formu lation to pred ict true corneal power u sing only
the single central postop erative read ing TK.

Koch/Wang Method22
K = 1.1141 * TKPO – 6.1
Koch and Wang analyzed several of these method s and obtained the best resu lts u sing
the Maloney method (d iscu ssed earlier) but on ly after increasing the constant from 5.5 to
6.1. They also offered a second method to calcu late true corneal power if the change in
the patient’s refractive error (RC) is know n. The formu la is:
K = Kt PO – (0.19 × RC)

Savini-Barboni-Zanini Method23
K = 1.114 * Kt PO – 4.98
Th is m ethod on ly requ ires the postop erative Sim-K from topography.
 Special Circumstan ces: Post-Laser Refractive Surg ery Eyes 185

Ta b le 3 2 -1 .

RO SA C O RRECTIO N FACTO R TABLE

22 to <23 1.01
23 to <24 1.05
24 to <25 1.04
25 to <26 1.06
26 to <27 1.09
27 to <28 1.12
28 to <29 1.15
>29 1.22

Shammas No History Method24

K = 1.14 * KPO – 6.8
Sham m as stud ied a series of eyes that had had LASIK. H is analysis led h im to propose
a formu la to pred ict the effective power of the cornea w ithout need ing any of the patient’s
clin ical h istory, on ly the postop erative K read ing obtained w ith m anu al keratom etry.

Adjusted Refractive Index Methods

1. Ferrara M ethod25
K = ((–0.0006 * AL2 + 0.0213 * AL + 1.1572) – 1)/ (KPOr/ 1000)
Th is m ethod requ ires the AL measu rement and the postoperative K read ing in
rad iu s of cu rvatu re.

2. Rosa M ethod26
K = (1.3375 – 1)/ (( KPOr * RCF)/ 1000)
Th is m ethod requ ires the postop erative K read ing in rad iu s of cu rvatu re and the
u se of a table to obtain a factor (RCF) based on AL (Table 32-1). Un fortu nately, they
u sed the SRK II regression formu la in their computation, wh ich I d isagree w ith.

3. Haigis M ethod27
K = -5.1625 * Kr +82.2603 – 0.35
Th is method requ ires on ly the postoperative K read ing form the Zeiss IOLMaster in
rad iu s of cu rvatu re (or converted to d iopters u sing the index of refraction setting in
the IOLMaster).
186 Ch apter 32

Figure 32-5. The O cu-

lus Pentacam provides
a topographic analy-
sis of the corneal front
and back surfaces as
well as central corneal
thickness.

Oculus Pentacam
A comprehensive Eye Scan ner, the Ocu lu s Pentacam (Ocu lu s, Inc, Wetzlar, Gem any,
w w w.ocu lu su sa.com) im ages the anterior segment of the eye by a rotating Scheimpflug
camera m easu rem ent (Fig. 32-5). Th is rotating process supplies pictu res in th ree d imen-
 Special Circumstan ces: Post-Laser Refractive Surg ery Eyes 187

A B

C
Figure 32-6. Pentacam study showing that the
Holladay EKR recommendation of a 4.5 mm zone
is less accurate than using a 3 mm zone compared
to the Clinical History Method.

sion s, provides a topograph ic analysis of the corneal front and back su rfaces as well as
central corneal th ickness, and generates a TrueNetPower m ap of the cornea.
The TrueNetPower m ap of the p ostoperative cornea produced has been proposed as
an accu rate measu re of the true corneal power. In itial resu lts were d isappointing and the
software was recon figu red in early 2007. Fig. 32-6 show s the resu lts reported by Savin i et
al 28-29 that demon strate that the closest to the Clin ical H istory Method is the Equ ivalent K
from the Pentacam 3.0 m m zone, not the 4.5 m m zone as recom mended. A UCLA study 30
has show n the Pentacam read ings as much as 2.00 D off from the back-calcu lated K read-
ing in p ost-LASIK eyes. There are several other stud ies on th is new software that have
also not lived up to exp ectations and newer changes are being proposed.

The BESSt Formula31

Pu blished by Borasio, it u ses the anterior and p osterior corneal cu rvatu res as well as
the central pachymetry from the Pentacam u n it to produce a pred icted central corneal
p ower. The formu la is qu ite complicated, but it is incorp orated into the Hoffer/ Savin i
LASIK Tool. A Version 2.0 of the formu la has ju st been released.
188 Ch apter 32

A B

Figure 32-7. The Ziemer Galilei dual Scheimpflug camera. (A) Instrument, (B) anterior segment
scans.

Figure 32-8. Galilei elevation maps.

Ziemer Galilei
A new in stru ment has been introduced called the Galilei (Figs. 32-7 and 8) that utilizes
two Scheimpflug cam eras w ith a placido d isk to attempt to better evalu ate the anterior
segm ent structu res to produce a true corneal power. Stud ies are qu ite prom ising and
ongoing.
 Special Circumstan ces: Post-Laser Refractive Surg ery Eyes 189

A
Figure 32-9. CSO Sirius Scheimpflug Corneal
Analyzer (A) examiner view, (B) patient view,
(C) summary report.

CSO Sirius
An Italian company (CSO, Costru zione Stru menti Oftalm ici, Florence, Italy) has also
introduced a Scheimpflug camera in stru m ent that has also been show n to be accu rate and
repeatable (Fig. 32-9).
190 Ch apter 32

Me tho ds to Adjus t/ Calcula te the

Targ e t Intra o cular Le ns Po w e r

THOSE NEEDING CLINICAL H ISTORY

Aramberri Double-K Method32
Use KPRE to calcu late ELP and KPO to calcu late IOL power.
One of the most important developm ents to improve the pred iction of corneal p ower in
eyes that have had refractive su rgery was proposed in 2001 and is termed the “Double-K”
m ethod by Jaime Aram berri of San Sebastian, Spain. H is prop osal m akes em inent sen se.
The modern theoretic formu las (except the H aigis) u se the input of corneal power for two
pu rposes: the first is to pred ict the u ltim ate position of the IOL (ACD or ELP) and the
second (along w ith AL, target refraction, and ELP) is to calcu late the power of the IOL.
The formu lations and algorith m s u sed to pred ict the ELP are based on the anatomy of
the anterior segment, wh ich is not changed by corneal refractive su rgery (on ly the center
is flattened and th in ned). Therefore, if the postoperative refractive su rgery K read ing
(wh ich is flatter) is u sed to calcu late the ELP, it w ill produce an erroneou s ELP value.
Becau se the anatomy has not changed, Aram berri recom mend s the u se of the preopera-
tive K read ing to calcu late the ELP. The IOL power is then calcu lated u sing the postopera-
tive K read ing, thu s the u se of two K read ings (“Double-K”). H is analysis of a sm all series
of eyes proved the benefit of th is idea.

Feiz-Mannis Formula33
P = P E – RC S/ 0.7
In th is method you calcu late the em m etropic IOL power u sing the preoperative K read-
ing and adju st that value (P E) u sing the su rgically induced refractive change.

Feiz-Mannis Method34
Th is method utilizes the change in refractive error to offset the calcu lated target IOL
p ower. There is one formu la for myopic eyes and another for hyperopic:
Myopic Eye P = P TARG – 0.595 * RCC + 0.231
Hyperopic Eye P = P TARG – 0.862 * RCC + 0.751

Latkany Methods35
Myopic Eye P = P TARG FlatK – 0.47 * RPRE + 0.85
Hyperopic Eye P = P TARG FlatK + 0.27 * RPRE + 1.53
This method requ ires knowledge of the pre-LASIK refractive error and the calcu lation of
the target IOL power u sing the flattest postoperative K rather than the u sual average K.

Masket Method36
P = P TARG – 0.323 * RCC + 0.138
[SRK/ T: myopes; Hoffer Q: hyperopes]
 Special Circumstan ces: Post-Laser Refractive Surg ery Eyes 191

Ta b le 3 2 -2 .

EXAMPLE C ALCULATIO NS USING THE MASKET FO RMULA

Myopic Eye Hyperopic Eye
SRK/T calculates 16.0 D IO L Hoffer Q calculates 22.0 D
Change in Rx = -6.0 D Change in Rx = +3.0 D
-0.323 * (-6) + 0.138 = +2.076 -0.323 * (+3) + 0.138 = -0.82
P = 16.0 + 2.0 = 18.0 D P = 22.0 – 1.0 = 21.0 D

Th is m ethod is a play on the Latkany m ethod, wh ich adju sts the power of the IOL
calcu lated u sing the postoperative m easu red d ata and the knowledge of the su rgically
induced refractive change. He recom mend s u sing the SRK/ T formu la for myopic ALs and
the Hoffer Q for hyp eropic ALs. Example calcu lations are show n in Table 32-2.
In a series of 28 p ost-LASIK eyes, he rep orted 43% of the eyes obtain ing a p ostop erative
refractive error of plano, 95% w ith in ± 0.50 D of pred iction, and a total error range from
-0.75 D to +0.50 D.

Wake Forest Method37

Use RPRE as the RXTARG u sing measu red AL and KPRE
In 2005, Gagnon et al (from Wake Forest Un iversity) published an alternative calcu la-
tion m ethod that has been d iscu ssed by others over the past 20 years. Th is method sim-
ply u ses the patient’s preoperative refraction before LASIK as the target or “desired” PO
refraction in the calcu lation and the m easu red AL and K read ings w ithout mod ification.

THOSE NOT NEEDING CLINICAL H ISTORY

Aramberri Double-K Method32
Use 43.5 or 44.00 to calc ELP & KPO to calc IOL p ower.
The u se of a stand ard norm al K read ing in the Double-K method is a great improve-
m ent over u sing the calcu lated very flat K read ing.

Ianchulev Intraoperative Aphakic Refraction Method38

P = 2.02 * AR + (A – 118.4)
In 2003, Ianchu lev et al proposed calcu lating IOL p ower by p erform ing an aphakic
refraction on the operating room table u sing a hand-held autom ated refractor im m ed i-
ately after the cataract has been removed and the AC is in flated to norm al statu s. The
resu ltant refraction is mod ified by the formu la.
H is early resu lts are qu ite prom ising (see Chapter 44). Th is m ethod wou ld completely
elim inate the need for axial length, corneal power measu rem ents, and the problem s w ith
LASIK and silicone oil-filled eyes. However, it wou ld requ ire a large IOL inventory in the
operating room .
192 Ch apter 32

Mackool Secondary Implant Method39

P = 1.75 * AR + (A – 118.84)
Th is method is sim ilar to the previou s method, except the patient is removed from the
operating room w ithout an IOL implanted, then refracted in a refraction lane and taken
back to the operating room for second ary lens implantation. It is my impression that th is
m ethod wou ld not be popu lar w ith most su rgeon s.

Fo rm ula Le g e nds
A = the IOL A constant for plan ned IOL style
AL = axial length
AR = aphakic refractive error (SE)
B = base cu rve, PCL = power of CL, NoCL = bare refraction
CL = contact len s
K = pred icted PO corneal power
KPO = the average PO corneal power by m anu al keratom etry (in d iopters D)
KPOFLAT = flattest measu red PO m anu al keratom etry
KPOr = the average PO corneal p ower by IOLMaster (in rad iu s r [m m])
KPRE = refractive su rgery preoperative corneal power (K read ings)
P = IOL Power
P EMM = the IOL power calcu lated for em metropia
P FlatK = IOL power calcu lated for RxTARG u sing the PO flattest K
P TARG = the target IOL power to produce the PO desired refractive error
R = refractive error: PRE = preop erative, PO = postop erative
RCC = su rgical change in refractive error (SE) vertexed to corneal plane
RCF = Rosa Correction Factor based on axial length
RC S = su rgical change in refractive error (SE) at sp ectacle plane
RPO = refractive su rgery PO refractive error (spherical equ ivalent)
RPRE = refractive su rgery preoperative refractive error (spherical equ ivalent)
RxTARG = plan ned postoperative target refractive error
TK = average PO top ography central Sim-K or EffRP
TKCTR = exact singu lar PO topography central K

Re fe re nce s
1. Koch DD, Liu JF, Hyd e LL, Rock RL, Em ery JM. Refractive complication s of cataract su r-
gery after rad ial keratotomy. A m J Ophthalmol. 1989;108(6):676–682.
2. Hoffer KJ. The Hoffer Q formu la: A comparison of theoretic and regression formu las.
J Cataract Refract Surg. 1993;19(6):700–712. Errata: 1994;20(6):677 and 2007;33(1):2–3.
3. Hollad ay JT, Prager TC, Chand ler TY, et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–24.
4. Retzlaff J, Sand ers DR, Kraff MC. Developm ent of the SRK/ T intraocu lar len s implant p ower
calcu lation formu la. J Cataract Refract Surg. 1990;16(3):333–340. Erratu m : 1990;16(4):528.
5. H aigis W. The H aigis formu la. In: H J Sham m as, ed. Intraocular Lens Power Calculations.
Thorofare, N J: SLACK Incorp orated ; 2003:41–57.
6. Hollad ay JT. IOL calcu lation s follow ing rad ial keratotomy su rgery. Refract Corneal Surg.
1989;5:36A.
 Special Circumstan ces: Post-Laser Refractive Surg ery Eyes 193

7. Hoffer KJ. Intraocu lar len s p ower calcu lation for eyes after refractive keratotomy. J Refract
Surg. 1995;11(6):490–493.
8. Hoffer KJ. Calcu lating intraocu lar len s power after refractive su rgery. A rch Ophthalmol.
2002;120(4):500–501.
9. Rid ley F. Developm ent in contact len s theory. Trans Ophthalmol Soc UK. 1948;68:385–401.
10. Sop er JW, Goffm an J. Contact lens fitting by retinoscopy. In : Sop er JW, ed. Contact Lenses.
New York, N Y: Stratton Intercontinental Med ical Book Corp.; 1974;99.
11. Od enthal MTP, Eggin k CA, Melles G, et al. Clin ical and theoretical resu lts of intraocu lar
len s p ower calcu lation for cataract after photorefractive keratectomy for myopia. A rch
Ophthalmol. 2002;120(4):431–438.
12. H am ed AM, Wang L, Misra M, Koch D. A comparative analysis of five m ethod s of
d eterm in ing corneal refractive p ower in eyes that have u nd ergone myopic laser in situ
keratom ileu sis. Ophthalmology. 2002;109(4):651–658.
13. Sp eicher L. Intra-ocu lar len s calcu lation statu s after corneal refractive su rgery. Curr Opin
Ophthalmol. 2001;12(1):17–29.
14. Seitz B, Langenbucher A, Nguyen N X, Ku s MM, Kuch le M. Underestim ation of intraocu lar
len s p ower for cataract su rgery after myopic photorefractive keratectomy. Ophthalmology.
1999;106(4):693–702.
15. Seitz B, Langenbu cher A. Intraocu lar len s p ower calcu lation in eyes after corneal refrac-
tive su rgery. J Refract Surg. 2000;16(3):349–361.
16. Jarade EF, Abi Nader FC, Tabbara KF. Intraocu lar len s power calcu lation follow ing LASIK:
Determ ination of the new effective ind ex of refraction. J Refract Surg. 2006;22(1):75–80.
17. Ronje. LASIK IOL calcu lation. Eyenet M agazine. 2004;20:23–24.
18. Savin i G, Barbon i P, Zan in i M. Intraocu lar len s p ower calcu lation after myopic refractive
su rgery: Theoretical comparison of d ifferent m ethod s. Ophthalmology. 2006;113(8):1271–
1282.
19. Cam ellin M, Calossi A. A new formu la for intraocu lar len s p ower calcu lation after refrac-
tive corneal su rgery. J Refract Surg. 2006;22(2):187–199.
20. Jarad e EF, Tabbara KF. New formu la for calcu lating intraocu lar len s p ower after laser in
situ keratom ileu sis. J Cataract Refract Surg. 2004;30(8):1711–1715.
21. Sm ith RJ, Chan WK, Maloney RK. The pred iction of su rgically induced refractive change
from corneal topography. A m J Ophthalmol. 1998;125(1):44–53.
22. Koch D, Wang I. Calcu lating IOL power in eyes that have had refractive su rgery. J Cataract
Refract Surg. 2003;29(11):2039–2042.
23. Savin i G, Barbon i P, Zan in i M. Correlation between attempted correction and kerato-
m etric refractive ind ex of the cornea after myopic excim er laser su rgery. J Refract Surg.
2007;23(5):461–466.
24. Sham m as H J, Sham m as MC, Garabet A, Kim JH, Sham m as A, LaBree L. Correcting the
corneal p ower m easu rem ents for intraocu lar len s p ower calcu lation s after myopic laser
in situ keratom ileu sis. A m J Ophthalmol. 2003;136(3):426–432.
25. Ferrara G, Cen namo G, Marotta G, Loffredo E. New formu la to calcu late corneal power
after refractive su rgery. J Refract Surg. 2004;20(5):465–471.
26. Rosa N, Capasso L, Lan za M, Iaccarino G, Rom ano A. Reliability of a new correcting fac-
tor in calcu lating intraocu lar len s p ower after refractive corneal su rgery. J Cataract Refract
Surg. 2005;31:1020-1024.
27. H aigis W. IOL calcu lation after refractive su rgery for myopia: The H aigis-L formu la.
J Cataract Refract Surg. 2008;34(10):1658–1663.
28. Savin i G, Barbon i P, Carbonelli M, Hoffer KJ. Agreem ent between Pentacam and video-
keratography in corneal power assessm ent. J Refract Surg. 2009;25:534-538.
29. Savin i G, Barbon i P, Carbonelli M, Hoffer KJ. Accu racy of Scheimpflug corneal p ower m ea-
su rem ents for intraocu lar len s p ower calcu lation. J Cataract Refract Surg. 2009;35(7):1193-
1197.
194 Ch apter 32

30. Tang Q, Hoffer KJ, Olsen MD, Miller KM. Accu racy of Scheimpflug Hollad ay equ iva-
lent keratom etry read ings after corneal refractive su rgery. J Cataract Refract Surg. 2009;
35(7):1198-1203.
31. Borasio E, Steven s J, Sm ith GT. Estim ation of true corneal p ower after keratorefrac-
tive su rgery in eyes requ iring cataract su rgery: BESSt formu la. J Cataract Refract Surg.
2006;32(12):2004–2014.
32. Aram berri J. Intraocu lar len s p ower calcu lation after corneal refractive su rgery: dou ble-K
m ethod . J Cataract Refract Surg. 2003;29(11):2063–2068.
33. Feiz V, Man n is MJ, Garcia-Ferrer F. Intraocu lar len s p ower calcu lation after laser in
situ keratom ileu sis for myopia and hyp eropia: A stand ard ized ap proach. Cornea.
2001;20(8):792–797.
34. Feiz V, Mosh irfar M, Man n is MJ, et al. Nomogram -based intraocu lar len s power
ad ju stm ent after myopic photorefractive keratectomy and LASIK: A new approach.
Ophthalmology. 2005;112(8):1381–1387.
35. Latkany RA, Choksh i AR, Sp eaker MG, Abram son J, Soloway BD, Yu G. Intraocu lar len s
calcu lation s after refractive su rgery. J Cataract Refract Surg. 2005;31(3):562–570.
36. Masket S, Masket SE. Simple regression formu la for intraocu lar len s power ad ju stm ent in
eyes requ iring cataract su rgery after excim er laser photoablation. J Cataract Refract Surg.
2006;32(3):430–434.
37. Walter KA, Gagnon MR, Hoop es PC Jr., Dickenson PJ. Accu rate intraocu lar lens power
calcu lation after myopic laser in situ keratom ileu sis, bypassing corneal p ower. J Cataract
Refract Surg. 2006;32(3):425–429.
38. Ianchu lev T, Salz J, Hoffer K, et al. Intraop erative optical intraocu lar lens power estim a-
tion w ithout axial length m easu rem ents. J Cataract Refract Surg. 2005;31(8):1530–1536.
39. Mackool RJ, Ko W, Mackool R. Intraocu lar len s p ower calcu lation after laser in situ ker-
atom ileu sis: The aphakic refraction tech n ique. J Cataract Refract Surg. 2006;32(3):435–437.
 33

Sp e cia l Circu msta n ce s:

Ha ig is -L IOL Fo rmu la
Wolfgang Haigis, M S, PhD

T
he nu m ber of patients presenting w ith cataract after refractive corneal su rgery has
been continuou sly increasing over the years. These patients still present a chal-
lenge to IOL calcu lation, although the special problem s associated w ith these eyes
are well u nderstood tod ay.

Pro ble m s fo r IOL Calculatio n

Essentially there are 3 sou rces of errors in IOL calcu lation after refractive su rgery.
First, there is the rad iu s m easu rem ent error stem m ing from the fact that K read ings are
not taken at the optical axis, but a little p eripherally. Th is error is relevant in cases of
preced ing laser su rgery for myopia, not for hyp eropia. Second, the keratom eter ind ex
error is du e to the fact that the corneal cu rvatu re ratio is d eliberately altered by refrac-
tive su rgery, thu s lead ing to m ean ingless K valu es. Th ird, som e IOL p ower formu las
m aking u se of K valu es to pred ict the effective len s p osition d erive a w rong valu e since
the cu rrent K does not represent the eye’s geom etry anym ore, as it does in its u ntou ched
state. Th is IOL formu la error cau ses a hyp eropic refractive sh ift in patients after laser
su rgery for myopia.

195
Hoffer KJ. IOL Power (pp. 195 -198).
© 2011 SLACK Incorporated.
196 Ch apter 33

Appro a che s in the Lite ra ture

A variety of approaches to hand le eyes after previou s refractive su rgery can be fou nd
in the literatu re (see Chapter 32). Formu las to estim ate the effective corneal power, as
well as formu las to fudge the calcu lated IOL p ower, are available. Often h istorical d ata,
add itional measu rem ents, and/ or special measu rement parameters are requ ired. There is
no m agic general formu la valid for all cases; u su ally solution s are described for specific
m easu rement in stru m ents and/ or specific IOL power formu las. IOL calcu lation method s
for eyes after refractive su rgery d iffer most in whether they requ ire h istorical patient d ata
or whether they rely on ly on cu rrent m easu rements. Obviou sly, no-h istory m ethod s are
clin ically the most u sefu l ones. Among them are the No-H istory m ethod of Sham m as, the
Pentacam-based BESSt formu la of Borasio, and the H aigis-L formu la 1 for the IOLMaster
(see Chapter 30). A spread sheet (Hoffer/ Savin i Tool) program m ed w ith virtually all
algorith m s h itherto pu blished is available for dow n load from w w w.EyeLab.com at no
cost. Also, on the ASCRS website (w w w.ascrs.org) an on line calcu lator is implemented
offering the free u se of a lim ited variety of published calcu lation schemes, but there are
lim itations as to the formu las u sed.

The Haig is -L Fo rm ula

The H aigis-L formu la con sists of the regu lar H aigis formu la 2 and a sp ecial correction
for the corneal rad iu s (r m eas) measu red by the Zeiss IOLMaster:
331.5
r corr = ______________________________
-5.1625 * r m eas + 82.2603 – 0.35
where r = IOLMaster rad iu s of cu rvatu re of the cornea, corr = corrected and m eas =
m easu red.
It is important to rem em ber th is correction is sp ecific on ly for the Zeiss IOLMaster.

Clinical Re s ults With the Haig is -L Fo rm ula

At present, we are studying the clin ical resu lts w ith the H aigis-L formu la sp ecific for
the IOLMaster; for 222 eyes after IOL implantation w ith previou s myopic and 56 w ith pre-
viou s hyperopic laser vision correction. The previou sly myopic eyes received 35 d ifferent
IOL types by 64 d ifferent su rgeons from all over the world ; form er hyperopic eyes were
implanted w ith 13 d ifferent IOL types by 15 d ifferent su rgeon s. All patients had biom-
etry and keratometry u sing the Zeiss IOLMaster. IOL calcu lation was perform ed from
cu rrent measu rem ents u sing the H aigis-L formu la (wh ich is included in the IOLMaster
software).
The mean arith m etic pred iction errors (ME) were -0.08 ± 0.71 D for myopic and -0.06
± 0.77 D for hyperopic eyes. The respective m ed ian absolute errors (MAE) were 0.37 D and
0.40 D. Of the myopic eyes, 98.6% were correctly pred icted w ith in ±2 D, 82.9% w ith in ±1
D, and 59.9% w ith in ± 0.5 D. The resp ective percentages for eyes after su rgery for hypero-
pia were 96.4%, 82.1%, and 58.9%. These resu lts compare well w ith norm al eyes, although
the error ranges in pred icted refraction are a little h igher in eyes after refractive su rgery.
 Special Circumstan ces: Haig is-L IOL Formula 197

Figure 33-1. Percentages

of correct refraction pre-
dictions obtained with the
standard Haigis (HAIstd)
and Haigis-L formulas
(HAIL), as well as the stan-
dard SRK/T (SRKTstd) and
the Double-K corrected
(SRKT2k) SRK/T formulas
for 60 eyes with an Alcon
SN60WF with previous
laser surgery for myopia
(unpublished data).

SRK/ T3 is the IOL power formu la most affected by the formu la error. The formu la
error can be lessened by applying the Aram berri4 Dou ble-K correction. The Hoffer Q
formu la is least effected, but the H aigis formu la, on the other hand, does not su ffer from
th is error becau se it does not u se the K read ing as a pred ictor for the effective lens posi-
tion (ELP).
Fig. 33-1 show s the perform ances of the H aigis and SRK/ T formu las expressed in cor-
rect refraction pred ictions in their stand ard version s as well as in their ad apted version s
for preced ing refractive su rgery (H aigis-L and SRK/ T w ith Dou ble-K correction) for 60
eyes w ith an Alcon SN60WF. The H aigis-L formu la, as can be seen, compares very well
w ith the other calcu lation approaches.

Re fe re nce s
1. H aigis W. IOL calcu lation after refractive su rgery for myopia: The H aigis-L formu la. J
Cataract Refract Surg. 2008;34(10):1658-1663.
2. H aigis W, Lege B, Miller N, Sch neider B. Comparison of im m ersion u ltrasou nd biom etry
and partial coherence interferom etry for intraocu lar len s calcu lation accord ing to H aigis.
Graefe’s A rch Clin Exp Ophthalmol. 2000;238:765-773.
3. Retzlaff J, Sand ers DR, Kraff MC. Developm ent of the SRK/ T intraocu lar len s implant
p ower calcu lation formu la. J Cataract Refract Surg. 1990;16(3):333-340.
4. Aram berri J. Intraocu lar len s p ower calcu lation after corneal refractive su rgery: Dou ble K
m ethod . J Cataract Refract Surg. 2003;29(11): 2063-2068.
 34

Sp e cia l Circu msta n ce s:

Do u ble -K Me t h o d
Jaime Aramberri, M D

A
ll theoretical IOL power calcu lation formu las perform two consecutive calcu lations:
first the position of the IOL w ithin the eye (ELP: effective lens position in thin lens
term inology 1) is estim ated from d ifferent independent variables; and then the
power of the implant is calcu lated u sing optical vergence or ray tracing formu lation.
Since the introduction of 3rd generation formu las in 1988,2 corneal cu rvatu re has been
u sed as a powerfu l ELP pred icting variable in most of them ; the steep er the cornea the
h igher the ELP, and therefore the h igher the IOL p ower. However th is anatom ical correla-
tion fails in abnorm ally steep or flat corneas, the latter due to corneal refractive su rgery
eyes. The Dou ble-K Method avoid s the ELP pred iction error in these eyes.
It mu st also be remem bered that there is another sou rce of error after corneal refractive
su rgery wh ich is the K measu rement error. Topographers and keratometers overestim ate
K value after myopic su rgery and u nderestim ate K value after hyperopic su rgery due to
the altered corneal anterior/ posterior ratio. Th is error and its correction w ill be explained
elsewhere.

Co rne al Re fra ctive Surg e ry Anato m ical Chang e s

Laser corneal refractive su rgery (LASIK/ PRK) flattens the anterior corneal su rface but
does not change the posterior su rface accord ing to Scheimpflug measu rements or steepen
it accord ing to Orbscan m easu rements. The latter has been reported to be an artefact due
to error in bou nd aries recogn ition or m agn ification.3 Anterior cham ber depth decreases
a sm all and non sign ificant amou nt.4
199
Hoffer KJ. IOL Power (pp. 199 -206).
© 2011 SLACK Incorporated.
200 Ch apter 34

Ta b le 3 4 -1 .

ELP PREDICTING VARIABLES USED BY THEO RETIC FO RMULAS

Formula K AL ACD LT CD Rx Age
Binkhorst 2 No Yes No No No No No
SRK/T Yes Yes No No No No No
Hoffer Q Yes Yes No No No No No
Holladay 1 Yes Yes No No No No No
Holladay 2 Yes Yes Yes Yes Yes Yes Yes
Haigis No Yes Yes No No No No
Olsen Yes Yes Yes Yes No Yes No

K = corneal power, AL = axial length, ACD = preoperative anterior chamber depth (ante-
rior cornea to anterior lens), LT = lens thickness, Rx = Refraction, and CD = corneal
diameter (horizontal white-to-white distance.)

ELP Pre dic tio n

Th ird generation formu las u se 2 independ ent variables to pred ict ELP (Table 34-1).
Hoffer Q,2 Hollad ay 1, 5 and SRK/ T6 u se AL and K. H aigis7 u ses AL and preop erative
ACD (anterior corneal vertex to anterior len s). Fou rth generation formu las u se more than
2 variables for the sam e task. The Hollad ay 28 u ses 7: AL, K, ACD, len s th ickness (LT), cor-
neal d iam eter (CD), refraction (Rx), and age. Olsen 9 (2006) u ses 5 variables: AL, K, ACD,
LT, and Rx. The m ain d ifference among all formu las is the pred iction of ELP. In fact, if
ELP is fixed to test the optical formu la perform ance it can be seen that Hoffer Q, Hollad ay
1, and SRK/ T pred ict IOL power w ith in 0.5 D from each other. H aigis always calcu lates
0.50 to 1.00 D h igher power (Fig. 34-1). Th is d ifference is system atic so it can have an effect
on the a 0, a 1, and a 2 values u sed for the calcu lations. Knowledge of ELP pred iction p erfor-
m ance of each formu la helps in u nderstand ing the potential sou rces of error.
SRK/ T lim its ELP pred iction neither superior nor in feriorly (Fig. 34-2). H igh K and AL
values w ill produce illogically h igh ELP values, ie, ELP for K = 45 and AL = 30 is 7.65 m m
wh ich of cou rse is sen seless; no IOL sits that deeply. Th is w ill increase IOL power in a
m agn itude dep end ant on the p ower of the IOL itself. A cu riou s phenom enon is a sudden
decrease of ELP for K values over 46. The h igher the K is (over 46 D) the lower the AL
needed to produce th is pred iction change. The consequence is that SRK/ T pred icts ELP
more accu rately in long eyes when the K is very h igh (>47) than when the K is in the m id-
h igh range (45 to 47). For very low K values, SRK/ T pred icts the lowest ELP value of all.
H aigis pred icts ELP linearly as a fu nction of AL for a fixed ACD despite variable K
values (wh ich is not a pred icting variable as has been stated). Th is fu nction is not lim ited
either sup erior or inferiorly (Fig. 34-3).
 Special Circumstan ces: Double-K Meth od 201

Figure 34-1. IO L power prediction of 3rd generation formulas for a

fixed ELP value of 5.50 mm. Difference among Hoffer Q , Holladay 1,
and SRK/T is lower than 0.5 D. Haigis formula calculates a higher value
(0.50 to 1.00 D) throughout the range of AL.

Figure 34-2. SRK/T ELP prediction. When K >46, an abrupt fall of ELP occurs.
High K and AL values lead to an overestimation of ELP. ELP values above 7 mm
are seldom found in pseudophakic eyes.
202 Ch apter 34

Figure 34-3. Haigis ELP prediction is a linear function of AL for a fixed ACD.
The K is not a predicting variable. There are no upper or lower limits in this
function.

Figure 34-4. Hoffer Q ELP prediction has an upper limit of 6.5 mm.

Hoffer Q (Fig. 34-4) and Hollad ay 1 (Fig. 34-5) algorith m s lim it superior ELP pred iction
to avoid ELP overestim ation in long eyes and/ or steep corneas. The Hollad ay 1 lim its ELP
over 26 m m of AL in a value that dep end s on K. The Hoffer Q sets an absolute ELP lim it
of 6.50 m m. With very low K values, SRK/ T pred icts the shortest ELP value and Hoffer Q
the h ighest (Fig. 34-6) w ith sign ificant d ifferences.
 Special Circumstan ces: Double-K Meth od 203

Figure 34-5. Holladay 1 ELP prediction is limited beyond AX = 26 mm in a

value that depends on K.

Figure 34-6. ELP as a function of K. AL = 26 and ACD = 3.3. Hoffer Q predicts

the highest values for very flat corneas. This is why Single-K Hoffer Q induces
less hyperopia than Single-K Holladay 1 and SRK/T after myopic corneal refrac-
tive surgery. Haigis is not affected by K as it is not a predicting variable

ELP Pre dic tio n Afte r Co rne al Re fra c tive Surg e ry

As the relationsh ip between corneal cu rvatu re and pseudophakic ACD has changed
after corneal refractive su rgery, it is obviou s that u sing the post-su rgical K value to pre-
d ict that variable w ill lead to an error in any algorith m that u ses K in ELP pred iction.
After myopic su rgery, the new flat K value w ill induce an u nderestim ation of ELP and
204 Ch apter 34

Ta b le 3 4 -2 .

RESULTS O F ELP ERRO R O F HO FFER Q, HO LLADAY 1, AND SRK/T*

*Assuming that ELP estimation of the unchanged cornea (First row: K = 44, Refractive cor-
rection [Rx] = 0) is correct. The highest error is induced by the SRK/T algorithm (1.72),
whereas the Hoffer Q introduces the least error (0.44); with Holladay 1 in between (1.27).

AL Kpre Rx Hoffer Q Holladay 1 SRK/T

mm D D ELP mm ELP error ELP mm ELP error ELP mm ELP error
26 44 0 5.88 0.00 6.04 0.00 5.99 0.00
26 43 1 5.82 0.06 5.85 0.18 5.72 0.28
26 42 2 5.76 0.12 5.69 0.35 5.47 0.52
26 41 3 5.71 0.18 5.53 0.51 5.26 0.73
26 40 4 5.65 0.23 5.38 0.65 5.06 0.93
26 39 5 5.61 0.28 5.25 0.79 4.88 1.11
26 38 6 5.56 0.32 5.12 0.92 4.71 1.28
26 37 7 5.52 0.36 5.00 1.04 4.55 1.44
26 36 8 5.48 0.40 4.88 1.16 4.41 1.58
26 35 9 5.44 0.44 4.77 1.27 4.27 1.72

therefore an u nderestim ation of IOL power resu lting in a hyperopic refraction. After
hyp eropic su rgery, the new steep K value w ill induce an overestim ation of ELP, an over-
estim ation of IOL power yield ing a myopic refractive error.
Th is error dep end s on the refractive correction perform ed on the cornea, becau se the
h igher th is is the h igher the d ifference between the original and the post su rgical K value
is and consequently the induced ELP error w ill increase.
Another factor is wh ich formu la is being u sed becau se ELP pred iction algorith m s are
d ifferent, as stated above. Table 34-2 show s that SRK/ T induces the biggest error and
Hoffer Q the lowest (w ith a very sign ificant d ifference). The reason is that the slopes of
these fu nction s are very d ifferent; SRK/ T decreases ELP ~0.2 m m per D of K, Hollad ay 1
~0.15 m m per D of K, and the Hoffer Q on ly ~0.05 m m per D of K. Th is m ean s, in the case
of a 7 D p ost-LASIK calcu lation w ith AL = 26 m m, SRK/ T w ill u nderestim ate ELP by 1.44
m m wh ich w ill tran slate into 1.4 D of hyperopia, the Hollad ay 1 w ill u nderestim ate ELP
by 1.04 m m resu lting in 1 D of hyperopia, and the Hoffer Q w ill u nderestim ate ELP by
0.36 m m producing on ly 0.30 D of hyperopia (approxim ate nu m bers).
The m agn itude of IOL power and sp ectacle plane refractive error produced by any ELP
error w ill depend m ain ly on AL (Table 34-3) and to a lesser extent on K value.
 Special Circumstan ces: Double-K Meth od 205

Ta b le 3 4 -3 .

IO L ERRO R AND SPECTACLE PLANE REFRACTIVE ERRO R

PRO DUCED BY A 0.50 MM ERRO R IN ELP PREDICTIO N
The refractive translation of this error is clearly dependant on AL. Calculations were per-
formed by paraxial ray tracing.

AL (mm) IOL Error (D) Rx Error* (D)

21 1.76 1.23
23 1.10 0.80
27 0.70 0.50
30 0.40 0.20

*Spectacle plane refractive error. Calculations performed by paraxial ray tracing.

ELP Pre dic tio n Erro r Co rre ctio n

The easiest way to correct th is problem is not u sing the K as an ELP pred icting vari-
able. The on ly formu la program med in th is way is the H aigis formu la, wh ich estim ates
ELP u sing on ly AL and ACD.
If a formu la that u ses corneal power (K or r) for ELP pred iction is u sed, the formu la
mu st be program m ed in such a way that the K previou s to corneal refractive su rgery
(Kpre ) is input into the ELP pred icting algorith m and the K after corneal refractive su r-
gery (Kp ost) is input into the optical power formu la. Th is Kp ost is not the value measu red
by the keratom eter or topographer but a corrected K value that can be calcu lated u sing
d ifferent method s as is explained elsewhere in th is book (see Chapter 30). The u se of two
d ifferent K values in IOL p ower calcu lation has been called the Dou ble-K Method.10
In the original pu blished paper,10 the Double-K SRK/ T has been recom mended to
be the more accu rate in long eyes. But since then I have d iscovered (nonpublished d ata)
that th is formu la overestim ates ELP, esp ecially if Kpre is between 44 and 46 D—inducing
some myopia. Th is behavior was not so obviou s w ith nonop erated eyes, as the IOL power
pred iction d ifference is low w ith low-powered IOLs (needed in myop es). Better 3rd gen-
eration options are Dou ble-K Hoffer Q and Dou ble-K Hollad ay 1.

Ho w to Obtain Do uble -K Fo rm ula Calcula tio ns

All formu las (except Hollad ay 2) have been pu blished in the scientific p eer-reviewed
literatu re. It is advisable to get the articles and program them in a spread sheet in such
a way that the ELP-pred icting algorith m u ses Kpre as the independent variable and the
optical vergence formu la u ses Kp ost as the independent variable. It is important to be
aware of the crucial errata in the Hoffer Q and SRK/ T pu blications. Conversion tables
have been published 11 to translate Single-K calcu lation s to Dou ble-K ones.
206 Ch apter 34

Some u ltrason ic biometers have included Dou ble-K formu las in their software, such
as Axis II (Quantel) and Sonomed A-Scan s. Com m ercial IOL power calcu lation software
includ e Dou ble-K formu las: Hoffer Program s® (w w w.eyelab.com) allow s u sing th is
m ethod w ith all 3rd generation formu las. The Hollad ay IOL Consu ltant® program (w w w.
dochollad ay.com) u ses this method on ly w ith the Hollad ay 2 formu la (u npublished
d ata).

Re fe re nce s
1. Hollad ay JT. Stand ard izing con stants for u ltrason ic biom etry, keratom etry, and intraocu-
lar len s calcu lation. J Cataract Refract Surg. 1997;23(9):1356–1370.
2. Hollad ay JT, Prager TC, Chand ler TY, et al. A th ree-part system for refin ing intraocu lar
len s p ower calcu lation s. J Cataract Refract Surg. 1988;14(1):17–24.
3. Nawa Y, Masud a K, Ued a T, H ara Y, Uozato H. Evalu ation of ap parent ectasia of the
p osterior su rface of the cornea after keratorefractive su rgery. J Cataract Refract Surg.
2005;31(3):571–573.
4. H ashem i H, Meh ravaran S. Corneal changes after laser refractive su rgery for myopia:
Comparison of Orbscan II and Pentacam find ings. J Cataract Refract Surg. 2007;33(5):841–
847.
5. Hoffer KJ. The Hoffer Q formu la: A comparison of theoretic and regression formu las
[pu blished correction app ears in : J Cataract Refract Surg. 1994;20(6):677 and J Cataract
Refract Surg. 2007;33(1):2–3]. J Cataract Refract Surg. 1993;19(6):700–712.
6. Retzlaff J, Sanders DR, Kraff MC. Developm ent of the SRK/ T intraocu lar lens implant
p ower calcu lation formu la [pu blished correction app ears in : J Cataract Refract Surg.
1990;16(4):528]. J Cataract Refract Surg. 1990;16(3):333–340.
7. H aigis W. IOL calcu lation accord ing to H aigis. Available on line: htpp :/ / w w w.augen-
klin ik.u n i-w uerzbu rg.de/ u slab/ ioltxt/ haie.htm . Last revision: Decem ber 7, 1998.
8. Hoffer KJ. Clin ical resu lts u sing the Hollad ay 2 intraocu lar len s p ower formu la. J Cataract
Refract Surg. 2000;26(8):1233–1237.
9. Olsen T. Pred iction of the effective p ostop erative (intraocu lar len s) anterior cham ber
d epth. J Cataract Refract Surg. 2006;32(3):419–424.
10. Aram berri J. Intraocu lar len s power calcu lation after corneal refractive su rgery: Dou ble-K
m ethod . J Cataract Refract. Surg. 2003;29(11):2063–2068.
11. Koch DD, Wang L. Calcu lating IOL p ower in eyes that have had refractive su rgery.
J Cataract Refract Surg. 2003;29(11):2039–2042.
 35
Sp e cia l Circu msta n ce s:
In flu e n ce o f Sp h e rica l
Ab e rra tio n o n IOL Po w e r
Sverker Norrby, PhD

A
s the nam e suggests, spherical aberration is a property of spherical len ses. For a
p ostive spherical lens, peripheral rays are refracted more than rays close to the
axis (positive spherical aberration). How much dep end s on the cu rvatu re of the
refracting su rfaces and the shap e of the lens.
In Table 35-1, d imensions (m m) are show n for d ifferent shap es of 20 D IOLs m ade of a
m aterial w ith a refractive index (RI) = 1.460, when im m ersed in aqueou s (RI = 1.336). Also
included are the parameters for the Gu llstrand cornea. Note that rad ii convex toward s the
object are positive, and those concave are negative. Plano su rfaces are given a very large
rad iu s.
The values for the equ i-convex len s are input into the calcu lation spread sheet
(Figs. 35-1 and 2), wh ich can be dow n loaded from AAO at http :/ / one.aao.org/ lm s/
cou rses/ IOLPowerFormu las/ im ages/ LO14_GC01.xls and http :/ / one.aao.org/ lm s/ cou rs-
es/ IOLPowerFormu las/ im ages/ LO14_GC02.xls. The m arginal ray is traced at the edge of
the ap ertu re. The focu sing ray is traced at 1/ √2 of the ap ertu re. Th is rad iu s d ivides the
pupil into an in ner circle and an outer an nu lu s of equ al area. The focu sing ray can thu s
be considered an average ray and its focal point can be considered as “best focu s” in that
sen se. The tracing of these two rays is exact. The paraxial ray is traced at an incom ing
height of 1 m m for visibility. Paraxial ray tracing is an approxim ation and the focal p oint
is independent of the height chosen. Th is can be demonstrated by giving it other values
in the spread sheet.
The d istances are termed as follow s:
l From back vertex of lens to best-focu s: back focal length (BFL)

207
Hoffer KJ. IOL Power (pp. 207-210).
© 2011 SLACK Incorporated.
208 Ch apter 35

Ta b le 3 5 -1 .

D IMENSIO NS (MM) O F D IFFERENT SHAPES O F 20 D IO LS

Lens shape Equi-convex Convex-plano Plano-convex Meniscus Cornea
Front radius 12.336 6.200 1000000 3.876 7.70
Back radius -12.336 -1000000 -6.200 10.000 6.80
Thickness 1.50 1.50 1.50 1.50 0.50
Diameter 6.00 6.00 6.00 6.00 12.00
O bject RI 1.336 1.336 1.336 1.336 1
Image RI 1.336 1.336 1.336 1.336 1.336
Lens RI 1.460 1.460 1.460 1.460 1.376
where RI = refractive index.

Figure 35-1. Graphic Calculator Ray Tracing Spreadsheet: Paraxial thick lens ray
tracing of single lenses. The spreadsheet is interactive and will accept any input,
though the graph may not remain within the limits for excessive numbers.

l From principal plane to best-focu s: effective focal length (EFL)

l From principal plane to paraxial focu s: paraxial focal length (PFL)
IOLs shou ld, in accord ance w ith the international stand ard (ISO 11979-2), be labeled
w ith their paraxial p ower, wh ich is 1336/ PFL w ith PFL given in m m. The d istance
between m arginal focu s and paraxial focu s is the longitud inal spherical aberration (LSA),
wh ich is output in the spread sheet.
Input the values for the other lens shap es and watch the con sequences. Also try the
IOLs in air (object and im age RI = 1) and see how much h igher the power is in air. Also try
the cornea. Note in particu lar how the principal plane sh ifts w ith shape. For equ i-convex
lenses it is slightly posterior to the m idd le, for convex-plano len ses it is a little posterior
to the anterior vertex. For plano-convex len ses it is exactly at the posterior vertex, and for
m en iscu s len ses (includ ing the cornea) it is slightly anterior to the anterior vertex.
Also see how mu ch in flu ence shap e has on LSA and how th is in flu ence is d ifferent
in air and in aqueou s. The in flu ences of shap e are d ifferent w ith the len s in the con-
verging light beh ind the cornea. Therefore, no in ference can be m ad e on how an IOL
w ill p erform in the eye from its p erform ance in isolation in neither air nor aqu eou s. In
 Special Circumstan ces: In fluen ce of Sph erical Aberration 209

Figure 35-2. Graphic Calculator Ray tracing of a normal eye with a 5 mm pupil,
with marginal (blue), focusing (green), and paraxial (red) rays drawn. Dimensions
are chosen to exactly focus with a 20 D fully correcting aspherical IO L.
Top view: Entire eye with spectacle. The lens haptic plane (LHP) is shown.
Bottom view: Focusing of rays at the receptor plane (brown line). The thin black
line depicts the inner limiting membrane 0.25 mm anterior to the receptor plane
of the retina. Activation of the spreadsheet allows 4 cases to be illustrated. Try
the buttons on the spreadsheet and see how spectacle power and rays at the
retina change when the optic is reversed, the IO L is flipped around LHP, or when
the lens is spherical.

the p seudoph akic eye it is the com bined spherical aberration of the cornea and the IOL
that m atters.
Popu lar IOL p ower formu las u se th in lens theory, in wh ich p ower is associated w ith
the principal plane of lenses. However, the p ower in the principal plane of the cornea is
about 1.5 D less in its principal plane than given by most keratometers.2
The principal plane of the IOL sh ifts anteriorly in the converging light beh ind the cor-
nea.3 The AL measu red by A-scan u ltrasou nd (and also by the IOLMaster) is shorter than
the d istance from the corneal principal plane to the im age plane on the photo receptors,
wh ich is the d istance that shou ld be u sed in an optical calcu lation. With norm al (positive)
spherical aberration, the average pseudophakic eye is approxim ately 0.5 D stronger than
w ithout spherical aberration for a pupil size of 4 m m. All these d ifferences are system atic
and are absorbed in the IOL con stants that apply to the power formu las. Becau se both
keratometers and AL measu ring instru ments can d iffer system atically from each other,
IOL constants shou ld be “p ersonalized” to give, on the average, zero error in refractive
outcome in a given setting.
Aspheric IOLs that correct all or part of the spherical aberration of the cornea requ ire
adju sted formu la constants to give the desired refractive outcome. Therefore an aspheric
210 Ch apter 35

IOL w ith otherw ise the same design as its spherical cou nterpart, w ill have formu la con-
stants that w ill resu lt in h igher powers being ind icated by the calcu lation.
Activation of the spread sheet allow s 4 cases to be illu strated. Try the buttons on the
spread sheet and see how spectacle power and rays at the retina change when the optic is
reversed, the IOL is flipped arou nd LH P, or when the len s is spherical.

Re fe re nce s
1. ISO 11979-2. Ophthalm ic implants—Intraocu lar len ses—Part 2: Op tical prop erties and
test m ethod s. Geneva, Sw itzerland : International Organ ization for Stand ard ization;
2000.
2. Norrby S. Letter: Pentacam keratom etry and IOL p ower calcu lation. J Cataract Refract Surg.
2008;34(1):3;reply 4.
3. Hollad ay JT, Maverick KJ. Relation sh ip of the actu al th ick intraocu lar len s optic to the
th in len s equ ivalent. A m J Ophthalmol. 1998;126(3):339–347.
 36
Sp e cia l Circu msta n ce s:
Mu ltifo ca ls a n d To ric
IOLs
John M oran, M D, PhD

C
alcu lating the optim al power of toric and mu ltifocal IOLs can be viewed as an
extension and refinem ent of the method s u sed to calcu late the power of spheri-
cal IOLs. The sam e principles of m in im izing system atic and random errors by
u sing consistent biometric method s and su rgical tech n iques along w ith optim ized lens
con stants and a modern power formu la (H aigis, Hoffer Q, Hollad ay, SRK/ T) apply.
Add itionally, the desired postoperative d istant and near refractive targets, as well as the
su rgically induced astigm atism, requ ire carefu l con sideration to ach ieve optim al resu lts.

Multifo cal IOLs

The IOL model and power shou ld be selected to m eet the patient’s most important d is-
tance and near visual tasks. For example, a cellist read ing sheet mu sic and a seam stress
requ ire d ifferent near points.
IOLs are available w ith d ifferent strength add s. Two recently introduced IOLs, the
Alcon Acrysof Restor (Alcon USA, Fort Worth, TX) (Fig. 36-1) and the AMO Rezoom
(Abbott Laboratories Inc, Abbott Park, IL) (Fig. 36-2), have 4 D and 3.5 D add s at the IOL
plane that yield approxim ate add powers of about 3.2 D and 2.8 D at the corneal plane,
resp ectively. The optimu m read ing d istance of any IOL model can be adju sted at the
expense of the d istance refraction. For a p ositively powered IOL, increasing the myopic
d istance refraction w ill strengthen the effect of the add power and shorten the read ing

211
Hoffer KJ. IOL Power (pp. 211-214).
© 2011 SLACK Incorporated.
212 Ch apter 36

Figure 36 -1. The Alcon Acrysof Restor multifocal Figure 36 -2. The AMO Rezoom
IO L. multifocal IO L.

d istance wh ile a hyperopic d istance refraction w ill do the opposite. Although modern
mu ltifocal IOLs provide some interm ed iate vision, it is adequ ate to con sider on ly the near
and far point of the postoperative eye for plan n ing pu rposes.
In practical term s, the su rgeon mu st select the IOL add power and d istance refraction
that represents the best comprom ise for each patient. In Table 36-1, the PreVize Mu ltifocal
IOL Plan n ing Gu ide (Moran Research and Con su lting, Inc, Hou ston, TX; PreVize
Calcu lation Center, w w w.previze.com) show s the optimu m read ing d istance and expected
d istance UCVA for several com binations of refractive target and IOL models w ith in a
clin ically realistic range.
As an example, an avid golfer and sport fisherm an who w ishes to see the ball on the
tee and also tie fish ing line wou ld likely be more content w ith a +4.0 D add and -0.75 D
refraction than a +3.5 D add and a plano refraction. A seam stress who prefers a short
working d istance for intricate need le work and also enjoys bird watch ing wou ld be
expected to prefer the latter IOL and refraction.

To ric IOLs
The best possible resu lt from a toric IOL implantation is ach ieved when the IOL and
the principle merid ians of the postoperative cornea are precisely aligned. Wh ile the actu al
power and astigm atism of the postoperative cornea are obviou sly u n know n preopera-
tively, they can be approxim ated by add ing the m ean change in keratom etry, determ ined
for a group of previou sly operated patients, to the preoperative keratometry of the eye to
 Special Circumstan ces: Multifocals an d Toric IOLs 213

Ta b le 3 6 -1 .

PREVIZE MULTIFO CAL IO L PLANNING G UIDE

IO L with +3.5 D Add IO L c + 4 D Add
(eg, AMO Rezoom) (eg, Alcon AcrySof ReStor)
Target Target
Near Point Far Point Distance Near Point Far Point Distance
Ref Ref
D (SEQ ) Inches Feet UCVA* D (SEQ ) Inches Feet UCVA*
1.00 22 N/A 20/40 1.00 18 N/A 20 /40
0.75 19 N/A 20/30 0.75 16 N/A 20/30
0.50 17 N/A 20/25 0.50 15 N/A 20/25
0.25 15 N/A 20/20 0.25 13 N/A 20/20
0.00 14 Infinity 20/20 0.00 12 Infinity 20/20
-0.25 13 13.1 20/20 -0.25 11 13.1 20/20
-0.50 12 6.6 20 /25 -0.50 11 6.6 20 /25
-0.75 11 4.4 20 /30 -0.75 10 4.4 20/30
-1.00 10 3.3 20/40 -1.00 9 3.3 20 /40

*This approximation assumes <0.5 D cylinder and a 3 to 4 mm pupil.

be operated. Th is average keratometric change, know n as the mean su rgically induced

astigm atism (SIA), can be easily calcu lated on line.1 Cylinder correction is h igh ly dep en-
dent on the proper rotational position ing of the IOL relative to the cornea. A 10 degree
m isalign m ent w ill reduce the efficacy of the cylinder correction by 30%.
The refractive target shou ld be chosen so as not to “flip the axis of astigm atism .”
Patients do not tolerate an inversion of the astigm atism axis they have grow n accu stom ed
to. In practice, it is best to choose an add power that leaves the patient w ith u ndercor-
rected rather than overcorrected cylinder.

Furthe r Re fine m e nts

Most cu rrently available software for calcu lating toric IOL power u ses simple vector-
based m athem atics (AcrySof Toric IOL Web Based Calcu lators). These method s are lim ited
to corneas w ith orthogonal principal merid ians. Corneas w ith principle merid ian s that
deviate from orthogonal by at least 10 degrees (present in about 10% of astigm ats) can not
be analyzed by these method s.1 More soph isticated ray tracing and m atrix method s mu st
be u sed to analyze these cases prop erly. A toric IOL power calcu lator that can correctly
analyze nonorthogonal Ks is available on line (PreVize Calcu lation Center, w w w.previze.
com and PreVize Optim ized IOL Power Calcu lation Web Service for the STAAR Toric IOL).
Finally, the ability to correct astigm atism is also lim ited by the reliance on corneal su r-
face measu rements, such as keratometry, to deduce the optical properties of the cornea.
The shape and refractive power of the back su rface of the cornea is on ly loosely related
214 Ch apter 36

to the shap e of the anterior su rface. The relative align ment, tilt, and rotation of back
su rface toru s w ith respect to the anterior su rface, all of wh ich effect the optical power of
the cornea, varies from eye to eye. Th is w ill continue to add a random error to toric IOL
power calcu lation s u ntil more accu rate m ethod s of measu ring the optical properties of
the cornea are available.

Ke y Po ints
l Use consistent biom etric and su rgical tech n iques.
l Use a modern IOL power formu la and optim ize IOL constants, wh ich can be easily
calcu lated u sing Hoffer Program s®, Hollad ay IOL Con su ltant®, or on line (PreVize
Calcu lation Center, w w w.previze.com).
l Match the d istance refraction and add power of mu ltifocal IOLs to the patients most
important visu al tasks.
l Determ ine and u se p ersonalized SIA in plan n ing IOL procedu res.
l Select toric add p owers that do not over correct refractive cylinder.

Re fe re nce
1. H arris WF. Interpretating nonorthogonal keratom etric m easu rem ents. Ophthalmic Physiol
Opt. 2001;21(3):253–254.
 37

Sp e cia l Circu msta n ce s:

Pe d iat ric Eye s
Scott K. M cClatchey, CA PT, MC, USN, M D

Backg ro und

T
here are two concerns in choosing an IOL power for a ch ild : the in itial IOL calcu la-
tion and the grow th of the eye. For ch ild ren you nger than about age 10, the grow th
of the eye has an overwhelm ing impact.
Formu las for calcu lating the in itial IOL power for you ng ch ild ren are relatively inac-
cu rate, compared to the same formu las in adu lts. And reo et al1 stud ied 47 con secutive
p seudophakic patients age 3 months to 16 years and fou nd no sign ificant d ifference in
accu racy between the several IOL calcu lation formu las: the average in itial p ostop erative
refractive error was between 1.2 and 1.4 D for all formu las.
Gordon and Don zis2 showed that a norm al ch ild’s eye has little change in refraction
(0.9 D from birth th rough adu lthood) becau se the power of the natu ral lens decreases
d ram atically as the eye grow s. McClatchey and Parks 3 showed that the refraction of apha-
kic ch ild ren’s eyes has a large myopic sh ift (10 D from in fancy th rough adu lthood). The
m ean refraction of these eyes follow s a logarith m ic cu rve from in fancy th rough age 20.
The optics of a grow ing pseudophakic eye resu lt in a magnification of th is aphakic myopic
sh ift. The grow th of a pseudophakic ch ild’s eye therefore resu lts in a large myopic sh ift.

215
Hoffer KJ. IOL Power (pp. 215 -218).
© 2011 SLACK Incorporated.
216 Ch apter 37

Ta b le 3 7 -1 .
1
PREDICTIO N TABLE FO R TYPICAL PEDIATRIC PSEUDO PHAKIC EYES
Predicted refractions at a given age1
Age at IOL power2 D Initial 1 yr. 2 yr. 4 yr. 8 yr. 20 yr.
surgery PO Rx2
3 mo. 26.9 +7.00 3.30 1.68 0.07 -1.54 -3.64
6 mo. 25.1 +6.50 4.23 2.13 0.04 -2.03 -4.73
1 yr 24.5 +5.00 5.00 2.77 0.70 -1.35 -4.02
2 yr 22.3 +4.00 NA 4.00 2.01 0.03 -2.53
3 yr 22.3 +3.00 NA NA 2.17 0.19 -2.40
4 yr 22.0 +2.25 NA NA 2.25 0.28 -2.29
6 yr 21.0 +1.50 NA NA NA 0.70 -1.82
8 yr 20.4 +1.00 NA NA NA 1.00 -1.50

Assumed A-constant = 118.0. Calculations are based on eyes with a normal RRG; variations in
RRG and initial ocular measurements will significantly affect these predictions. The large variance
in RRG will lead to a large range of ultimate refractions: these are the expected averages. These IOL
power and initial postoperative refractions are for example only, and are not our recommendations.

Rate o f Re fra ctive Gro w th

Becau se aphakic refraction follow s a logarith m ic decline, a plot of refraction vs. the
log of age for these eyes is a straight line. The slop e of th is line is defined as the “Rate of
Refractive Grow th” (RRG). RRG has u n its of d iopters, and can be calcu lated for pseudo-
phakic eyes by m athem atically “removing” the IOL. There is little d ifference in the RRG
between aphakic and pseudophakic eyes over age 6 months. Recent u npublished analysis
ind icates that there is no sign ificant d ifference, ie, putting an IOL in a ch ild’s eye does not
affect its refractive grow th.
RRG is sm aller in ch ild ren who have su rgery at less than 6 months of age. However, in
ch ild ren who have su rgery after 6 month s of age, RRG is not affected by age at su rgery,
type of cataract, in itial refraction, or controlled glaucom a. No other stud ied factor has
been fou nd to consistently in fluence RRG. The value of RRG has a large stand ard devia-
tion (ie, some eyes grow faster than others).

Dis cus s io n
RRG is u sefu l in clin ical practice becau se it allow s pred iction of refraction in pseudo-
phakic eyes. It is also u sefu l in research. I believe it is the best way to analyze refractive
changes in pseudophakic ch ild ren.4 Calcu lating RRG elim inates the large con fou nd ing
factors of non-linear grow th of ch ild ren’s eyes and the variables of age at su rgery, length
of follow-up and variations in IOL power (Table 37-1).
 Special Circumstan ces: Pediatric Eyes 217

Figure 37-1. Graph of axial length growth

from birth to age 10.

No t e Fr o m t h e Ed it o r
A nother consideration is decreasing the power of the primary IOL by the difference that will
occur with aging to maturity. The graph of eye growth in infancy and childhood (Fig. 37-1) can be
helpful in this regard by projecting the change from implantation to adulthood. That subtracted
power can be implanted as a second piggyback IOL over the top of the primary or as a phakic
IOL, either of which could be removed when the patient grows up. In the extremely young, this
could be done using three lenses.

Re fe re nce s
1. And reo LK, Wilson ME, Sau nd ers RA. Pred ictive valu e of regression and theoretical
IOL formu las in p ed iatric intraocu lar len s implantation. J Pediatr Ophthalmol Strabismus.
1997;34(4):240–243.
2. Gordon RA, Don zis PB. Refractive developm ent of the hu m an eye. A rch Ophthalmol.
1985;103(6):785–789.
3. McClatchey SK, Parks MM. Myopic sh ift after cataract removal in ch ild hood. J Pediatr
Ophthalmol Strabismus. 1997;34(2):88–95.
4. McClatchey SK, Hofm eister EM. Intraocu lar len s p ower calcu lation for ch ild ren. In : Wilson
ME, Trived i RH, Pand ey SK, ed s. Pediatric Cataract Surgery: Techniques, Complications and
M anagement. Ph iladelph ia, PA: Lippincott, William s & Wilkin s; 2005:30–37.

The views expressed in this chapter are those of the author and do not necessarily reflect the
official policy or position of the Department of the N avy, Department of Defense, or the United
States Government.
 38

Sp e cia l Circu msta n ce s:

Pig g yba ck IOLs
Kenneth J. Hoffer, M D

G
ayton 1 was the first to p erform and rep ort implanting a piggyback IOL. A pig-
gyback len s is defined as an IOL that is in serted on top of another IOL. They m ay
be placed together in the capsu lar bag d irectly in contact w ith each other, in the
ciliary su lcu s away from the other IOL, or in the anterior cham ber. Piggyback len ses can
be either placed prim arily (w ith the first IOL) or second arily over a previou sly implanted
IOL. The need for piggyback lenses arises in extrem ely h igh hyp eropes (prim ary) and in
cases of postoperative IOL power error (second ary).

Prim ary Pig g yba ck Calcula tio n

If both IOLs are placed w ith in the capsu lar bag, it is estim ated that the anterior IOL
forces the p osterior IOL more p osteriorly, a d istance equ al to half the central th ickness of
the anterior len s. Th is was show n by Bau meister and Koh nen 2 u sing Scheimpflug stud ies
in 2006. They showed th is d id not occu r if the piggyback lens was placed in the ciliary
su lcu s. The sh ift cau ses the posterior lens (whose focal p oint is moved more posteriorly)
to requ ire more power to m aintain the same focu s. Th is effect d im in ishes the th in ner
(lower power) the anterior lens is, and a th in ner len s is easier to remove if that shou ld
be necessary later. Prim ary piggyback lenses need special calcu lations to adju st for th is
posterior len s sh ift. Th is can be done by simply add ing one-half the central th ickness
of the anterior IOL to the ELP being u sed by the formu la to calcu late the power of the
posterior IOL.
219
Hoffer KJ. IOL Power (pp. 219 -220).
© 2011 SLACK Incorporated.
220 Ch apter 38

Here is an example of how th is wou ld be done. If we have an eye that requ ires a 36 D
IOL power and we have to split the p ower between two lenses, the h ighest p ower we can
obtain for a posterior cham ber (PC) lens is 30 D. Since we know we w ill need more than
36 D in th is situ ation, we obtain from the m anu factu rer the central th ickness of a 7 D
IOL, wh ich is 0.942 m m. We now add one-half of th is th ickness (0.47 m m) to the ELP the
formu la calcu lates for the posterior lens (eg, 4.37 m m) to obtain the ELP we need to u se
(4.37 + 0.47 = 4.84.) We shou ld now u se 4.84 as the ELP to perform the final calcu lation.
The Hollad ay IOL Consu ltant ® has th is calcu lation bu ilt in to the program, m aking th is
easier to p erform.

Se co ndary Pig g yba ck Calcula tio n

Second ary len ses can be calcu lated u sing the refraction formu la of Hollad ay 3 (u sed for
ahakic eyes and phakic IOL power calcu lation) or by a more simple formu lation based on
the fact that the healed prim ary IOL is more stable. Due to the d ifferent effect on vertex
p ower changes between plu s and m inu s lenses, the follow ing formu lation works well:
Hyperopic Error: Piggyback IOL = 1.5 x RxError
Myopic Error: Piggyback IOL = 1.0 x RxError
where RxError = the postoperative spherical equ ivalent refractive error needed to be
corrected.
H abot-Wilner and associates 4 from Israel rep orted excellent pred iction resu lts u sing a
slight variation of the formu lation for hyp eropic error:
Hyperopic Error: Piggyback IOL = 1 +1.4 x RxError
They reported a mean pred iction error on 10 eyes of 0.46 ± 0.40 D.
Sham m as 5 has suggested a formu lation utilizing the A con stant of the IOL as fol-
low s:
Hyperopic Error: Piggyback IOL = [RxError/ [(0.03 * (138.3 – A)] – 0.5
Myopic Error: Piggyback IOL = [RxError/ [(0.04 * (138.3 – A)] – 0.5

Re fe re nce s
1. Gayton JL, Sand ers V, Van Der Karr M, Raanan MG. Piggybacking intraocu lar implants
to correct p seudophakic refractive error. Ophthalmology. 1999;106(1):56–59.
2. Bau m eister M, Koh nen T. Scheimpflug m easu rem ent of intraocu lar len s p osition after
piggyback im plantation of fold able intraocu lar len ses in eyes w ith h igh hyp eropia.
J Cataract Refract Surg. 2006;32:2098–2104.
3. Hollad ay JT. Refractive p ower calcu lation s for intraocu lar len ses in the phakic eye. A m J
Ophthalmol. 1993;116(1):63–66.
4. H abot-Wilner Z, Sach s D, Cahane M, et al. Refractive resu lts w ith second ary piggy-
back implantation to correct p seudophakic refractive errors. J Cataract Refract Surg.
2005;31(11):2101–2103.
5. Sham m as H J. The Sham m as refractive equ ation s. In : H. Sham m as, ed. IOL Power
Calculations. Thorofare, N J: SLACK Incorp orated ; 2004:60-61.
 39

Sp e cia l Circu msta n ce s:

Silico n e Oil Po w e r
Kenneth J. Hoffer, M D

I
n calcu lating IOL power, there are 2 m ajor problem s when the posterior segment is
filled w ith silicone oil:

1. A physical one affecting u ltrasou nd AL measu rem ent due to the slow US
velocity th rough the oil filling the large length of the eye.
2. An optical one affecting the overall optical power of the eye due to the
refractive index of the oil compared to vitreou s.
Often, such eyes have seriou s retinal cond itions wh ich preclude a postoperative acu ity
of 20/ 40 or better and thu s postop erative stud ies are qu ite d ifficu lt.

Silico ne Oil Ultra s o und Effe c t

Silicone oil-filled eyes are an especially vexing problem becau se the u ltrasou nd wave is
so slowed dow n crossing the posterior segm ent that it is often impossible to get a read ing
at all. The sou nd velocity d rops from the vitreou s velocity of 1532 m/ sec to the silicone
oil velocity of 980 m/ sec; a 36% decrease. There are also d ifferences in velocity depend-
ing upon the type of silicone oil. Preset u ltrasou nd instru ments that do not allow chang-
ing the velocities m akes it impossible to obtain an accu rate read ing. It is also d ifficu lt in
some eyes to determ ine what p ercentage of the vitreou s body is filled and dep end ing on
the position of the patient, what parts the beam is going th rough. Some have suggested

221
Hoffer KJ. IOL Power (pp. 221-222).
© 2011 SLACK Incorporated.
222 Ch apter 39

mu ltiplying the AL read ing obtained w ith u ltrasou nd by 0.71. I have no exp erience w ith
th is suggestion.
Silicone oil eyes are best measu red w ith the IOLMaster or the LenStar becau se the
optical laser is not appreciably affected by the silicone oil, though fixation m ay be a prob-
lem .
Th is problem cou ld easily be solved if all su rgeons replacing the vitreou s w ith silicone
oil wou ld ensu re that an im mersion A-scan or IOLMaster AL measu rement was taken
and recorded prior to doing so. The resu lt shou ld be given to the patient w ith in structions
to keep it safe and provide it to any su rgeon plan n ing to perform cataract su rgery.
Other alternatives are not perform ing a prim ary implant and later perform ing a sec-
ond ary IOL after the aphakic refraction is obtained. Also, one cou ld consider a piggyback
lens or phakic IOL to correct the error.

Silico ne Oil Re frac tive Effe c t

The second problem that arises when the vitreou s is replaced w ith silicone oil is that
the refractive index of the oil is much less than that of the vitreou s and the silicone oil
acts as a negative-powered lens in the eye. Th is mu st be offset w ith more power in the
IOL. Obviou sly, th is factor can be ignored if the silicone oil w ill be removed completely
at some tim e in the futu re.
Th is optical effect is heavily dependent upon the shap e factor of the back su rface of the
IOL, such that a men iscu s len s w ith the concave su rface facing posteriorly cau ses practi-
cally no effect. But these IOLs are no longer com m ercially available. With a plano-convex
lens, w ith the plano su rface facing posteriorly, cau ses a moderate effect such that 2 to 3 D
mu st be added to the IOL power to compensate for th is silicone effect.
The greatest problem is w ith biconvex IOLs that have a convex su rface facing the sili-
cone oil. Th is situ ation requ ires 3 to 5 D of added IOL power to offset th is effect.
 40
Sp e cia l Circu msta n ce s:
Effe ct o f IOL Tilt o n
Ast ig matis m
Susana M arcos, PhD, FOSA, FEOS
with Patricia Rosales, PhD; A lberto de Castro, M Sc; and Ignacio Jiménez-A lfaro, M D, PhD

Intro ductio n

A
s IOLs becom e more soph isticated, the question arises whether the p otential
improvem ent in ocu lar optical qu ality of the new design s m ight be comprom ised
by proper IOL centration.1 The u se of aberrometry allow s u s to measu re the opti-
cal aberration s of the eye includ ing astigm atism and, in particu lar, to evalu ate optical
qu ality after cataract su rgery in pseudophakic patients. However, add itional new tools
allow fu ll evalu ation of the d ifferent contributions to optical degrad ation (astigm atism, in
particu lar) in eyes w ith IOLs. Astigm atism in pseudophakic eyes can arise from variou s
sou rces, includ ing natu ral corneal astigm atism, astigm atism induced by the incision, the
eccentric fixation of the fovea, or tilt and decentration of the IOL. Among them, the inci-
sion-induced astigm atism is the most relevant.2-4
Whether norm al amou nts of tilt and decentration of the IOL can cau se sign ificant
amou nts of astigm atism (and other h igher order aberrations) is of h igh interest. It is par-
ticu larly imp ortant in new aspheric designs, where the correction of spherical aberration
(and improved retinal im age qu ality) is aim ed. Despite in itial concerns that aspheric IOLs
m ay be more su sceptible to increasing h igher order aberrations (such as com a),1 recent
stud ies show that th is does not seem to be the case in comparison w ith spherical IOLs.5-8
Futu re IOL trend s aim ing at ind ividu al cu stom ization of designs also rely on the absence
of relevant induced IOL tilt and decentration.9
This chapter w ill present simple computations on the theoretical astigm atism induced
by a tilted lens. However, it w ill be show n that real measu rements are essential to assess the
223
Hoffer KJ. IOL Power (pp. 223-230).
© 2011 SLACK Incorporated.
224 Ch apter 40

impact of real amou nts of tilt and decentration on astigm atism. Cu stom-developed instru-
mentation to measu re IOL tilt and decentration w ill be described. In add ition, a cu stom
model eye w ill be presented that allows accu rate estim ations of the astigm atism induced
by IOL tilt and decentration (in comparison w ith a perfectly centered IOL), and in relation
to other sou rces of astigm atism, in particu lar, and optical degradation in general.

The o re tical As tig m atis m

Induce d by a Tilte d Le ns
Simple aberration theory show s that a narrow beam of light entering a spherical lens
obliquely (not parallel to the axis of the lens), w ill show marginal oblique astigmatism. In
general, a horizontal or vertical tilt of the len s potentially induces an astigm atic effect.
Although d ifferences are fou nd w ith the geometry of the len s, simple estim ations can be
p erformed for a th in len s. Theoretically, a th in lens of power P in air, located in the pupil
plane and tilted by angle α , generates an astigm atism of A w ith an angle p erpend icu lar
to the tilt (α ) axis given by 10
(1) A = P [1 + (sin α ) 2 / 3] * (tan α ) 2
where A = astigm atism (D), P = power of IOL (D)
For example, an ophthalm ic glass in air of +22 D, tilted 10° wou ld produce an oblique
astigm atism of 0.69 D.
A = 22 [1 + (sin 10°) 2/ 3] * (tan 10°) 2 = 22 [1 + (0.1736) 2/ 3] * (0.1763) 2
A = 22 [1.01005] * (0.03109) = 22(0.0314) = 0.6908 D

Although ind icative of the order of m agn itude of the astigm atism induced by a tilted
IOL, equ ation (1) is largely simplified to be applied in pseudophakic eyes, where:
a. the IOL is im mersed in aqueou s
b. the rays of light converge on the IOL from the cornea
c. the cornea and IOL form a compou nd optical system
d. the IOL is not a th in lens, and its design has an impact on optical qu ality, includ ing
the degrad ation cau sed by off-axis view ing
e. the IOL does not lie on the pupil plane of the system
f. the eye is not a centered optical system, w ith the fovea tilted w ith respect to the
“optical axis”
g. the IOL is both tilted and decentered w ith resp ect to the pupillary axis
h. several factors, includ ing corneal astigm atism and incision-induced astigm atism,
contribute to the total astigm atism in the eye
i. the actu al amou nt and orientation of tilt and decentration of the IOL shou ld be con-
sidered for correct estim ates of their impact on im age qu ality

Me a s ure m e nt o f IOL Tilt

IOL tilt and decentration can be measu red in vivo u sing Pu rkinje im aging (Fig. 40-
1A),11-14 and Scheimpflug im aging (Fig. 40-1B).15-16 Both tech n iques have been valid ated
 Special Circumstan ces: Effect of IOL Tilt on Astig matism 225

Figure 40 -1. Examples of

images obtained from Purkinje
imaging (A) and Scheimpflug
imaging (B) in pseudophakic
eyes. The reflections are PI
(anterior cornea,) PIII (anterior
IO L surface) and PIV (posterior
IO L surface.) The surface imag-
es in the Scheimpflug photo
have been fitted by conics.

u sing physical eye models and u sed to estim ate tilt and decentration of both the natu ral
lens in phakic eyes and IOLs in patients.
Pu rkinje im ages are reflections from the ocu lar su rfaces of the eye. A w idespread
implementation of a Pu rkinje im aging system to m easu re tilt and decentration assu mes
that the locations of the Pu rkinje im ages of a p oint sou rce are linearly related w ith eye
rotation (β), IOL tilt (α ), and IOL decentration (d).11,17,18
PI=Aβ
PIII =Bβ + Cα + Dd
PIV=Eβ + Fα +Gd
where PI, PIII and PIV are the location s of the Pu rkinje reflection s from the anterior
cornea, anterior and posterior lens respectively, and A-G are coefficients cu stom ized to
the anatom ical parameters of the eye.
Scheimpflug im aging allow s captu re of anterior segment im ages w ith a large depth of
focu s, although they are su bject to optical and geometrical d istortion.19 The d istortion-
corrected Scheimpflug im ages can be processed to obtain the pupillary axis (join ing the
center of rotation of the cornea and the pupil center) and the IOL axis (join ing the center
of cu rvatu re of the anterior and posterior IOL su rfaces), and therefore the tilt of the IOL.
Fig. 40-1 show s typical im ages (Pu rkinje A, Scheimpflug B) obtained in experimental
system s on pseudophakic eyes from the Visu al Optics and Biophoton ics Laboratory in
Mad rid, Spain. Fig. 40-2 show s the amou nts of tilt measu red u sing Pu rkinje im aging in
30 p seudophakic eyes. The average tilt of the IOL was 0.25° (arou nd the horizontal axis)
and 1.79° (arou nd the vertical axis). The average amou nt of decentration of the IOL (not
depicted) was 0.237 m m horizontally and -0.039 m m vertically. Tilt and decentration
tended to be m irror-sym metric in right and left eyes.

Cus to m ize d Ps e udo phakic Co m pute r Eye Mo de ls

The impact of IOL tilt (and decentration) can be accu rately evalu ated u sing cu stom ized
computer eye models.7,20 These models are bu ilt u sing the anatom ical parameters mea-
su red for an eye. The u se of these models overcomes all the lim itation s of the theoretical
simple computations (Formu la 1, previou sly) and allow s u nderstand ing the contribution
226 Ch apter 40

Figure 40-2. IO L tilt measured using Purkinje

images in 30 eyes (O D and O S). Horizontal
tilts stands for tilt around the horizontal axis
and vertical tilt is for tilt around the vertical
axis.

of each factor to the optical qu ality (includ ing astigm atism) in p seudophakic eyes, and
most imp ortantly, to p erform th is evalu ation on an ind ividu al basis.
We have presented cu stom ized eye models7 that include anterior su rface topography
(measu red w ith videokeratoscopy), anterior cham ber depth and axial length (measu red
w ith low coherence interferom etry), IOL tilt and decentration and foveal m isalign m ent
(u sing Pu rkinje im aging), and IOL geom etry. The posterior corneal su rface (u n measu red)
was assu med spherical.
Fig. 40-3 show s a schem atic d iagram of the pseudophakic eye model u sed to evaluate
the impact of the d ifferent factors (includ ing IOL tilt) on the optical aberrations (includ ing
astigm atism), along w ith the variou s sou rces of the m easu red anatom ical parameters.

Effe c t o f Re al IOL Tilt o n Ocular As tig m atis m

Computer eye models allow pred iction of the ocu lar aberrations by ray tracing. These
estim ates can be compared to actu al m easu rements of aberrations on the sam e eyes.
We computed the aberrations in 12 cu stom ized pseudophakic eye models implanted
w ith aspheric IOLs (Acrysof IQ) and compared them to aberrations m easu red on those
patients.7 The evaluation of the impact of IOL tilt and decentration on the aberration s was
ach ieved by comparing the optical qu ality, assu m ing that the IOL was centered on the
pupillary axis (no IOL tilt and decentration), or subject to the actu al measu red com bina-
tions of tilt and decentration.
Pred icted and measu red aberrations are described by sets of Zern ike coefficients (up to
the 7th order.) For the pu rposes of th is chapter, we analyzed the astigm atism coefficients
in the Zern ike expansion (astigm atism at 0/ 90° Z 22 and astigm atism at 45° Z 22).
Fig. 40-4 compares corneal astigm atism, m easu red total astigm atism, and simu lated
astigm atism (assu m ing no IOL tilt/ decentration and w ith their real amou nts). We fou nd a
good correlation between the simu lated and measu red astigm atism . There is a consistent
 Special Circumstan ces: Effect of IOL Tilt on Astig matism 227

Figure 40-3. Computer eye model showing the relevant axes and angles (IO L
tilt a and foveal misalignment l ,) and the instruments used to customize it for
each eye. Estimated aberrations (including astigmatism) by ray tracing on this
eye model were compared to the aberrations measured using Laser Ray Tracing
Aberrometry.

Figure 40-4. Astigmatism computed from the Zernike polynomial expansions

in simulated and measured wave aberrations in 12 pseudophakic eyes showing:
anterior corneal astigmatism (green); simulated astigmatism from computer eye
models; a) assuming a centered IO L (yellow), and b) assuming a tilted /decen-
tered IO L (red) as well as measured astigmatism (blue).
228 Ch apter 40

overestim ation of the simu lated astigm atism by 17.3% for Z 22 and by 8.7% for Z 2-2 w ith
resp ect to the real m easu rements. Th is resu lt is ind icative of a comp ensatory role of the
astigm atism by the real posterior corneal su rface.21-22 The resu lts for the cornea alone are
sim ilar to those of the model w ith the IOL (either lined up or tilted/ decentered), ind icat-
ing that the cornea is the m ajor sou rce of astigm atism in pseudophakic eyes and IOL tilt
and decentration play a negligible role in the postoperative astigm atism.

Co nclus io ns
1. IOL tilt can be measu red accu rately u sing im aging tech n iques, provided w ith vali-
d ated routines for im age analysis and qu antification.
2. With state-of-the-art monofocal IOLs, tilt is typically less than 5°.
3. Computer eye models can be u sed to accu rately estim ate the contribution of IOL tilt
and decentration on optical qu ality. These models con sider the ind ividu al interac-
tions among all geometrical param eters of the eye.
4. In modern cataract su rgery w ith monofocal IOLs, the effect of IOL tilt on ocu lar
astigm atism is neglible.

Re fe re nce s
1. Atch ison DA. Design of aspheric intraocu lar len ses. Ophthalmic Physiol Opt. 1990;11(2):137-
146.
2. Gu irao A, Tejedor J, Artal P. Corneal aberration s before and after sm all-incision cataract
su rgery. Invest Ophtalmol Vis Sci. 2004;45:4312-4319.
3. Marcos S, Rosales P, Llorente L, Jim enez-Alfaro I. Change of corneal aberration s after
cataract su rgery w ith two typ es of aspheric intraocu lar lenses. J Cataract Refract Surg.
2007;33:217-226.
4. Jacobs BJ, Gaynes BI, Deutsch TA. Refractive astigm atism after oblique clear corneal
phacoemu lsification cataract incision. J Cataract Refract Surg. 1999;25:949–952.
5. Barbero S, Marcos S, Jim enez-Alfaro I. Optical aberration s of intraocu lar lenses m easu red
in vivo and in vitro. J Optom Soc. 2003;20:1841-1851.
6. Marcos S, Barbero S, Jim énez-Alfaro I. Optical qu ality and depth-of-field of eyes implant-
ed w ith spherical and aspheric intraocu lar len ses. J Refract Surg. 2005;21:223-235.
7. Rosales P, Marcos S. Cu stom ized computer mod els of eyes w ith intraocu lar len ses. Optics
Express. 2007;15:2204-2218.
8. Marcos S, Rosales P, Llorente L, Barbero S, Jim énez-Alfaro I. Balance of corneal horizontal
com a by internal optics in eyes w ith intraocu lar artificial len ses: Evid ence of a passive
m echan ism . Vision Res. 2008;48:70-79.
9. Barbero S, Marcos S. Analytical tools for cu stom ized d esign of monofocal intraocu lar
len ses. Optics Express. 2007;15:8576-8591.
10. Fan n in T, Grosvenor TP (Ed s). Clinical Optics (2nd ed ition). New York, N Y: Elsevier,
Inc.;1996:48-50.
11. Rosales P, Marcos S. Phakom etry and len s tilt and d ecentration u sing a cu stom-d evelop ed
Pu rkinje im aging apparatu s: Valid ation and m easu rem ents. J Opt Soc A m A . 2006;23:509-
520.
12. Tabernero J, Ben ito A, Nou rrit V, Artal P. In stru m ent for m easu ring the m isalign m ents of
ocu lar su rfaces. Optics Express. 2006;14:10945-10956.
 Special Circumstan ces: Effect of IOL Tilt on Astig matism 229

13. Schaeffel F. Binocu lar len s tilt and decentration m easu rem ents in healthy su bjects w ith
phakic eyes. Invest Ophthalmol Vis Sci. 2008;49:2216-2222.
14. Rosales P, d e Castro A, Jim énez-Alfaro I, Marcos S. Intraocu lar len s align m ent from
Pu rkinje and Scheimpflug im aging. Clinical and Experimental Optometry. 2010;93;400-408.
15. d e Castro A, Rosales P, Marcos S. Tilt and d ecentration of intraocu lar len ses in vivo from
Pu rkinje and Scheimpflug im aging - Valid ation study. J Cataract Refract Surg. 2007;33:418-
429.
16. Copp en s JE, van der Berg TJ, Budo CJ. Biom etry of phakic intraocu lar lens u sing
Scheimpflug photography. J Cataract Refract Surg. 2005;31(10):1904-1914.
17. Ph illip s P, Perez-Em m anuelli J, Rosskothen H D, Koester CJ. Measu rem ent of intraocu lar
len s d ecentration and tilt in vivo. J Cataract Refract Surg. 1988;14:129-135.
18. Barry JC, Du n ne M, Kirsch kamp T. Phakom etric m easu rem ent of ocu lar su rface rad iu s
of cu rvatu re and align m ent: Evalu ation of m ethod w ith physical model eyes. Ophthalmic
Physiol Opt. 2001;21:450-460.
19. Rosales P, Marcos S. Pentacam Scheim pflug qu antitative im aging of the crystalline len s
and intraocu lar len s. J Refract Surg. 2009;25:421-428.
20. Tabernero J, Piers P, Ben ito A, Redondo M, Artal P. Pred icting the optical p erform ance
of eyes implanted w ith IOLs to correct spherical aberration. Invest Ophthalmol Vis Sci.
2006;47:4651-4658.
21. Du n ne M, Royston J, Barnes D. Posterior corneal su rface toricity and total corneal astig-
m atism . Optom Vis Sci. 1991;68:708-710.
22. Du bbelm an M, Sicam V, Van der Heijd e GL. The shap e of the anterior and posterior su r-
face of the aging hu m an cornea. Vision Research. 2006;46:993-1001.
 41
Sp e cia l Circu msta n ce s:
An ise iko n ia a n d
An iso m e tro p ia
Kenneth J. Hoffer, M D

A
nisometropia is defined as a d ifference in refractive error between the two eyes, and
is often con fu sed w ith an iseikon ia. A niseikonia is defined as the binocu lar statu s of
u nequ al im age sizes projected on the two m acu las (iseikonia means equ al im ages).
An iseikon ia can lead to a d iscom forting an noyance termed asthenopia if the percentage
d ifference is greater than 14%. Obviou sly, in you ng ch ild ren u nder 6 years of age th is
situ ation can lead to amblyopia.
The cau ses of an iseikon ia are due to a com bination of axial length, corneal power, lens
power, and spectacle correction. Becau se of variou s biological d ifferences in the above
parameters, it is possible for an ind ividu al to have an isom etropia but still be iseikon ic as
well as the opposite being true.
Herm an n Gernet of Mü nster (now living in Wü rzbu rg), Germ any, devoted a lot of time
in the late 1960s and early 1970s to the evalu ation and prevention of th is cond ition, espe-
cially after IOL implantation. He u sed complex formu las and m ain fram e computers to
perform the calcu lations. When I visited h im in 1974, it became obviou s that h is calcu la-
tions wou ld be a very complex m ethod to u se as a routine for the average su rgeons begin-
n ing lens implantation in the US at that time. Follow ing Gernet’s lead, Colenbrander 1
included a specific formu la to calcu late an an iseikon ic IOL power in h is land m ark paper
of 1973.
In 1974, I2 began my work in th is field u sing the Colenbrander formu la and performed
a comparison of the em m etropic and an iseikon ic formu las to calcu late IOL powers. The
resu lt yielded an eye that wou ld be approxim ately 1 to 1.50 D more myopic than the target
refraction originally aim ed for. Since my goal was em m etropia and not pu rposely cau s-

231
Hoffer KJ. IOL Power (pp. 231-232).
© 2011 SLACK Incorporated.
232 Ch apter 41

ing myopia, I refrained from u sing the an iseikon ia formu la. Over the next several years,
there d id not seem to be any patient complaints attributable to an iseikon ia. Th is is the
reason you do not hear th is su bject being d iscu ssed regard ing IOL power. On the other
hand, p erhaps some of the vague complaints patients relay after IOL implantation m ight
be attributable to th is problem . More research is needed.

Re fe re nce s
1. Colenbrand er MC. Calcu lation of the p ower of an iris clip len s for d istant vision. Br J
Ophthalmol. 1973;57(10):735–740.
2. Hoffer KJ. Intraocu lar lens calcu lation : The problem of the short eye. Ophthalmic Surg.
1981;12(4):269–272.
 42

Pre ve n t in g IOL
Po w e r Erro rs
Kenneth J. Hoffer, M D

T
able 42-1 gives a su m m arized outline of the m any th ings that can be done in clin i-
cal practice to improve IOL power calcu lation accu racy and prevent the d readed
IOL power su rprise.

Pre ve ntio n o f Co m m o n IOL Po w e r Erro rs

l Know more about IOL power calcu lation than you r employees.
l Employ a well-trained, exp erienced tech n ician.
l Use on ly the IOLMaster, H aag-Streit LENSTAR LS900, or im mersion A-scan to m ea-
su re the AL.
l Carefu lly evalu ate the IOLMaster scan for reliability.
l If importing external K read ings, be su re to set the index of refraction (IR) to 1.3375
in the setup screen of the IOLMaster for the Hoffer Q formu la to yield correct
resu lts.
l Su spect a staphylom a in eyes >25 m m : Use IOLMaster and/ or Sham m as A/ B-scan
tech n ique.
l Use CALF AL method : measu re the eye u sing 1532 m/ s and add +0.32 m m to the
resu lt to correct for any error in sou nd velocity (u sefu l in very long and short eyes).
l Silicone oil eyes need IOLMaster if possible or u ltrasou nd AL times 0.71.

233
Hoffer KJ. IOL Power (pp. 233-236).
© 2011 SLACK Incorporated.
234 Ch apter 42

Ta b le 4 2 -1 .

SUMMARY FO R PREVENTING IO L POWER ERRO RS

Summary
• Calculate power accurately
• O rder PC and AC powers
• Fill out O R sheet
• Place O R sheets on wall and scope
• Prior to insertion, check power against sheet

Figure 42-1. O perating

room IO L sheet listing
the various powers of
potential IO Ls that may
be needed, as well as
amount and axis of cyl-
inder and the endothelial
cell count. Note red sheet
for right eyes.

l Regu larly calibrate m anu al keratometers.

l Keep the CL out completely for 2 weeks prior to keratometry (at least in one eye).
l Use the Hoffer Q formu la in eyes <22 m m and in post-refractive su rgery short eyes.
l Use the Hollad ay 1 formu la in eyes 24.5 to 26 m m in length.
l Use the SRK/ T formu la in eyes longer than 26 m m .
l Never use t he SR K Regression for mula s (SR K I or II).
l Personalize you r ELP factors in the formu las.
l The su rgeon shou ld personally select the IOL power for the ind ividu al patient.
l Prepare a sheet w ith all IOL powers that m ay be needed (Fig. 42-1) and place it on
the wall in the OR for the staff (Fig. 42-2) and a m in im ized copy on the m icroscope
 Preven tin g IOL Pow er Errors 235

Figure 42-2. O perating

room IO L sheet attached
full size to the wall, read-
ily available to the nursing
staff. Note Yellow sheet for
Left eyes.

Figure 42-3. O perating

room IO L sheet attached
as a minimized copy to
the operating microscope,
readily available to the
surgeon while operating.

for the su rgeon to verify the correct IOL power (Fig. 42-3). Use red paper for right
eyes and yellow paper for left eyes. Other important d ata can also be added to the
sheet.
l Use the variou s method s available to better calcu late an IOL power in p ost-refrac-
tive su rgery corneas:
© Clin ical H istory Method if pre-LASIK refraction and K are available.
© Contact Lens Method if eye can be refracted effectively. H ave a set of PMMA hard
CLs in the clin ic and be su re the staff know s where they are.
© Sham m as “No H istory” Formu la: K = 1.14 * KPO – 6.8.
© Maloney or Koch Corneal topography m ethod s.
236 Ch apter 42

© Use the Aram berri Double-K: Calcu late the ELP u sing the preoperative K and the
IOL power u sing the PO K.
© Sp eicher/ Seitz, Savin i, and Masket Method s.
© H aigis-L formu la.
l Consider delaying the IOL implantation u ntil the cornea has healed after a penetrat-
ing keratoplasty, rather than perform ing a “triple procedu re.”
 43

Dia g n o s in g a n d Tre at in g
IOL Po w e r Erro rs
Kenneth J. Hoffer, M D

T
h is chapter w ill d iscu ss ways to d iagnose IOL power errors early and treat them
qu ickly for the benefit of the patient as well as the su rgeon.

Diag no s ing IOL Po w e r Surpris e s

The biggest th ing one can do to prevent m ed icolegal problem s w ith IOL p ower error is
to m ake it a routine to p erform a K read ing and fu ll m an ifest refraction on the first post-
op erative d ay. With tod ay’s su rgery, th is is p ossible in 95% of cases and takes no longer to
do than it does a week or two later. It has been so ingrained in su rgeon’s practices to ju st
pin hole at one d ay PO and carefu lly refract much later.
Pin hole vision testing can m iss p ower errors and delay d iagnosis and treatment for
several weeks. Day 1 refraction allow s you to d iscover the problem early enough to take
the patient back to the OR and correct the problem in the first 12 to 48 hou rs. The su rgery
w ill be less trau m atic (than later) becau se the incision w ill be easy to reop en, the capsu le
has not scarred and healed arou nd the IOL so it is much easier to remove and replace
a new lens. The patient is im med iately pleased w ith the improved u ncorrected visu al
acu ity (UCVA) and u su ally forgets that they had two procedu res performed when asked
months later. With excellent UCVA (thu s no d am ages), the opportu n ities for med icolegal
action s are almost completely elim inated.
On the other hand, when weeks go by before the d iagnosis is m ade and then several
months pass as the su rgeon attempts to mollify the patient’s concern s, the patient often

237
Hoffer KJ. IOL Power (pp. 237-240).
© 2011 SLACK Incorporated.
238 Ch apter 43

Figure 43-1. McReynolds IO L power analyzer.

is recom m ended to see another ophthalmologist by their friend s and relatives, u nbe-
know nst to the su rgeon. With an obviou s IOL power error, it is d ifficu lt for the con su lting
su rgeon to refrain from u sing the word s, “…the w rong len s was inserted.” That one word
“w rong” becomes m isinterpreted by the patient as “the su rgeon m ade an error.” All th is
can be prevented by early d iagnosis and treatment. The patient is the first to benefit by
having the problem solved in <48 hou rs.

Tre a ting IOL Po w e r Surpris e s

l In trying to determ ine the cau se of the IOL power error, it is cu stom ary to re-m ea-
su re the AL of the newly p seudophakic eye. If u sing u lrtrasou nd, th is necessitates
the u se of the proper u ltrasou nd velocity depend ing on the m aterial of the IOL as
described in Chapter 4. It is also important to ignore the som etimes large US spikes
that appear in the m id-vitreou s wh ich are reduplication echoes cau sed by the IOL.
l It obtain ing a remeasu re of the AL, it m ight be w ise to ask that it be perform ed in
another colleague’s office in case there m ay be a problem w ith you r equ ipment. It is
also w ise if that colleague is not considered a close friend of you rs.
l The best option is to have the AL remeasu red u sing an IOLMaster, if that has not
already been done. Be su re to adju st the menu to the appropriate p seudophakic eye
typ e.
l If the eye has already healed and the cap su lar bag is tightly scarred arou nd the
IOL, consider the u se of a piggyback IOL in the ciliary su lcu s or a phakic IOL. Th is
is appropriate when the eye has healed beautifu lly and removal of the errant IOL
wou ld be more trau m atic to the eye. For myopic error u se 1 tim es the error and for
hyp eropic errors u se 1.5 times the error.
l In those patients where rep eat intraocu lar su rgery is not possible or advisable, con-
sider a m in im al corneal refractive su rgery.
l If one is interested in measu ring the power of a removed IOL in the OR, there is a
device that can do that. It is called the McReynold s Analyzer (Fig. 43-1), and can
 Diag n osin g an d Treatin g IOL Pow er Errors 239

Figure 43-2. IO L Power Club founding members and Executive Committee. Front
row left to right: Jaime Aramberri, MD, San Sebastian, Spain; Kenneth J. Hoffer,
MD, Santa Monica, CA; and Thomas O lsen, MD, Århus, Denmark. Back row
left to right: Sverker Norrby, PhD, Groningen, Holland; H. John Shammas, MD,
Lynwood, CA; and Wolfgang Haigis, PhD, Würzburg, Germany.

on ly be obtained from William McReynold s, MD. (He can be contacted at 217-222-

6656.) It is also possible to request the m anu factu rer of the removed IOL to bring
equ ipm ent to you r OR to perform the measu rement.

Co nclus io n
Th is is the next to last chapter in ou r IOL power book and we all hope you have fou nd
the details we have delved into u sefu l in you r clin ical practice to improve you r ability to
calcu late the IOL power you r patients need. I w ish to personally than k all the mem bers of
the IOL Power Clu b who have taken tim e to participate in th is large endeavor. The Clu b
was fou nded in 2005 by a group of su rgeons and engineers (Fig. 43-2) to promote the sci-
ence of th is field (see App end ix A).
Simple steps and attention to detail can be very u sefu l in preventing IOL power errors.
Since perform ing the first Am erican u ltrasou nd IOL power calcu lation 1 in 1974, the past
35 years have seen great improvement in the accu racy of postoperative refractive pred ic-
tion. Futu re improvements m ay somed ay elim inate the problem s we have left.

Re fe re nce
1. Hoffer KJ. The h istory of IOL p ower calcu lation in North America. In : Kw itko ML,
Kelm an CD, ed s. The History of M odern Cataract Surgery. The H ague, Netherland s: Kuglen
Pu blication s; 1998:193–20.
 44
Fu tu re Dire ctio n s in
IOL Po w e r Ca lcu la t io n :
In tra o p e rat ive
Re fra ctive Bio m e try
Tsontcho [Sean] Ianchulev, M D, M PH

S
uccess of cataract su rgery is in herently lin ked to the accu racy of IOL calcu lation. As
cataract su rgery has evolved over the years, so have conventional m ethod s of IOL
power calcu lation. Ever since Dr. Fyodorov described the first IOL power formu la
in 1967 based on preoperative keratometry and axial length, a long array of theoretical
and empiric formu las have lead to sign ificant incremental improvements in IOL p ower
calcu lation. Colenbrander (1973), Hoffer (1974), Bin khorst (1975), Van der Heijde (1975),
SRK I (1980), Hollad ay 1 (1988), SRK II (1990), Hoffer Q (1993), Olsen (1995), Hollad ay 2
(1996), and H aigis (2000) are the more salient em bod iments of continuou s in novation in
the field of IOL biom etry. Tod ay, the 3rd and 4th generation formu las are able to pred ict
the final em metropic power in more than 90% of stand ard cataract cases w ith in ±1 D. As
a resu lt of th is, as well as the concu rrent advancements in phacoemu lsification tech nol-
ogy, modern cataract su rgical intervention has moved to refractive len s exchange.
Despite sign ificant d ifferences across the variou s IOL formu las, they all share the sam e
basic principle deriving from Fyodorov’s original equ ation—they are based on preopera-
tive anatom ic parameters, such as axial length and corneal cu rvatu re. In essence, they
derive an optical variable (d iopters) from non refractive, anatom ic ocu lar parameters. Th is
has been greatly enabled by the advances in u ltrasou nd and optical biometry and kera-
tometry. It also successfu lly bypasses any con fou nd ing effect the cataractou s len s wou ld
have on any preoperative refractive measu rements, if such were to be u sed as a su rrogate
estim ate of lenticu lar corrective power.
Tod ay ou r patients have a much greater exp ectation of ach ieving their desired post-
operative refraction, but some trad itional challenges rem ain and some new ones have

241
Hoffer KJ. IOL Power (pp. 241-246).
© 2011 SLACK Incorporated.
242 Ch apter 44

emerged. Conventional preop erative non refractive biometric approaches continue to

struggle w ith the d ifficu lties associated w ith very short and very long eyes, where most
formu las see an appreciable attrition of effectiveness and precision. More sign ificantly,
the conventional approaches fail to deliver the same sup erior efficacy in cases after
previou s LASIK and PRK su rgery. The laser ablation of the cornea complicates proper
estim ation of keratometric values wh ich are at the core of all conventional formu las. For a
detailed description of these challenges and potential m itigation strategies and solution s
please refer to earlier chapters in th is book.

Intra o pe ra tive Re fra c to m e try:

Initial Co nce pt and Me tho do lo g y
One alternative methodology for IOL calcu lation offers a rather d ifferent approach
to IOL calcu lation, wh ich m ay present potential opportu n ities to add ress some of the
rem ain ing need s in the field. Intraop erative autom ated refractive biom etry can be u sed
to obtain the aphakic spherical equ ivalent of the eye right after the extraction of the
cataractou s lens. In th is transiently-aphakic state (after lens extraction and before IOL
implantation,) one can take a “refractive” biop sy of the eye w ith an autorefracting device
(autoretinoscop e, wavefront aberrometer, etc.) to measu re the lower-order aberrations
(sphere, cylinder) wh ich can then provide the aphakic spherical equ ivalent. Assu m ing
m in im al d istortion of ocu lar optics du ring su rgery (as is typical of tod ay’s m in im ally
invasive phaco tech n iques) as well as h igh accu racy of autorefracting devices, the aphakic
spherical equ ivalent in form s u s about the refractive deficit of the aphakic eye at the vertex
d istance of measu rement. Converting or correlating th is to the power at the intraocu lar
plane of final len s position can provide the necessary em m etropic IOL power. Theoretical
analyses based on Ben nett–Rabbetts1 schem atic eye variants demon strate that the expect-
ed ratio between the aphakic spherical equ ivalent and the final em m etropic power is in
the range of 1.75 to 2.01.
Th is new m ethodology is comp elling in a nu m ber of ways. Becau se of its pu rely
refractive approach, one wou ld pred ict little con fou nd ing by the effect of prior corneal
alteration from refractive su rgery. AL and keratometry are not needed, as their resu ltant
optical effect shou ld be factored into the aphakic autorefraction. In fact, preop erative len s
calcu lation s wou ld be elim inated in th is intraop erative refractive parad igm where d iag-
nostic biometry is done “on the table” after the len s extraction. And the su rgical effect of
the phaco incision on the cornea wou ld also be captu red in th is intraop erative setting.
Wh ile aphakic refraction has been described previou sly as a u sefu l alternative for IOL
p ower calcu lation du ring second ary IOL implantation, applying the tech n ique intraop -
eratively is not trivial. In itial clin ical exp erience has show n that in order to ach ieve the
fu ll potential of th is m ethod, control over and experience w ith a nu m ber of variables is
imp ortant. A reliable autorefractor such as the portable Retinom ax (Nikon, Kanagawa,
Japan) or the Nidek AR-20 device (Nidek, Co. Ltd, H iroish i, Japan) shou ld be u sed (Fig.
44-1A), but ideally it wou ld be better w ith a su rgical m icroscope-integrated device (Fig.
44-1B) such as the ORange (WaveTec Vision, Aliso Viejo, CA). Manu al refraction is d is-
cou raged due to operator dependence and reliability. Vertex d istance, visu al axis centra-
tion, and parallax are important, as are post-phaco corneal statu s, intraocu lar pressu re
(over/ u nder-filled AC), and type of viscoelastic cham ber m aintainer. Fortu nately, w ith
the advent of new intraop erative autorefractive devices, m any of those consideration s w ill
be resolved by tech n ique stand ard ization.
 Future Direction s in IOL Pow er Calculation 243

A
B

Figure 44-1. Intraoperative refractometry after cataract removal and prior to IO L implantation. (A)
Portable autorefractor used during cataract surgery. (B) Microscope-integrated wavefront autorefractor
(O Range).

Sim ilar to conventional m ethod s of IOL calcu lation, the intraoperative refractive tech-
n ique is dependent on a nu m ber of variables that m ay in fluence the final correlation
between aphakic spherical equ ivalent and em m etropic IOL power. Wh ile th is ratio can
show some variability w ith resp ect to both AL and corneal cu rvatu re, by far the most
important factor is the final IOL position. A theoretical evaluation of th is correlation 2
illu strates the impact of final lens position on the correlation coefficient between the
aphakic spherical equ ivalent and final em metropic p ower (ie, the theoretical ratio.) The
d ifference between a more anterior final IOL position versu s a more p osterior placement
can exceed 0.75 D of final em metropia across a w ide range of IOL p owers. Th is effect is
con siderably larger than the one cau sed by variations of corneal cu rvatu re, for example.

Early Clinical Re s ults

The first formu la for intraoperative autorefraction was derived by myself and col-
leagues 3 in 2005 in a series of 38 eyes, 6 of wh ich were post-prior LASIK patients. The
range of the axial length was 21.4 to 25.2 m m w ith a range of IOL power implanted from
12.0 to 28.5 D. Autorefraction vertex d istance was 13.1 and A con stant of the IOL u sed was
118.40. A strong linear correlation was fou nd in a series of 38 eyes across a w ide range of
em metropic IOL powers (Fig. 43-2).
Using a linear regression, the follow ing empiric formu la was derived based on a strong
“linear fit” between aphakic spherical equ ivalent and em metropic IOL:
Ianchu lev Formu la: P = 2.01 x ASE
where P = em metropic IOL power, ASE = aphakic spherical equ ivalent.
In the pu blished series, more than 93% of the variability of the final em metropic power
is accou nted for by the linear relationsh ip w ith aphakic spherical equ ivalent—in stand ard
eyes, the conventional formu las and the optical refractive model showed equ ivalent pre-
d ictive efficacy w ith a correlation coefficient of 0.96. In add ition, 83% of the LASIK eyes
and 100% of the norm al eyes were w ith in ±1 D of the final IOL p ower when aphakic
autorefraction was u sed, compared w ith 67% LASIK eyes and 100% of the norm al eyes
u sing the conventional m ethod.
244 Ch apter 44

Figure 44-2. Linear regression of the original

Ianchulev et al3 series between emmetropic
IO L power and aphakic spherical equivalent
(ASE).

Several other stud ies offer add itional valid ation of the original tech n ique and formu la
described by u s. In a prosp ective, noncomparative consecutive case series of 82 myopic
eyes w ith a m ean preoperative spherical equ ivalent of -12.80 D (range -3 D to -27 D),
Leccisotti et al4 derived a mod ified formu la for h igh myopia p opu lation:
Leccisotti Formu la: P = 1.3 x ASE + 1.45
where P = em metropic IOL power, ASE = aphakic spherical equ ivalent
A more definitive study by Wong et al5 compared the Ianchu lev formu la w ith and
w ithout a Leccisotti mod ification in a series of 182 eyes. Th is study demonstrated that
wh ile the Ianchu lev formu la hold s across the w ide spectru m of IOL powers, the Leccisotti
mod ification perform s slightly better in myopic eyes (AL >25). In add ition, another set of
intraoperative aphakic refractive formu las were derived from th is series as follow s:
For AL <25.5 m m : P = 1.97 x ASE
For AXL $ 25.5m m : P = 0.015 x ASE2 + 1.5 x SE + 1.5
where P = em metropic IOL power, ASE = aphakic spherical equ ivalent
Another more advanced development of the intraoperative m ethodology is the Wavetec
Intraoperative Aberrometer u nder the com mercial name ORange (WaveTec Vision, Aliso
Viejo, CA). Th is is a Talbot-Moire based wavefront aberrometer wh ich is attached to the
su rgical m icroscop e and has been ad apted for intraoperative aphakic measu rements,
particu larly in term s of dynam ic range and accu racy of intraoperative aphakic read ings.
Th is tech nology elim inates a sign ificant part of the variability associated w ith portable
autorefraction where centration and vertex can be problem atic. ORange comes w ith pre-
loaded A-con stants and autom atically in form s the IOL power for each len s type. An add i-
tional d ifferentiating featu re is that in stead of u sing a linear regression, ORange utilizes
the refractive vergence formu la in wh ich the aphakic SE replaces the term in the stand ard
vergence formu la that incorporates the axial length. Th is method requ ires an estim ation of
the len s position (ELP) wh ich ORange accomplishes w ith a u n ique formu la. Prelim inary
resu lts are very prom ising u sing th is approach, a recent analysis of post-su rgical refrac-
tive outcom es in wh ich the most recent ORange-optim ized software was u sed to derive
IOL power recom mend ation s, show s that in a popu lation of 1977 non-p ost-refractive eyes,
81% fell w ith in ± 0.50 D and 94% fell w ith in 1.00 D of intended target. The MAVPE for th is
group was 0.35 w ith a stand ard deviation of 0.48. These are comp elling outcomes wh ich
are con sistent w ith the increasing dem and s in the cataract-refractive space, where not
±1.00 D but <± 0.5 D from emetropia is the futu re m etric of su rgical success.
Ultim ately, the original methodology by Ianchu lev et al m ay establish a pu rely refrac-
tive intraop erative parad igm for IOL calcu lation wh ich helps solve important aspects
of IOL estim ation in p ost-Lasik eyes. It can also be applicable to the stand ard cataract
 Future Direction s in IOL Pow er Calculation 245

case where refractive biom etry can refine and verify the final IOL calcu lation. With the
development of new integrated equ ipment that stream lines the autom ated refraction at
the point of su rgery, sign ificantly h igher accu racy can be ach ieved from measu rement
stand ard ization, better centration of autorefraction, and incorporation of add itional intra-
op erative parameters in the optical analysis such as keratom etry. Intraop erative refractive
biometry for IOL calcu lation m ay u ltim ately represent another imp ortant tangential point
along the expand ing interface between cataract and refractive su rgery.

Re fe re nce s
1. Rabbetts RB. Bennett & Rabbetts’ Clinical Visual Optics (4th ed). Ph ilad elph ia, PA:
Butterworth Heinem an n Elsevier; 2007:237.
2. Sheppard A, Du n ne M, Wolffsoh n J, Davies L. Theoretical evalu ation of the cataract
extraction-refraction-implantation tech n iqu es for intraocu lar len s p ower calcu lation
Ophthal Physiol Opt. 2008;28(6):568–578.
3. Ianchu lev T, Salz J, Hoffer KJ, et al. Intraop erative op tical intraocu lar len s p ower estim a-
tion w ithout axial length m easu rem ents. J Cataract Refract Surg. 2005;31(8):1530–1536.
4. Leccisotti A. Intraocu lar len s calcu lation by intraop erative autorefraction in myopic eyes.
Graefes A rch Clin Exp Ophthalmol. 2008;246:729-733.
5. Wong AC, Mak ST, Tse RK. Clin ical evalu ation of the intraop erative refraction tech n ique
for intraocu lar len s power calcu lation. Ophthalmology. 2010;117:711–771.
 Appendix A
IOL Po w e r Clu b
M embers: Jaime Aram berri, MD, San Sebastian, Spain
H an Bor Fam, MD, Singapore, Malaysia
Massimo Cam ellin, MD, Rovigo, Italy
Claud io Carbonara, MD, Rom e, Italy
Jean-Pierre Colliac, MD, Paris, France
Oliver Find l, MD, Vien na, Au stria
Wolfgang H aigis, MS, PhD, Wü rzbu rg, Germ any
Ken neth J. Hoffer, MD, Santa Mon ica, CA
Douglas Koch, MD, Hou ston, TX
Gabor Koranyi, MD, Vâxjô, Sweden
Scott McClatchey, CAPT, MC, USN, MD, San Diego, CA
Joh n Moran, MD, Hou ston, TX
Sverker Norrby, PhD, Gron ingen, Holland
Thom as Olsen, MD, Århu s, Den m ark
Giacomo Savin i, MD, Bologna, Italy
H. Joh n Sham m as, MD, Lynwood, CA

Honorary M embers: Svyataslav Fyodorov, MD, Moscow, Ru ssia (deceased)

Herm an n Gernet, MD, Wü rzbu rg, Germ any
Joh n Retzlaff, Med ford, OR
Rob G.L. van der Heijde, PhD, Am sterd am, Holland

Office rs
2005
Presid ent: Ken neth J Hoffer, MD
Vice-Presid ent: Jaim e Aram berri, MD
Secretary: Thom as Olsen, MD
Treasu rer: Wolfgang H aigis, MS, PhD
EC Mem bers: Sverker Norrby, PhD, H Joh n Sham m as, MD

247
Hoffer KJ. IOL Power (pp. 247-248).
© 2011 SLACK Incorporated.
248 Appen d ix A

2007
Presid ent: Jaim e Aram berri, MD
Vice-Presid ent: Ken neth J. Hoffer, MD
Secretary: Thom as Olsen, MD
Treasu rer: Wolfgang H aigis, MS, PhD
EC Mem bers: Sverker Norrby, PhD, H. Joh n Sham m as, MD

2009
Presid ent: H. Joh n Sham m as, MD
Vice-Presid ent: Sverker Norrby, PhD
Secretary: Thom as Olsen, MD
Treasu rer: Wolfgang H aigis, MS, PhD
EC Mem bers, Past-Presidents: Ken neth J. Hoffer, MD, Jaim e Aram berri, MD
 Fin a n cia l Disclo s u re s

Jaime A ramberri, M D was previou sly a consu ltant for CSO In stru ments, Italy.

Wolfgang Haigis, M S, PhD is a consu ltant for Carl Zeiss Med itec.

Kenneth J. Hoffer, M D has no financial or proprietary interest in the m aterials presented

herein.

Tsontcho [Sean] Ianchulev, M D, M PH is the Ch ief Med ical Officer for Tran scend Med ical,
is a consu ltant for Wavetec Vision, is the inventor and patent holder for Intraoperative
Autorefraction Method s, is a Clin ical Assistant Professor at UCSF, and is a ventu re
partner at Tu llis Health Investors.

Susana M arcos, PhD, FOSA , FEOS has no financial or proprietary interest in the m ateri-
als presented herein.

Scott K. M cClatchey, CA PT, M C, USN, M D has no financial or proprietary interest in the

m aterials presented herein.

John R. M oran, PhD, M D is a consu ltant for STAAR Su rgical Co and AMO.

Sverker N orrby, PhD is a con su ltant for AMO.

Thomas Olsen, M D is a share holder of IOL In novation s (w w w.phacooptics.com).

Karl Ossoinig, M D is a consu ltant for Qu antel Med ical.

Thomas C. Prager, PhD, M PH is a consu ltant for ESI (Min neapolis, MN ) and Qu antel
Med ical (Bozem an, MT), as well as a patent holder for the Prager shell.

Giacomo Savini, M D has no financial or proprietary interest in the m aterials presented

herein.

H. John Shammas, M D has no financial or proprietary interest in the m aterials presented

herein.

249