Eximer BL Osu1086104689

THE ASSESSMENT OF POSTOPERATIVE REFRACTIVE SURGERY
PATIENTS IN CLINICAL RESEARCH
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Melissa Diane Bailey, OD, MS
*****
The Ohio State University

2004
Dissertation Committee:
Approved by
Professor Karla Zadnik, Advisor
Professor Joseph A. Barr
Associate Professor Mark A. Bullimore

Advisor
Vision Science Graduate Program
ABSTRACT
We examined existing data from Food and Drug Administration (FDA) clinical
trials for approved ophthalmic lasers to determine if trends for improvement in outcomes
occurred with changes in laser technology. The repeatability of autorefraction
measurements, visual acuity measurements, and the National Eye Institute Refractive
Error Quality of Life (NEI-RQL-42) Questionnaire was evaluated in laser in situ
keratomileusis (LASIK) patients. We also compared the performance of the NEI-RQL-42
to another survey for refractive surgery patients to see if similar results were obtained
when subjects completed the surveys simultaneously. Finally, we evaluated patient
survey responses and opinions of eye care practitioners to provide direction for
generating hypotheses for future research studies.
We found that cycloplegic autorefraction measurements with the Grand Seiko
autorefractor were repeatable and agreed well with subjective refraction. The NEI-RQL-
42 was determined to have an acceptable level of repeatability and produced results that
were similar to other studies that tested refractive surgery patients. Measurements of low-
contrast visual acuity under various lighting conditions were also found to have
acceptable levels of repeatability, and 5% contrast visual acuity appears to be sensitive to
differences between LASIK and non-LASIK patients. When evaluating reasons why
patients would recommend LASIK to others, we determined that the primary reason was
ii
to eliminate the need for glasses and/or contact lenses. Focus group sessions with eye
care practitioners revealed that practitioners would like to have more information about
the relationship between patient personality and psychological issues and satisfaction
with the results of the procedure. The eye care practitioners defined a successful LASIK
procedure as one in which the patients had good uncorrected visual acuity, refractive
error within ± 0.50 D of the intended correction, and no new or increase in night vision or
dry eye symptoms.
Future studies in corneal refractive surgery patients should be able to use the
measurement techniques evaluated in these studies to answer the following two research
questions: What factors are associated with the dependence on glasses and/or contact
lenses five years after LASIK, and what psychological factors are associated with the
patient’s view of a LASIK procedure?
iii
Dedicated to Daniel Brian Bailey, my husband, and my family for their support. I would
also like to dedicate this to my mentor for teaching me how to conduct high quality
research while balancing a personal and professional life, and for helping me find a job I
love.
iv
ACKNOWLEDGEMENTS
I would like to acknowledge my co-authors of the studies contained in this dissertation:
Karla Zadnik, OD, PhD Deepinder K. Dhaliwal, MD
Michael D. Twa, OD, MS Michael D. Olson, OD, PhD
Mark A. Bullimore, MCOptom, PhD Joseph Shovlin, OD
G. Lynn Mitchell, MAS Stephen Pascucci, MD
Lisa A. Jones, PhD Brian Boxer-Wachler, MD
Timothy McMahon, OD
NIH/NEI grant #T32-EY013359 and #R24-EY14792, American Optometric Foundation
William C. Ezell Fellowships (CIBA Vision and the AAO Section on Cornea and Contact
Lenses), Ohio Lions Eye Research Foundation. I would also like to acknowledge Donald
O. Mutti, OD, PhD and Charles Campbell, PhD for their assistance in determining the
design principles and infrared beam diameter for each autorefractor used in this study.
v
VITA
March 1, 1975 Born – Gillette, Wyoming
1997 B.S. Neuroscience, Muskingum College
2001 O.D. The Ohio State University
2001 M.S. Vision Science, The Ohio State University
2001 Graduate Teaching Associate, The Ohio State University
2001- present Postdoctoral Fellow, The Ohio State University
PUBLICATIONS
1. Bailey MD, Olson MD, Bullimore MA, Jones LA, Maloney RK. (2004) The effect of
LASIK on best-corrected high and low contrast visual acuity. Optometry and Vision
Science, 81:362-368.
2. Bailey MD, Mitchell GL, Dhaliwal DK, Boxer Wachler BS, Olson MD, Shovlin JP,
Pascucci SE, Zadnik K. (2004) Reasons why patients recommend laser in situ
keratomileusis to others. Journal of Cataract and Refractive Surgery, in press.
3. Twa MD, Bailey MD, Hayes J, Bullimore MA. (2004) Estimation of pupil size by
digital photography. Journal of Cataract and Refractive Surgery, 30: 381-389.
4. Bailey MD, Mitchell GL, Dhaliwal DK, Zadnik K. (2003) Patient satisfaction and
night vision symptoms after laser in situ keratomileusis. Ophthalmology, 110:1371-
1378.
5. Rah MJ, Barr JT, Bailey MD. (2002) Corneal pigmentation in overnight
orthokeratology: a case series. Optometry, 73:425-434.
6. Rah MJ, Jackson JM, Jones LA, Marsden HJ, Bailey MD, Barr JT. (2002) Overnight
orthokeratology: preliminary results of the lenses and overnight orthokeratology
(LOOK) study. Optometry and Vision Science, 79:598-605.
vi
7. Bailey MD, Walline JJ, Mitchell GL, Zadnik K. (2001) Visual acuity in contact lens
wearers. Optometry and Vision Science, 78:726-731.
8. Walline JJ, Bailey MD, Zadnik K. (2000) Vision-specific quality of life and modes of
refractive error correction. Optometry and Vision Science, 77:648-652.
FIELDS OF STUDY
Major Field: Vision Science
vii
TABLE OF CONTENTS
Page
Abstract ............................................................................................................................... ii
Acknowledgements..............................................................................................................v
Vita..................................................................................................................................... vi
List of Tables .......................................................................................................................x
List of Figures .................................................................................................................. xii
Chapters:
1. Introductions ..................................................................................................................1
1.1 Common unsuccessful LASIK outcomes ..............................................1
1.2 Selection of refractive error correction for individual patients..............7
1.3 Future studies of LASIK patients ........................................................11
2. Trends in outcomes for LASIK for myopia in FDA approved lasers..........................13
2.1 Methods................................................................................................14
Data source.....................................................................................14
Data entry and analysis ..................................................................17
2.2 Results..................................................................................................18
Summary of all studies...................................................................18
Trends in visual acuity and refractive error outcomes...................18
2.3 Conclusions..........................................................................................29
3. Defining a successful LASIK procedure .....................................................................33
3.1 Methods................................................................................................33
Sample............................................................................................33
Data collection ...............................................................................34
Data analysis ..................................................................................35
3.2 Results..................................................................................................35
Primary and secondary themes ......................................................36
4. Reasons why patients recommend laser in situ keratomileusis to others ....................42
4.1 Methods................................................................................................43
Subjects ..........................................................................................43
Informed consent ...........................................................................44
Questionnaire .................................................................................45
Data collection and analysis...........................................................46
4.2 Results..................................................................................................48
Response rate .................................................................................48
Subject demographics ....................................................................48
viii
Reasons for recommending LASIK...............................................49
Dissatisfaction................................................................................49
5. The repeatability of autorefraction and axial length measurements after LASIK .......54
5.1 Materials and methods .........................................................................55
Subjects ..........................................................................................55
Refractive error measurements ......................................................56
IOLMaster measurements..............................................................57
Analysis of refractive error data- conversion to power vector
representation.................................................................................58
Statistical methods .........................................................................59
Measurement of autorefractor infrared beam ................................60
5.2 Results..................................................................................................61
Repeatability ..................................................................................61
Agreement between subjective refraction and autorefraction .......61
Axial length....................................................................................70
5.3 Discussion ............................................................................................70
6. Clinical evaluation of the NEI-RQL-42 in LASIK patients ........................................75
6.1 Methods................................................................................................77
Study 1: Repeatability of the NEI-RQL-42 in LASIK patients.....77
Study 2: Comparison of the NEI-RQL-42 and QOV.....................78
Subjects and informed consent ..........................................78
Questionnaires....................................................................78
Data analysis ......................................................................79
6.2 Results..................................................................................................79
Repeatability of the NEI-RQL-42..................................................79
7. Assessment of visual function in patients after LASIK...............................................93
Subjects ..........................................................................................96
Refractive error measurements ......................................................96
IOLMaster measurements..............................................................97
7.2 Results..................................................................................................98
Repeatability ..................................................................................99
Comparisons of visual acuity between groups.............................103
8. Conclusions................................................................................................................112
List of references..............................................................................................................115
ix
LIST OF TABLES
Table Page
1.1 Previous studies of low-contrast visual acuity and contrast sensitivity in refractive
surgery.........................................................................................................................4
1.2 Summary of literature review for factors associated with unsuccessful outcomes ....8
1.3 Refractive surgery recommendations: 2002 ASCRS Survey .....................................9
1.4 Refractive surgery recommendations: 2002 ISRS Survey........................................10
2.1 Excimer lasers approved by the Food and Drug Administration for laser in situ
keratomileusis ...........................................................................................................15
2.2 List of variables recorded from summaries for data analysis ...................................16
2.3 Proportion of all patients in each category of laser with postoperative symptoms ..27
2.4 Proportion of low myopes in each category of laser with postoperative

symptoms ..................................................................................................................28
3.1 Desirable and undesirable outcomes emerging from focus group session ...............38
4.1 Reasons for recommending LASIK to a friend ........................................................50
5.1 Bias and limits of agreement for each measure of refractive error and axial
length.........................................................................................................................62
5.2 Mean and standard deviation of refractive error measurements...............................63
5.3 Agreement between subjective refraction measurements and autorefraction

measurements............................................................................................................64
6.1 Repeatability of the NEI-RQL-42 subscales across two visits .................................78
6.2 Mean scores for each of the NEI-RQL-42 and QOV subscales at each of the study
visits ..........................................................................................................................82
x
6.3 Mean scores for each of the QOV questionnaire subscales at each of the study
visits ..........................................................................................................................83
6.4 Tukey adjusted post hoc comparisons of the repeated measures analysis of
variance test for each subscale..................................................................................84
6.5 Comparison of the repeatability of the National Eye Institute Refractive Error
Quality of Life (NEI-RQL-42) Questionnaire across studies ...................................89
6.6 Comparison of preoperative and postoperative NEI-RQL-42 scores across

studies .......................................................................................................................90
7.1 Contrast sensitivity from a report by Lee, Hu and Wang (2003)..............................92
7.2 Descriptive statistics by group ..................................................................................96
7.3 Visual acuity scores for each level of contrast and lighting level by group,
examiner, and visit ....................................................................................................97
7.4 Within examiner test-retest repeatability ..................................................................98
7.5 Between examiner test-retest repeatability ...............................................................99
7.6 Difference between LASIK and Non-LASIK patient visual acuity scores ............100
7.7 Relationship between mean visual acuity scores (across light levels) and axial
length.......................................................................................................................101
xi
LIST OF FIGURES
Figure Page
1.1 Associations of visual function after LASIK..............................................................2
2.1 Percentages of low myopes versus all patients with Snellen visual acuity of 20/20
or better for each laser type.......................................................................................19
2.2 Percentages of low myopes versus all patients with Snellen visual acuity of 20/40
or better for each laser type.......................................................................................20
2.3 Percentages of all patients, low myopes, and high myopes with spherical
equivalent refractive error within 0.50 D of intended correction for each laser
type............................................................................................................................21
2.4 Percentages of all patients, low myopes, and high myopes with spherical
equivalent refractive error within 1.00 D of intended correction for each laser
type............................................................................................................................22
2.5 Percentages of all patients, patients with spherical myopia only, and patients with
myopic astigmatism with Snellen visual acuity of 20/20 or better for each laser
type............................................................................................................................23
myopic astigmatism with Snellen visual acuity of 20/40 or better for each laser
Type ..........................................................................................................................24
myopic astigmatism with spherical equivalent refractive error within 0.50 D of
intended correction....................................................................................................25
myopic astigmatism with spherical equivalent refractive error within 0.50 D of
intended correction....................................................................................................26
3.1 Photograph of flip chart for participants’ comments during discussion of a

successful “cut-point” for each measurement of success .........................................40
xii
5.1 Difference versus mean plot for the agreement in spherical equivalent between
Grand Seiko autorefraction and subjective refraction without cycloplegia..............66
Grand Seiko autorefraction and subjective refraction with cycloplegia...................67
Humphrey autorefraction and subjective refraction without cycloplegia.................68
Humphrey autorefraction and subjective refraction with cycloplegia......................69
6.1 Internal reliability, Cronbach’s Alpha, of the individual items for each subscale
of the NEI-RQL-42 and QOV at the preoperative visit............................................80
7.1 Mean axial length versus difference between pairs at 100% contrast in dim
lighting conditions ..................................................................................................102
7.4 Mean axial length versus difference between pairs at 2.5% contrast in dim
xiii
CHAPTER 1
INTRODUCTION
Like many segments of health care, the area of refractive error correction has
experienced an explosion of technological advances in the last few decades. So many
different types of glasses, contact lenses, and refractive surgery procedures are currently
available that an individual patient may have to choose among dozens of options for
refractive error correction. This has led to many dilemmas for the eye care practitioner
and patient. What are the rates of unsuccessful outcomes associated with each of the
options available? How does one counsel an individual patient as to which mode of
refractive error correction will provide the best quality of vision for him or her? Which
option will provide the best quality of life? These questions are especially important
when one considers the most common corneal refractive surgery, laser in situ
keratomileusis (LASIK), because this particular option is irreversible. The purpose of this
dissertation is to address the above questions as they relate to LASIK outcomes and to the
design of future studies of LASIK outcomes.
1.1 Common unsuccessful LASIK outcomes
Previous studies overwhelmingly support two conclusions about the LASIK
procedure when it is used to correct myopia:
1
• The vast majority of LASIK patients (over 90%) can achieve good uncorrected
and best-corrected, high-contrast visual acuity after LASIK.1
• Patient reported satisfaction is generally high, ranging from 82 to 98%.2-5
Nonetheless, some patients have subjective complaints of decreased vision after
LASIK. The subjective complaints range from decreased clarity of vision to difficulties
with night driving caused by symptoms such as glare, halos, and starbursts.
Consequently, studies of both (photorefractive keratectomy) PRK and LASIK have
shown that losses of low-contrast visual acuity can occur, even in patients with normal
high-contrast visual acuity.6, 7 Figure 1 is an outline of a series of relationships. These
relationships summarize studies of several outcomes after LASIK-induced changes in
optical aberrations, contrast sensitivity, and patient reports of visual function and
satisfaction.
Increased Ocular
Aberrations
Previous studies of low-
1
contrast visual acuity or
Decreased Contrast
contrast sensitivity after Sensitivity
LASIK have not found similar 2 ?
results. Some studies have Night Vision

Symptoms
found only transient losses of
3
low-contrast visual acuity or
Decreased
contrast sensitivity that resolve Satisfaction
after a period of time,8, 9 and
others have reported a Figure 1.1 Associations of visual function after
LASIK
2
persistent decrease in visual performance, but only in dark conditions or in higher
myopes.7, 10 It appears that discrepancies among previous studies are due to
methodological differences, primarily differences in the luminance levels used during
data collection (Table 1.1). Losses of contrast sensitivity after corneal refractive surgery
have been reported in multiple studies under low luminance or dilated pupil conditions.
This is consistent with patient complaints of vision problems at night.
Reports of changes in contrast sensitivity following corneal refractive surgery
have fueled an interest in how optical aberrations might also change after these
procedures. Optical aberrations increase after both PRK and LASIK.11-13 This increase in
optical aberrations correlates with losses of low-contrast visual acuity and accounts for
most of the decrease in contrast sensitivity after LASIK (Figure 1, association #1).11, 13
Certain values of preoperative corneal shape may help researchers predict loss of
visual function after LASIK. Several factors correspond with either the increase in ocular
aberrations or the loss of low-contrast visual acuity after LASIK. The natural aspheric
shape of the cornea (prolate, a flattening ellipse) is known to partially reduce the optical
aberrations of the eye.14 It has been demonstrated that corneas become more oblate (i.e.,
less prolate) in shape after myopic LASIK.10 Increasing values of asphericity (more
oblate) have also been shown in calculated models to increase the optical aberrations of
the eye.14 Thus, a cornea that is already flatter than average or less prolate prior to LASIK
may be more likely to experience increased ocular optical aberrations (and thus glare,
halos, and starbursts) postoperatively. Larger pupil sizes and decentered laser ablation
zones are also associated with increased ocular aberrations/decreased low-contrast visual
acuity.6, 15
3
Testing Surgical Testing Time From
Method Procedure Conditions Surgery Results Authors
LCVA PRK Normal and 3 to 34 Loss of LCVA at both Gauthier et al. 16
low luminance months luminance levels*
Dilated and 12 months Persistent loss of LCVA, Bullimore et al.17
undilated with losses greater
pupils under dilated
conditions*
Dilated and 12 months Loss of LCVA with Verdon et al.6
undilated natural and dilated
pupils pupils*
LASIK Normal 3 months Loss of LCVA in higher Nakamura et al.7
luminance only myopes only
Normal and 6 months Persistent loss of LCVA Holladay et al.10
low luminance under low luminance*
Low luminance 6 months Persistent loss of LCVA, Bailey et al.18
especially in higher
myopes
CS PRK Normal and 6 months Persistent loss of CS Montes-Mico et
low luminance under low luminance* al.19
LASIK Normal 3 months Loss of CS at 12 Mutyala et al.8
luminance only cycles/degree, although
within “normal” range
compared to a historical
control group
Normal 6 months No persistent loss of CS Perez-Santonja et
luminance only al.9
Low 6 months Persistent loss of CS at Anera at al.20
Luminance higher spatial
only frequencies in higher
myopes only
*Denotes studies where losses of LCVA or CS are reported. These are the same studies that also
included low-luminance or dilated conditions. LCVA= low contrast visual acuity, CS=contrast
sensitivity
Table 1.1 Previous studies of low contrast visual acuity and contrast sensitivity in
refractive surgery
4
The bottom half of Figure 1 depicts what the literature suggests about the
patient’s perspective of visual function after LASIK. Although satisfaction is generally
high, ranging from 82 to 98%,2-5 reports of night vision symptoms still occur frequently.
Rates for night vision symptoms range from 12% to 57%.3, 21 Night vision symptoms can
affect driving after refractive surgery. In a study of PRK and LASIK patients, 29.5% of
subjects experienced a worsening of their driving subscale score when compared to
preoperative subscale scores on the Refractive Status and Vision Profile (RSVP)
questionnaire.22
In a previous study by our group, we provided evidence that postoperative night
vision symptoms are associated with decreased postoperative satisfaction (Figure 1,
association #3).23 Two recent studies suggest that changes in objective visual function,
increased optical aberrations and reduced contrast sensitivity, are associated with the
subjective night vision symptoms that have been reported after LASIK (Figure 1,
association #2).24, 25
Obviously, objective and subjective visual quality are important considerations
when choosing a mode of refractive error correction; however, convenience/dependence
and comfort are also considerations. If a mode of refractive error correction is too
uncomfortable and/or inconvenient, patients will probably consider that mode of
correction unsuccessful for them. Counseling prospective LASIK patients about the long-
term convenience and comfort is somewhat problematic because the vast majority of
LASIK studies have only followed patients for six months or less. With so little
information on the long-term effects of LASIK, it is difficult to estimate the number of
subjects who remain truly free of optical correction for distance targets long after their
5
refractive surgery. The proportion of subjects who remain free of optical correction has
rarely been reported, let alone used as a primary outcome measure in previous studies of
LASIK. One recent report on patients with six years of follow-up after LASIK suggests
that a fairly significant percentage of subjects required some use of glasses or contact
lenses postoperatively, as 29% of the subjects were unhappy with their uncorrected visual
acuity.26
When considering the aspect of comfort after LASIK, one must consider the
considerable body of work assessing dry eye after LASIK. One explanation for dry eye
after LASIK is a simple consideration of corneal physiology. In the human eye, basal
tearing occurs when environmental factors stimulate the corneal surface.27 When lamellar
flaps are cut with a microkeratome during LASIK, many of the nerve fibers that innervate
the cornea are severed. Neural transmission of the stimulus for basal tearing might be
interrupted at least temporarily when the corneal nerves are severed during LASIK.
Support for this conclusion is provided by studies that have found a decrease in corneal
sensitivity following LASIK, which may persist for as long as 16 months
postoperatively.28-30 A recent study suggests that preoperative dry eye may be a risk
factor for severe post-LASIK dry eye and that decreased corneal sensitivity may persist
longer in those patients who had preoperative dry eye.31
6
Dry eye symptoms are significantly increased following LASIK, and may persist
for as long as 16 months.28 Symptoms of dry eye may or may not affect postoperative
satisfaction. Hovanesian and co-workers found that postoperative dry eye significantly
affected subject satisfaction, while Toda and co-workers report no significant effect of
dry eye on satisfaction.31, 32 Changes in the tear film may also be associated with glare
disability;33 however, this effect has not been evaluated in LASIK-related dry eye
patients.
1.2 Selection of refractive error correction for individual patients
Beyond informing patients of the risks associated with having LASIK, it is more
clinically useful to be able to predict which patients are more likely to have an
unsuccessful result. Although studies attempting to address this issue do not necessarily
provide the level of evidence required to truly predict outcomes, multiple studies suggest
some important and consistent associations. When reviewing the literature for factors that
are associated with various unsuccessful outcomes, i.e. night vision symptoms, decreased
low-contrast visual acuity/contrast sensitivity, under-correction, dry eye, enhancement,
regression, and insufficient fixation of the microkeratome, it is somewhat surprising to
find that some of the same factors are associated with more than one of these events
(Table 1.2); for example, the higher levels preoperative myopia and flatter preoperative
corneal curvature are both associated with multiple unsuccessful outcomes.
Although there are eight different factors listed in Table 2, there are four basic
categories of factors that seem to be associated with unsuccessful outcomes: preoperative
refractive error (myopia, astigmatism, and required ablation depth), corneal shape
7
Unsuccessful Outcome
Decreased Low-
Contrast Visual Insufficient
Associated Night Vision Acuity/Contrast Under- Dry Fixation of the
Factor Symptoms Sensitivity correction Eye Enhancement Regression Microkeratome
Larger
amounts of 9 9 9 9 9
myopia
Increasing age
9 9
Increased
9 9
ablation depth
Increased
9
corneal toricity
Flatter corneal
9 9 9
8
curvature
Larger
amounts of 9 9
astigmatism
Female
9
Table 1.2. Summary of Literature Review for Factors Associated with Unsuccessful Outcomes
Recommendations (% of respondents)
No Phakic CLE with
Example Patient surgery RK PRK LASIK LASEK ICR IOL IOL Waiting
30-year-old 9% 1% 4% 79% 3% 0% NA NA 3%
–3.00 D myope
30-year-old 3% 0% 1% 86% 4% 0% 0.6% NA 4%

–7.00 D myope
30-year-old 13% 0% 2% 22% NA 0.5% 15% 7% 38%
–12.00 D myope
No Phakic CLE with
surgery PRK LASIK LASEK LTK CK IOL IOL Waiting
45-year-old 27% 1% 48% 2% 3% 8% 0.4% 1%* 10%
9
+1.00 D hyperope
45-year-old 9% 1% 62% 2% 2% 3% 1% 7% 12%

+3.00 D hyperope
45-year-old 17% 0.4% 17% 0.5% 1% 0.4% 5% 27% 32%
+5.00 D hyperope
†
Table created from data in text of Leaming DV. Practice styles and preferences of ASCRS members-2002 survey.
J Cataract Refract Surg 2003;29:1412-20. RK = radial keratotomy, PRK = photorefractive keratectomy, LASIK =
laser in situ keratomileusis, LASEK = subepithelial keratomileusis, ICR = intrastromal corneal rings, IOL =
intraocular lens, CLE = clear lens extraction
Table 1.3. Refractive surgery recommendations: 2002 ASCRS Survey†

Recommendations (% of respondents)
No Phakic
Example Patient surgery RK PRK LASIK LASEK ICR IOL CLE Waiting
30-year-old NA 1% 1% 97% NA 1% 0% 0% 0%
–3.00 D myope
30-year-old NA NA 0% 96% MA MA 2% 0% 0%
–7.00 D myope
30-year-old NA NA 2% 36% 4% NA 22% 5% 31%
–12.00 D myope
No Phakic
surgery PRK LASIK LASEK LTK CK IOL CLE Waiting
45-year-old NA 2% 80% 2% 3% 8% NA 3% 3%
10
+1.00 D hyperope
45-year-old NA 1% 93% 2% 0% 0% 0% 3% 1%
+3.00 D hyperope
45-year-old NA 0% 33% 2% NA NA NA 33% 33%
+5.00 D hyperope
†
Table created from data in text of Duffey RJ, Leaming D. US trends in refractive surgery: 2002 ISRS survey. J
Refract Surg 2003;19:357-63. RK = radial keratotomy, PRK = photorefractive keratectomy, LASIK = laser in situ
keratomileusis, LASEK = subepithelial keratomileusis, ICR = intrastromal corneal rings, IOL = intraocular lens,
CLE = clear lens extraction, LTK = laser thermokeratoplasty, CK= conductive keratoplasty
Table 1.4. Refractive surgery recommendations: 2002 ISRS Survey†

(curvature, asphericity, and toricity), age, and gender. In addition, there is likely to be an
interaction between preoperative refractive error and corneal shape, as patients with
higher levels of refractive error and astigmatism require deeper and/or asymmetrically
shaped ablation profiles that results in a very different postoperative shape. It is then very
easy how patients who possess an interaction of effects between higher levels. of myopia
and flatter or less prolate corneal curvatures may have an increased susceptibility to fall
into the cascade of observations outlined in Figure 1
Even though strong evidence for predicting LASIK outcomes may not be
available yet, practitioners must still make recommendations to patients regarding corneal
refractive surgery on a daily basis. In a report by the American Society of Cataract and
Refractive Surgery (ASCRS), Practice Styles and Preferences of ASCRS Members– 2002
Survey, the investigators report on the recommendations study participants would make
for various types of patients (Table 1.3).34 A similar study was also conducted by
surveying members of the International Society Refractive Surgery (ISRS) (Table 1.4).35
In both surveys, LASIK is the surgical procedure that is most often recommended to all
types of patients. The exceptions to this rule are patients with higher levels of myopia and
hyperopia.
1.3 Future studies of LASIK patients
Based on the many retrospective and small prospective studies evaluating LASIK
outcomes, one could easily generate hypotheses for studies that would provide for the
predictive information clinicians need to confidently and accurately select candidates for
LASIK. When one begins to design these studies, however, many problems arise. In
11
order to conduct a prospective study of LASIK patients, researchers must be able to
accurately measure LASIK patients and must be confident that constantly emerging
changes in technology will not make the results of these studies obsolete. In the
remaining chapters, we address the many challenges of measuring corneal refractive
surgery patients, including problems with autorefraction and visual function
measurements, as well as evaluating evidence related to technological advances. This
body of evidence should then allow researchers to answer questions related to the
predicting LASIK outcomes.
12
CHAPTER 2
TRENDS IN OUTCOMES FOR LASIK FOR MYOPIA IN FDA APPROVED LASERS
Many of the companies manufacturing ophthalmic lasers have implemented rapid
technological changes in laser design and function. The reports supplied by the laser
companies often suggest that these technological advances have dramatically improved
the outcomes of refractive surgical procedures such as laser in situ keratomileusis
(LASIK) and photorefractive keratectomy (PRK). Such technological advances include
devices to track the eye movements of the patient, and lasers that attempt to reduce, or at
least not increase, whole eye optical aberrations by performing ablations that are guided
by an aberrometer. Such rapid changes in technology make studying the outcomes of
corneal refractive surgery difficult, because results of studies with one generation of laser
are often considered obsolete when newer generations of lasers are marketed.
Nonetheless, there are very few studies that directly compare types of lasers to determine
if technological advances have made measurable improvements in refractive surgery
outcomes.36-39 Summaries of the safety and efficacy data for each of the lasers approved
by the US Food and Drug Administration (FDA) for use in LASIK are readily available
on the internet; therefore, we integrated these data sets in order to summarize expected
outcomes of LASIK eye surgery and to address the following hypothesis: Visual acuity,
13
refractive error, and postoperative symptoms have improved as ophthalmic laser
technology has evolved from broad beam design to wavefront-guided.
2.1 METHODS
Data source
The Summary of Safety and Effectiveness Data for each of the ophthalmic
excimer laser systems approved by the FDA for LASIK were downloaded from the
FDA’s webpage: http://www.fda.gov/cdrh/lasik/laser.htm. For simplification during data
presentation and analysis, each device was assigned a number 1 through 10 in order of
approval (Table 2.1). Variables were recorded for five categories of patients, i.e., the
entire cohort, high myopes, low myopes, patients receiving spherical only ablations, and
patients receiving spherocylindrical ablations. Data analyses were completed for each of
the five groups separately.
The summary reports for laser numbers 1, 4, and 6 categorized high myopes as
those patients with corrections equal to or more myopic than –7.00 D and low myopes as
those patients with corrections less myopic than –7.00 D. The summary reports for laser
numbers 2, 3, 5, and 7 categorized high myopes as more myopic than –7.00 D and low
myopes as equal to or less myopic than –7.00 D. At the time of data analysis, data for low
myopes only were available for lasers 8, 9, and 10; thus, data from these three lasers were
only included in analyses related to low myopes. For laser number 7, it was not possible
to make the division between high and low myopes at –7.00 D for the following
variables, UCVA
14
Date of Date of
Excimer laser trade FDA summary
Number name Applicant’s name Approval download
1 Kremer Excimer Photomed, Inc 07/30/1998 04/30/2003
Laser
2 SVS Apex Plus Summit 10/21/1999 04/30/2003
Excimer Laser Technology, Inc
Workstation
3 VISX STAR S2 CRS Clinical 11/19/1999 04/30/2003
Excimer Laser Research, Inc
System
4 Dishler Excimer Jon G. Dishler, MD, 12/16/1999 04/30/2003
Laser System FACS
5 TECHNOLAS 217A Bausch & Lomb 02/23/2000 04/30/2003
Excimer Laser Surgical, Inc.
System
6 LADARVision® Autonomous 05/09/2000 04/30/2003
Excimer Laser Technologies
System Corporation
7 Nidek EC-5000 Nidek 04/19/2000 04/30/2003
Excimer Laser Technologies, Inc.
System
8 LaserSight LaserSight 09/28/2001 04/30/2003
LaserScan LSX Technologies, Inc.
Excimer Laser
System
9 LADARVision 4000 Alcon 10/18/2002 04/30/2003
Excimer Laser
System
10 VISX STAR S4 VISX, Incorporated 05/23/2003
Excimer Laser
System and
WaveScan
WaveFront System
11 WaveLight SurgiVision 10/7/2003
ALLEGRETTO Refractive
WAVE Excimer Consultants, LLC
Laser System
Table 2.1. Excimer lasers approved by the Food and Drug Administration for laser in situ
keratomileusis (LASIK)
15
20/20 or better, UCVA 20/40 or better, BSCVA worse than 20/40, loss of greater than
two lines of BSCVA. These variables were reported in categories, one of which grouped
patients between –6.00 D and –7.99D in one group. We had no way of determining how
many of the patients would have been categorized at or above –7.00 D; therefore, high
myopes were categorized as all patients with spherical equivalent refractive error equal to
more myopic than –6.00 D. Low myopes were categorized as patients with spherical
equivalent refractive error less myopic than –6.00 D.
Data regarding patient symptoms, i.e., glare, dryness, etc., were included when
reported in individual summaries. The exact wording of the questions used to survey
patients was not available for any of the lasers. To complicate matters further, some
Variable Categories
Refractive error
Visual acuity outcomes outcomes Symptom outcomes
UCVA of 20/20 or better Within ± 0.50 D of Glare
intended correction
UCVA of 20/40 or better Within ± 1.00 D of Halos
intended correction
Loss of > 2 lines of Within ± 2.00 D of Night driving problems
BSCVA intended correction
BSCVA worse than 20/40 Change in spherical Dryness
equivalent of < 1.00 D
from 3 to 6 months
BSCVA worse than 20/25 Increase of > 2.00 D

cylinder
†
Individual variables all refer to the proportion of patients with the given outcome.
Table 2.2. List of variables recorded from summaries for data analysis.†
16
lasers asked for patients to compare their current symptoms to symptoms before surgery
and other lasers just queried patients about the severity of their postoperative symptoms.
For the purpose of these data analyses, the proportion of subjects for each laser who
experienced glare, halos, night driving problems, or dryness was defined as the number of
subjects who said these symptoms were either worse than before surgery, much worse
than before surgery, moderately severe, or severe. All data presented are six-month
outcomes.
Data entry and analysis
Data from each summary were entered into a Microsoft Excel spreadsheet. All
variables (Table 2.2) were extracted from the summaries as proportions, recording both
the numerator and denominator for data analysis. The Cochran-Armitage test for trend
was used to compare laser types (broad beam, scanning, scanning with eye tracker,
wavefront guided) for each of the variables. The analyses were performed for all five
categories of patients (the entire cohort, high myopes, low myopes, patients receiving
spherical only ablations, and patients receiving spherocylindrical ablations). The
Cochran-Armitage test for trend was chosen because the data were only available in the
form of binomial proportions.40 The null hypothesis for this test is that no trend exists
across levels of the explanatory variable.
17
2.2 RESULTS
Summary of all studies
The mean age of subjects across all FDA studies of LASIK ranged from 36 to 43
years old. A total of 8519 eyes were initially enrolled in these studies, with data reported
for 6596 (77%) eyes at 6 months after the initial LASIK procedure. Fifty-three percent
(3878/7282) of the subjects were women. Uncorrected visual acuity of 20/20 or better
was achieved by 60% (3402/5685) of eyes, while 93% (5263/5685) of eyes achieved
uncorrected visual acuity of 20/40 or better. Postoperative spherical equivalent refractive
error was within ± 0.50 D of the intended correction for 71% (4348/6120) of eyes and
was within ± 1.00D of the intended correction for 89% (5467/6120) of eyes. Overall,
21% (700/3396) patients experienced glare, 18% (617/3475) of patients experienced
halos, 21% (608/2877) of patients reported night vision problems, and 19% (244/1305) of
patients experienced symptoms of dryness.
Trends in visual acuity and refractive error outcomes
For the percentage of patients with uncorrected visual acuity of 20/20 or better,
trends for improvement were found for all five groups of patients (low myopes, high
myopes, patients with spherical myopia only, patients with myopic astigmatism, and the
entire cohort) (Figures 2.1 and 2.4). When looking at the percentage of patients who were
20/40 or better, however, the trend for improvement in this outcome was only present for
the analyses including low myopes, patients with spherical myopia only, patients with
myopic astigmatism, and the entire cohort (Figures 2.2 and 2.5). There was a statistically
significant trend across laser types for all groups except high myopes and patients with
18
100 All Eyes
Percent with UCVA of 20/20 or Better

90 High Myopes
Low Myopes
80
70
60
(Snellen)
50
40
30
20
10
0
Broad Beam Scanning Scanning with Wavefront-
Eye Tracker guided
Laser Type
Figure 2.1 Percentages of Low Myopes versus All Patients with Snellen Visual
Acuity of 20/20 or Better for Each Laser Type. There was a statistically significant
trend across groups for all eyes (Z = 20.76, p < 0.0001), low myopes (Z = 16.00,
p < 0.0001), and high myopes (Z = 7.89, p < 0.0001).
19
All Eyes
High Myopes
Percent with UCVA of 20/40 or Better (Snellen) 100 Low Myopes
90
80
70
60
50
40
30
20
10
0
Broad Beam Scanning Scanning with Eye Wavefront-guided
Tracker
Laser Type
Figure 2.2 Percentages of All Patients, Low Myopes, and High Myopes with
Snellen Visual Acuity of 20/40 or better for Each Laser Type. There was a
statistically significant trend across groups for all eyes (Z = 6.39, p < 0.0001) and
low myopes (Z = 4.08, p < 0.0001). No significant trend was found for high
myopes (Z = 2.55, p = 0.01) after adjusting for multiple comparisons.
20
All Eyes
100
High Myopes
Correction (Spherical Equivalent)

Percent Within 0.50 D of Intended
90 Low Myopes
80
70
60
50
40
30
20
10
0
Eye Tracker guided
Laser Type
Figure 2.3. Percentages of All Patients, Low Myopes, and High Myopes with
Spherical Equivalent Refractive Error within 0.50 D of Intended Correction for
Each Laser Type. There was a statistically significant trend across groups for all
eyes (Z = 11.31, p < 0.0001) and low myopes (Z = 7.68, p < 0.0001). No
significant trend was found for high myopes (Z = 3.70, p = 0.0002) after adjusting
for multiple comparisons.
21
All Eyes
High Myopes
100
Percent Within 1.00 D of Intended Low Myopes
90
80
70
60
50
40
30
20
10
0
Eye Tracker guided
Laser Type
Figure 2.4 Percentages of All Patients, Low Myopes, and High Myopes with
Spherical Equivalent Refractive Error within 1.00 D of Intended Correction for
eyes (Z = 10.55, p < 0.0001) and low myopes (Z = 7.61, p < 0.0001). No
significant trend was found for high myopes (Z = 3.28, p = 0.001) after adjusting
22
All Eyes

100 Myopic Astigmatism
90 Spherical Myopia
80
70
60
50
40
30
20
10
0
Eye Tracker guided
Laser Type
Figure 2.5 Percentages of All Patients, Patients with Spherical Myopia Only, and
Patients with Myopic Astigmatism with Snellen Visual Acuity of 20/20 or Better for
eyes (Z = 20.76, p < 0.0001), spherical myopia (Z = 5.44, p < 0.0001), and
myopic astigmatism (Z = 9.26, p < 0.0001).
23
All Eyes

Myopic Astigmatism
110
100 Spherical Myopia
90
80
70
(Snellen)
60
50
40
30
20
10
0
Eye Tracker guided
Laser Type
Patients with Myopic Astigmatism with Snellen Visual Acuity of 20/40 or better for
eyes (Z = 6.39, p < 0.0001). No significant trend was found for spherical myopia
(Z = 2.50, p = 0.01) and myopic astigmatism (Z = 2.49, p = 0.01) after adjusting
24
All Eyes
100
Myopic Astigmatism

Percent Within 0.50 D of Intended 90
Spherical Myopia
80
70
60
50
40
30
20
10
0
Eye Tracker guided
Laser Type
Patients with Myopic Astigmatism with Spherical Equivalent Refractive Error
within 0.50 D of Intended Correction for Each Laser Type. There was a
statistically significant trend across groups for all eyes (Z = 11.31, p < 0.0001)
and low myopes (Z = 4.17, p < 0.0001). No significant trend was found for
myopic astigmatism (Z = 0.13, p = 0.89) after adjusting for multiple comparisons.
25
All Eyes
Myopic Astigmatism
100 Spherical Myopia
Percent Within 1.00D of Intended
90
80
70
60
50
40
30
20
10
0
Eye Tracker guided
Laser Type
Patients with Myopic Astigmatism with Spherical Equivalent Refractive Error
within 1.00 D of Intended Correction for Each Laser Type. There was a
statistically significant trend across groups for all eyes (Z = 10.55, p < 0.0001)
and spherical myopia (Z = 4.54, p < 0.0001). No significant trend was found for
myopic astigmatism (Z = 1.59, p = 0.11) after adjusting for multiple comparisons.
26
Laser Category
Scanning
with Eye Wavefront-
Symptom Broad Beam Scanning Tracker guided
†
Glare 137/1259 103/572 429/572 31/349
(10.9%) (18.0%) (36.6%) (7.9%)
†
Halos 130/1259 165/651 286/1171 36/394
(10.3%) (25.4%) (24.4%) (9.1%)
Night driving 7/661 145/651 407/1171 49/394
problems† (1.1%) (22.3%) (34.8%) (12.4%)
Dryness‡ NA 107/572 82/339 55/394
(18.7%) (24.2%) (14.0%)
NA = data not available for laser category
†
Statistically significant trend across laser types (Glare: Z = 8.45, p < 0.0001;
Halos: Z = 4.37, p < 0.0001; Night driving problems: Z = 10.31, p < 0.0001)
‡
Test for trend across laser types is not statistically significant (Z = –1.56, p =
0.12)
Table 2.3 Proportion of all patients in each category of laser with postoperative
symptoms
myopic astigmatism for the percentage of patients within ± 0.50 D (Figures 2.3 and 2.7)
and for the percentage of patients within ± 1.00 D (Figures 2.4 and 2.8); however, it is
important to note that data on patients with myopic astigmatism were only available for
broad beam and scanning lasers. Statistically significant differences were also found for
all other variables listed in Table 2.2. These comparisons were not clinically meaningful,
however, as the largest proportion of patients with each variable was almost always less
than one percent.
Significant trends across laser types were found for symptoms related to night
vision problems, but not for dryness symptoms when the data were analyzed for all
27
patients (Table 2.3). When the analyses were repeated for low myopes only, the only
statistically significant trend across laser types was for the halos symptom. The difference
between the two laser types for halos was approximately seven percent. No significant
trends were found across laser types for low myopes for the other symptoms: glare, night
driving problems or dryness (Table 2.4). Note that the proportion of patients experiencing
each symptom in the wavefront-guided category in Table 2.3 is identical to the proportion
of patients experiencing each of the symptoms in Table 2.4; this reflects the fact that data
for higher myopes were not included in reports for the wavefront-guided lasers. Also,
comparisons could not be made across lasers for high myopes, patients with myopic
astigmatism, or patients with spherical myopia only because symptom data were not
available for these categories.
28
2.3 CONCLUSIONS
Based on these analyses, it appears that increases in the proportion of subjects
with good uncorrected visual acuity have occurred with changes in laser technology
(Figures 2.1, 2.2, 2.5 and 2.6). These trends were present for analyses including the entire
cohort of patients as well as for low myopes only. It also appears that a higher proportion
of patients are receiving the intended refractive error correction (Figures 2.3, 2.4, 2.7 and
2.8) as improvements in laser technology have occurred. Again, these trends were present
for analyses including the entire cohort of patients as well as for low myopes only.
Trends across laser types for subjective symptoms after LASIK were not as
conclusive. Trends showing improvements in night vision symptoms (glare, halos, and
Laser Category
Scanning
with Eye Wavefront-
Symptom Broad Beam Scanning Tracker guided
†
Glare NA 36/348 NA 31/394
(10.34%) (7.87%)
Halos‡ NA 69/427 NA 36/394
(16.16%) (9.14%)
Night driving NA 63/427 NA 49/394
problems† (14.75%) (12.44%)
Dryness† NA 59/348 NA 55/394
(16.95%) (13.96%)
NA = data not available for laser category
†
Test for trend across laser types is not statistically significant (Glare: Z = –1.17,
p = 0.24; Night driving problems: Z = –0.97, p = 0.33; Dry eye: Z = –1.12, p =
0.26)
‡
Statistically significant trend across laser types (Halos: Z = –3.01, p < 0.003)
Table 2.4 Proportion of low myopes in each category of laser with postoperative
symptoms
29
night driving problems) were found when analyses included the entire cohort of patients.
When the analyses were repeated for low myopes only, however, trends showing
improvements in night vision symptoms were found only for the symptom of halos. This
may indicate that trends for improvement in night vision symptoms found for all patients
are driven by differences among laser types in the mean amount of preoperative myopia;
patients in the wavefront-guided laser category were all low myopes and this significant
difference in patient selection criteria may be driving the trends for improvement when
analyzing data for the entire cohort of patients.
It is important to note that no trends for improvement in the proportion of patients
experiencing symptoms of dryness were found for either all patients or low myopes only.
This result was expected. Lamellar flaps were cut during the LASIK procedure for all
four types of laser, and the changes in technology that were implemented across laser
types were not intended or purported to improve symptoms of dryness.
Studies that are retrospective or comparative but not randomized have limitations
that are important to note. Limitations to this study include the use of binomial
proportions for each laser instead of individual patient data; therefore, it is impossible to
control for differences in patient selection criteria, such as the level of preoperative
myopia. Also, the results of this study are limited by our assumption that data reported in
the summaries of safety and efficacy for each laser are accurate. We were also limited by
the data available. For example, very few of the summaries included measurements of
contrast sensitivity/low-contrast visual acuity, which would have provided a more
objective measurement of quality of vision than patient-reported symptoms.
30
Another important limitation to this study is the fact that the patients in each laser
category were not randomly assigned to that category of laser. We compared data that
were collected over a number of years, and our comparison of laser types may be
confounded by a number of different variables, i.e. changes in patient selection criteria,
surgeon experience, new knowledge regarding the biomechanics of the cornea, etc.
Nonetheless, the unique nature of an FDA clinical study offers us some assurances as to
the quality of these data. This level of quality and consistency across studies is not
usually available for metanalysis studies of published literature, so this data set has
provided us with a unique opportunity to examine the relationship between changes in
laser technology and LASIK outcomes.
In a recent survey of members of the American Society of Cataract and Refractive
Surgery (ASCRS), survey participants commented on the “value” of customized or
wavefront-guided ablation. In that study, 8% of participants said the technology was
extremely valuable, 50% said it was moderately valuable, 39% said it was of little value,
and 3% said that this technology was of no value.34 Either many of the participants in that
survey study have not experienced a large change in outcomes in their practice when
beginning to use wavefront-guided technology, or many of the participants are waiting
for evidence that would demonstrate the value of wavefront-guided/customized ablations.
In consideration of the results of our analyses and the ASCRS survey data, future
studies should carefully examine differences between low and high myopes to determine
if improvements in laser technology have truly improved outcomes, or if the trends for
improvement noted in this study are due to other factors such as the level of preoperative
myopia of the cohort tested. These studies should be designed to make comparisons
31
between groups of patients who were randomly assigned to a laser type and should
include low-contrast visual acuity measurements.
Data collection should be conducted in a double-masked design; however, there
are potential pitfalls to this type of data collection. Randomizing eyes of patients instead
of randomizing patients would provide a more cost effective study design, but it would
make measurement of patient symptoms impossible with any currently available
questionnaires. This is because none of these instruments asks participants to answer the
questions based on a specific eye.22, 41-44 One might wonder if patients were truly capable
of making these distinctions in an accurate manner anyway.
In addition, randomization of eyes within an individual patient may also prove
difficult to mask. The entire surgical procedure takes longer when an ablation profile is
performed as a custom/wavefront-guided ablation. Patients enrolled in the study would
certainly have the potential to be unmasked if they equated the length of the procedure
with the sophistication of the technology. If patients were unmasked, it would render any
subjective inter-eye comparisons invalid. A study that randomized patients, however,
could avoid these pitfalls and further examine the relationship between changes in
technology and improvements in outcomes and could potentially provide information
leading to further improvements in laser technology.
In summary, there is evidence to suggest that visual acuity and the accuracy of the
refractive error correction have improved with changes in laser technology; however,
future studies of higher myopes and patients who are randomized to a laser type are
required to determine it changes in technology have improved problems with night vision
symptoms.
32
CHAPTER 3
DEFINING A SUCCESSFUL LASIK PROCEDURE
When attempting to identify factors that would predict unsuccessful outcomes of
corneal refractive surgery procedures, it is important to first define which outcomes are
considered unsuccessful. This definition could then be subjected to experimental testing
and researchers could identify factors that predicted the unsuccessful outcomes. In order
to assist in generating definitions of successful and unsuccessful LASIK procedures, we
chose to conduct focus group sessions with eye care practitioners who have experience
with the preoperative and postoperative care of refractive surgery patients as well as with
clinical research. The purpose of the present study is to obtain the opinions of eye care
practitioners regard LASIK outcomes. Specifically, we will ask them to outline the
elements of a successful LASIK procedure, and what might predict a successful or
unsuccessful procedure.
3.1 METHODS
Sample
An eye care practitioner was eligible to participate if he or she performed LASIK
on a regular basis and/or delivered preoperative or postoperative care to LASIK patients
on a regular basis. Sampling was accomplished by conducting focus group sessions
33
during the 2003 American Society of Cataract and Refractive Surgery (ASCRS) annual
meeting in San Francisco. Subjects were recruited for participation through email or fax
solicitations after being identified in the meeting program as someone who was
presenting research or educational lectures related to LASIK outcomes, or who had been
identified by a member of the study group as someone who had an interest in or history
of conducting research related to LASIK outcomes. Other subjects were recruited through
snowball sampling, i.e., as subjects were enrolled in the study they were asked to suggest
other participants.
These criteria were chosen for feasibility. Even the largest cities have only a
handful of eye care practitioners who would meet our criteria, so a scientific meeting was
the only practical location to obtain a large diverse sample. Also, we were seeking to
obtain a definition of success that we would be able to test experimentally; thus
participants needed to have some knowledge of research in order to understand what
would constitute a definition that could be tested experimentally.
Data collection
Two focus group sessions were held on the first day of the ASCRS meeting, one
at breakfast and the other at lunch. The third session was held at breakfast on the second
day of the meeting. All sessions were held in the same small conference room at a hotel
near the meeting. We evaluated the data after the third session in order to determine if
theoretical saturation had occurred. After the first two sessions, no new or relevant data
seemed to emerge; therefore data collection ended after the third session.
34
Sessions were conducted by a white, male interviewer who had facilitated focus
group sessions for more than 75 studies. The interviewer used a flexible discussion guide
for the focus group sessions that was adapted in later sessions to develop emergent issues
fully. The discussion guide for the focus groups was designed based on a widely-accepted
principle in psychological research: that the experience of maximum success occurs when
desirable outcomes are maximized and undesirable outcomes are minimized. Therefore,
participants were asked to verbalize, describe, and justify as wide a range of possible
desirable and undesirable outcomes of LASIK as possible.
All sessions were audiotaped and transcribed verbatim. Ideas and answers to
questions that were presented during the sessions were recorded on a flip chart during the
session. After each session, participants completed a brief questionnaire on demographic
information. A meal was provided in order to stimulate conversation within the group.
Data analysis
Immediately after each session, a debriefing occurred between the moderator and
the assistant moderators in order to contrast findings from earlier sessions and to capture
first impressions of the group. Once transcriptions of audiotapes were complete, the data
from the flip charts and transcripts were coded and then categorized. Finally, a report of
emerging themes was developed by topic and then for the overall sessions.
3.2 RESULTS
Overall, 18 subjects participated in discussions across the three sessions. Fifteen
of the subjects were males, and three subjects were females. Eleven subjects held Doctor
35
of Medicine (MD) degrees and seven subjects held Doctor of Optometry (OD) degrees.
Only one subject practiced in a solely academic practice; all other participants were in
private practice. The mean ± SD LASIK procedures conducted per month was 120 ± 185
(range: 8 to 750). On average, study participants primarily have practices devoted to care
of refractive surgery patients, with a mean ± SD percentage of the subjects’ practices
devoted to refractive surgery of 55 ± 36 % (range = 10% to 100%).
Primary and secondary themes
The primary theme to emerge from discussions related to desirable outcomes was
patient happiness. The patient’s perception of the procedure was valued as one of the
most important indicators of success on an individual basis. Comments related to this
theme included: “The patient is happy” or “We have certain ways that we measure 45, so
we have Snellen acuity, contrast sensitivity, corneal topography, wavefront testing. We
have all of those ways that we measure it, but what we’re really after is a happy patient.
We’ve met their needs.”
Discussions of meeting the patient’s needs and expectations were often
intertwined and/or accompanied by discussions related to the patient’s happiness. “It all
depends on what you do preoperatively. If you set high expectations, I mean that’s tough.
That’s an equation. Whoever said that famous quote that success is results minus
expectations? It’s true.”
Participants also expressed frustration about how patient happiness seemed to be
an arbitrary target that varied from patient to patient and did not appear to correlate well
with more objective outcomes. “We have all the tools that we have that we utilize to try
36
to measure [success], but I guess the frustration rests with this: Is there a good correlation
between those things? Even in our discussion on desirable outcomes, it came out that
we’ve got patients that are 20/40 that are happy, and we’ve got patients… that are 20/20
that are unhappy. We know that it doesn’t correlate one to one, and obviously that
frustrates us because we’d like a test where it takes 30 seconds to capture whether they’re
happy or unhappy.”
In the first two focus groups session participants discussed how they perceive the
effect of the patient’s personality on whether or not the patient is happy with the outcome
of the procedure. Comments included, “It’s indirectly proportional to their critical, their
type A nature. I mean the more acute they are about their awareness of things, the less
potential for success that you can have. You can satisfy those folks as well, but you have
to make sure that they understand the education involved. The pre-procedure education is
critical.” Participants also expressed an interest in finding ways to deal with the aspect of
patient personality in day-to-day patient care. “If you could somehow tie personality
traits with long-term success rates, you know that would be good… If you could tie
personality traits with anything up there, that would be pretty good!... I was just going to
say if you could give them a personality test... You could [then] correlate it with
[outcomes].”
Participants were reminded that the patient’s point of view would be fully covered
in similar sessions with LASIK patients and were encouraged to focus on outcomes that
would make a procedure successful from their point of view. A second dominant theme
emerged from these discussions in the ideas expressed by the participants and in the
criteria that were chosen for measuring success: a desire for the postoperative condition
37
Desirable Outcomes Undesirable Outcomes
1. Good/clear Vision 1. Increased aberrations/ degraded optics*
2. Happiness 2. Not meeting patient expectations**
3. Better quality of life 3. Unstable/irreversible complications***
4. Decreased dependence 4. Lost Flap**
on glasses/contact 5. Infections***
lenses 6. Diffuse lamellar keratitis***
5. Meet patient’s main goal 7. Double/ghost images**
6. Quick and painless 8. Night vision symptoms (halos, glare,
surgery starbursts)***
9. Ectasia**
10. Dry eye***
Table 3.1 Desirable and Undesirable Outcomes Emerging From Focus Group
Sessions. The number of asterisks (*) next to each item indicates the number of
sessions in which the topic was mentioned.
of the patient to be equal to or better than the preoperative condition. Desirable and
undesirable outcomes listed during the session are summarized in Table 3.1. It was not
uncommon for desirable outcomes to be expressed in the inverse manner under
undesirable outcomes in the same session or a different session, i.e., happy patient vs.
unhappy patient; however, the table only expresses each idea or category once.
This second theme was expressed in multiple ways throughout the sessions. As an
example, issues related to dry eye and night vision symptoms were mentioned in all three
sessions. The references frequently referred to “no new” or “no increase” in these
symptoms. When discussion turned towards measuring success, this theme remerged,
with participants indicating that they thought procedures were successful when
symptoms, visual acuity, and contrast sensitivity were no worse than baseline. After these
38
themes were noted in the first two sessions, the moderator adapted the guide to include an
element in the final session where participants were asked to pick a cut-point for success
on a scale from worst possible outcome to the most ideal outcomes for each of the
measurements that had been discussed in the two previous sessions. Methods for
measuring success included:
o Quality of Life/Satisfaction surveys
o Snellen Visual Acuity
o Refractive Error
• Within ± 0.50 D of the intended correction
• no change in refractive error over time, or < 0.50 D change in refractive
error over the course of 3 to 5 years
o Wavefront/ Aberrometry
o Corneas Clear
o Patient’s expectations met
o Contrast sensitivity
o Endothelial cell counts
o Less dependent on spectacles
• Older patients free of spectacles 85% to 90% of the time
The participants’ desire for the postoperative condition to be equal to or better than the
preoperative condition was evident in the cut-points offered by the participants. Figure
3.1 is a photograph of the flip chart used during this session to assist participants in the
selection of cut-points.
39
Figure 3.1 Photograph of flip chart for participants’ comments during discussion
of a successful “cut point” for each measurement of success
40
3.3 CONCLUSIONS
Two major themes emerged from collection of these data. Focus group
participants made judgments about the success of a procedure based on the patient’s
overall happiness or satisfaction. Participants could also give examples, however, of
patients who had poor vision but who were satisfied and of patients who had good vision
but who were not satisfied. When the participants were asked to give their criteria for a
successful procedure, independent of the patient’s judgment of the procedure, a second
theme emerged. Participants agreed that successful procedures were ones in which the
patient had vision equal to or better than preoperative conditions without the aid of
spectacles or contact lenses.
The results of this study indicate that practitioners rely on both subjective
evaluations and objective evaluations of LASIK procedures, even though they readily
admit that the two are often not related. In addition, participants also offered observations
that the patient’s personality traits appear to factor into this disparity. Based on the
discussions of participants in this study, successful outcomes should be defined as
procedures in which the patient has vision and comfort after the procedure that is equal to
or better than prior to surgery in future studies. Future studies should also seek to define
the relationship between the successfulness of the procedure and the patient’s personality.
41
CHAPTER 4
REASONS WHY PATIENTS RECOMMEND LASER IN SITU KERATOMILEUSIS

TO OTHERS
Previous investigations have characterized patient satisfaction and complaints,
such as night vision symptoms and dry eye, after laser in situ keratomileusis (LASIK).3,
31, 32, 46-48
Although night vision symptoms have been reported in 12% to 57% of patients3,
21, 46-50
and dry eye symptoms have been reported in 4% to 9% of subjects,4, 32 post-
LASIK satisfaction remains very high. For example, Hill reported that 15 (7.5%) of the
200 subjects considered their night vision symptoms to be “considerably worse” than
before LASIK.48 Still, 195 (97.5%) of 200 subjects in the same study reported that they
were “extremely happy,” and only one subject reported that he or she was “slightly
unhappy.” Similar disparities have been reported after photorefractive keratectomy
(PRK). Brunette and co-workers found that 31.7% of PRK patients reported a decrease in
night vision; nevertheless, 91.8% of subjects were satisfied or very satisfied with their
surgery.41
In a previous paper by our group, we characterized the factors associated with
night vision symptoms and decreased satisfaction following LASIK.50 In this study, we
also report a disparity between the relatively frequent report of postoperative night vision
symptoms and high patient satisfaction. Ninety-seven percent of the subjects in this
sample of LASIK patients would recommend LASIK to a friend.50 When asked to rate
42
their satisfaction with their vision, subjects reported a median score of 100% of the
maximum value on the visual analogue scale (mean=87.2%). Still, 30.0% of the subjects
reported halos, 27.2% reported glare, and 24.5% reported starbursts when asked about
these symptoms in a questionnaire. While this disparity could be explained by assuming
that the symptoms were not severe enough to adversely affect satisfaction reports, there
may be other explanations. The purpose of this study is to further investigate explanations
for this disparity by examining and categorizing the reasons why post-LASIK patients
say they would recommend LASIK to a friend. We also look for associations between
categories of reasons why patients recommend LASIK to others and factors known to be
associated with an increase in postoperative night vision symptoms and decreased
satisfaction following LASIK.
4.1 METHODS
Subjects
Previously myopic LASIK subjects were recruited from three refractive surgery
practices. All subjects were required to have had either LASIK or their last LASIK
enhancement procedure at least six months prior to completing the questionnaire to insure
corneal and refractive error stability. Two of the practices were located in Pennsylvania:
the University of Pittsburgh Eye and Ear Institute and the Northeastern Eye Institute in
Scranton. The third practice was at the University of California, Los Angeles. A total of
2100 questionnaires were mailed to eligible subjects. Questionnaires were mailed to 421
patients from the University of Pittsburgh, 979 patients from the Northeastern Eye
43
Institute, and 700 patients from the University of California, Los Angeles. In order to
generalize the results of this study to as many LASIK patients as possible, the only
exclusion criteria for this study were pre-LASIK refractive surgery, post-LASIK
enhancement with Intacs inserts, and/or failure to complete the required informed consent
document. Each surgeon in our study used only one laser, either the Summit Apex Plus
(Alcon Laboratories, Fort Worth, TX) with a 6.0 mm laser optical zone for spherical
myopic treatments and 5.5 mm x 6.5 mm optical zone for astigmatic myopic treatments
or the VISX Star S2 (VISX, Santa Clara, CA) with a 6.0 mm or 6.5 mm ablation zone for
spherical treatments. The zone for astigmatic treatments changed based on the amount of
astigmatic correction required for an individual patient.
Informed consent
The institutional review boards at the University of Pittsburgh, the University of
California in Los Angeles, and The Ohio State University approved the protocol for this
study. Because the Northeastern Eye Institute is a private practice and does not have an
institutional review board, the institutional review board at The Ohio State University
provided human subjects review for that site. The institutional review boards of the
University of Pittsburgh and The Ohio State University required that subjects sign and
return an informed consent document in addition to the questionnaire. Additional
informed consent documents were mailed to any subject who returned a questionnaire
without the required informed consent document. If neither this copy of the informed
consent document nor the original copy was returned, no medical chart data were
44
collected for that subject, and the subject never entered the study. The institutional review
board for the University of California, Los Angeles considered subjects who returned the
questionnaire to have provided informed consent by simply returning the questionnaire
and did not require them to complete an informed consent document.
Questionnaire
The questionnaire used for this study was designed to assess the visual symptoms
of postoperative LASIK patients. Any questions that required the subjects to rate their
response included a line without any divisions or numbering except for the maximum and
minimum response, a visual analogue scale.51 Subjects were asked to place a mark on the
line to indicate their response along the continuum from minimum to maximum. A
subject’s response was measured as the distance (in mm) from the start of the line to his
or her mark. These data were treated as continuous instead of categorical.
The questionnaire included three questions that addressed satisfaction after
LASIK. The first question asked: “Overall, how satisfied are you with your vision since
you had LASIK?” The respondent recorded his or her response on a visual analog scale
labeled “not satisfied” at the left end and “very satisfied” at the right end. The second
question was “Would you recommend LASIK to a friend?” The respondents replied
either yes or no. Lastly, the subjects were asked “Why or why not?” After this question
there was an open space with a line for subjects to write their reasons for recommending
LASIK to a friend.
45
All questionnaires were mailed to the patients between January 1, 2001 and
March 1, 2001. Any questionnaires that were returned because of undeliverable addresses
were mailed again, if the correct address was provided by the United States Postal
Service on the returned mail or if the correct address could be found in the telephone
book. We did not mail a second questionnaire to subjects who did not respond to the first
mailing, nor did we make telephone calls to any of the subjects to encourage them to
return the questionnaire.
Data collection and analysis
Statistical analyses were calculated using the SAS statistical software system
(SAS, Inc, Cary, NC). In order to score responses that contained visual analogue data, a
transparent ruler was placed over the line, and the distance between the subject’s
response and the left end of the line was measured in millimeters, with 0 mm being the
minimum value and 127 mm the maximum value. These values were adjusted to a range
of 0 to 100 for data analysis. Results from questionnaires using visual analogue lines of a
length 100 mm and over have been reported to be more repeatable than when shorter
lines were used.51 The measurements of the lines were made almost exclusively by one
person, and these data were entered by another person. Data from the medical charts
including pupil size, refractive error, visual acuity, corneal thickness, ablation diameter
and depth, corneal eccentricity, and keratometry of all patients who returned the
questionnaire (and an informed consent document, if required) were collected and entered
46
by a separate person. This served to mask the person entering medical chart data from
knowing how individual subjects had responded to the questionnaire.
The questionnaire did not ask subjects to specify if any postoperative symptoms
could be attributed to one eye or the other, so the symptoms reported by the subjects were
considered to be bilateral. For the purposes of data analysis, one eye was randomly
selected for each subject, except when subjects had only one eye treated with LASIK, and
then the eye that was treated with LASIK was selected for analysis.
Data from all questions that were not visual analogue scales were treated as
categorical. A large percentage of subjects designated the maximum value for the visual
analogue question that asked subjects to rate their overall satisfaction with their vision
after LASIK. Thus, the distribution of responses was highly skewed, making analysis as a
continuous variable not feasible. Responses to this question were divided into three
categories: those who responded with the maximum value (Very Satisfied), those who
responded with values between 50% of the length of the line and the maximum value
(Satisfied), and those who responded with values between 50% and the minimum value
(Not Satisfied).
The reasons for recommending LASIK that were provided by subjects in our
sample were categorized and counted. Some subjects provided multiple reasons. All of
the reasons given by each subject were counted, as we had no way of knowing which
reason was the primary reason for recommending LASIK to a friend. Chi square analyses
were used to make comparisons between each of the categories of reasons, demographic
factors and factors known to be associated with decreased postoperative satisfaction or
increased postoperative night vision symptoms.

47
4.2 RESULTS
Response rate
The response rate of the questionnaire was calculated using the total number of
questionnaires returned (841) divided by the total number of questionnaires that could
possibly be returned (1961), for a response rate of 42.9%. Overall, 604 subjects (31%)
were included in the analyses below. The other subjects were excluded from analysis
because of inclusion/exclusion criteria, the patient’s medical chart could not be located or
because they were returned after the deadline.50
Subject demographics
The mean (± SD) age of the 604 subjects was 43.0 ± 10.6 years. Women
comprised 63% of the sample. The mean (± SD) spherical equivalent preoperative
refractive error was –4.99 ± 3.14 D. In total, 30.4% of the subjects were corrected to
produce monovision at the time of the initial LASIK procedure. Overall, 82.7% of
subjects had had LASIK in either 1999 or 2000. The mean (± SD) time between the
subject’s initial LASIK procedure and completion of the questionnaire was 1.55 ± 0.58
years. Of the 604 subjects included in this analysis, 421 subjects had bilateral LASIK;
184 had only one eye treated.
Out of the 586 subjects who would recommend LASIK to a friend, 434 (74.1%)
answered the open-ended question, “Why or why not?” Due to the fact that 26% of the
subjects who recommended LASIK to a friend did not provide any reasons for
48
recommending LASIK to others, we compared satisfaction with vision between those
who provided reasons to those who did not provide a reason. Out of the subjects who
were dissatisfied with their post-LASIK vision, a significantly higher percentage did not
provide a reason for recommending LASIK (14.6% vs 3.07%, χ2 = 27.0, p<0.0001).
Reasons for recommending LASIK
We identified 16 categories of reasons for recommending LASIK to a friend
(Table 4.1). “No more spectacles/contact lenses” was the category most frequently listed
by subjects (180/434), followed by “better vision” (92/434) and “convenience” (65/434).
Overall, the categories most frequently cited by subjects primarily deal with issues of
convenience and quality of vision.
No significant associations were found between the level of satisfaction, the
degree of myopia, corneal curvature, or age (<45 years versus >45 years) and the
categories of reasons for recommending LASIK to a friend. There were, however,
associations between gender and the categories of reasons for recommending LASIK to a
friend. Female subjects were more likely to report better comfort (27 females vs. 3
males, χ2=8.99, p=0.003) and/or better quality of life (41 females vs. 9 males, χ2=7.36,
p=0.007).
Dissatisfaction
As we have reported previously, 35 of the subjects in our sample reported that
they were dissatisfied with their post-LASIK vision.50 Out of the 35 subjects who were
dissatisfied with their post-LASIK vision, 20 of them (57.1%) reported that they would
49
Category Number of Subjects
(n=434)
1. No more spectacles/contact lenses 179
2. Better vision 89
3. Convenience 65
4. “Freedom”* 59
5. I have good results 44
6. Better quality of life 50
7. Better comfort 30
8. Ease of the procedure 24
9. Sports or activities are easier 23
10. Good investment/saves money 15
11. “It is a miracle”* 15
12. The benefits outweigh the risks 9
13. Cosmesis 6
14. Helps others (not me) 4
15. Healthier eyes 3
16. Happy with monovision 3
Reasons for recommending LASIK were categorized and counted for each
subject. All reasons were counted, even if subjects provided more than one.
Thus, the total for all of the reasons (618) is greater than the total number of
subjects (434) who responded to the question.
*Several subjects wrote either “Freedom” or “It is a miracle” as one of the
reasons, and sometimes the only reason, for why they would recommend LASIK
to a friend. The other categories paraphrase the responses of the subjects.
Table 4.1 Reasons for Recommending LASIK to a Friend
still recommend LASIK to a friend.9 These were the only subjects who reported that
“LASIK helps others, but not me” as a reason for recommending LASIK to a friend.
4.3 CONCLUSIONS
In this report we identified 16 categories of reasons why LASIK patients would
recommend the procedure to a friend. The most common reason for recommending
LASIK to a friend was “no more spectacles/contact lenses.” Overall, reasons related to
50
convenience and better vision were the most frequently listed reasons. Although we did
not find an association between the categories of reasons and factors like age and degree
of preoperative myopia, female subjects were significantly more likely to cite better
comfort and/or better quality of life as reasons for recommending LASIK to a friend.
The categories of reasons for recommending LASIK to a friend are not very
different from the reasons patients give for seeking LASIK in the first place. In a recent
report by Khan-Lim and co-workers, the three most common reasons a myopic patient
sought LASIK involved inconvenience of, intolerance of, or a desire for freedom from
spectacles or contact lenses.52 Given that patients seek LASIK to be free of spectacles
and contact lenses and recommend LASIK to others for that same reason, it is surprising
that more studies do not report the percentage of subjects who do not require spectacles
or contact lenses after surgery, or even use this measure as a primary outcome measure of
success. In the few studies that have reported this information, the percentage of subjects
who no longer require spectacles or contact lenses for distance vision ranges from 55% to
95%.3, 5, 48
One of the more intriguing findings of this study is that the majority of the
subjects who were not satisfied with their post-LASIK vision would still recommend
LASIK to a friend (57%). Some of those patients appear to be making that
recommendation based on ideal results as opposed to their actual results. At minimum,
these results suggest that this question, “Would you recommend LASIK to a friend?” is
not a good measurement of visual function and that this question is not a valid measure of
satisfaction after LASIK.
51
There is more than one plausible explanation for the disparity between the
relatively common frequency of postoperative complaints and high levels of satisfaction
after LASIK. The first explanation is that the night vision symptoms and symptoms of
dryness after LASIK are so mild relative to the benefits of LASIK that patient satisfaction
still remains high. There are, however, other explanations.
The discrepancy between symptoms and satisfaction has also been reported after
PRK.41 Brunette and coworkers conclude: “[Patients’] self-conditioning to the belief in
the success of this expensive and irreversible surgery was not assessed in the present
study. However, if such a psychological process prevailed, one would expect patients to
deny the secondary effects such as glare and night vision problems. This was not the case
here. Patients may simply adapt to their new condition.” This idea that patients may adapt
to their new condition is a second explanation for the disparity between symptoms and
satisfaction.
The third explanation we can offer for this disparity is one that is dismissed by
Brunette and coworkers. Although a specific “psychological process” is not mentioned in
their report, they may be referring to a psychological process called cognitive dissonance.
The theory of cognitive dissonance suggests that individuals need to have consistency
among their attitudes or beliefs.53 The theory can apply to many types of decisions, i.e.
purchases, surgical procedures, etc., and the effects of cognitive dissonance have been
previously investigated in other non-ocular surgical procedures.54, 55 There are certain
criteria that are required for dissonance to be induced following a decision.56 First, the
decision must be important, and the consumer/patient must have invested substantial
money or psychological cost. Second, the consumer/patient must freely choose to make
52
the purchase or have the surgical procedure. Finally, the commitment must be
irreversible.
It is evident that the LASIK procedure meets all three of the criteria for inducing
cognitive dissonance in patients who would have post-LASIK complications/side-effects.
Homer and coworkers predict that cognitive dissonance might have less effect on specific
questions about symptoms, and a greater effect on questions about the overall success of
the surgical procedure, such as whether or not the patient would have the procedure again
or recommend the procedure to others.55 This prediction is supported by our previous
study in LASIK patients, and by Brunette and coworkers in PRK patients.41, 50 Cognitive
dissonance should be explored in future studies as an explanation for the disparity
between the relatively frequent occurrence of postoperative symptoms and high levels of
satisfaction. Future studies will also be able to evaluate the effects of more current
technology, such as larger ablation zone diameters, on postoperative symptoms,
satisfaction and the disparity between the two.
53
CHAPTER 5
THE REPEATABILITY OF AUTOREFRACTION AND AXIAL LENGTH

MEASUREMENTS AFTER LASIK
Refractive error and ocular component measurements are commonly used in
studies of the prevalence of refractive error, myopia progression, and emmetropization.57-

61
Studies where changes in ocular components or refractive error are monitored over
time require measurements that are repeatable so that true changes in refractive error can
be distinguished from measurement variability. The measurements used for research
purposes must also agree well with other widely used clinical measurements.
Zadnik and co-workers have reported that cycloplegic autorefraction
measurements not only agree well with the gold standard of subjective refraction
measurements but that cycloplegic autorefraction measurements are highly repeatable in
normal patients.61 Sheng and co-workers have reported that axial length measurements
made with the IOLMaster provide repeatability superior to that of A-scan
ultrasonography and that there was modest agreement between the two measurements.62
Based on these studies, investigators of myopia progression commonly apply a criterion
of –0.50 D as representing a clinically significant change if cycloplegic autorefraction is
used to measure refractive error.61 Similarly, investigators also apply a criterion of a 0.10-
mm clinically significant change in axial length if the IOLMaster is used to measure axial
54
length.62 Changes of less than these amounts are likely to be due to measurement
variability.
Although cycloplegic autorefraction is often used as the primary outcome
measure in studies of refractive error, this measurement has proven to be problematic in
patients who have had corneal refractive surgery. Previous studies comparing
autorefraction and subjective refraction in corneal refractive surgery patients have
reported a lack of agreement between the two measurements and high variability in
autorefraction measurements.63-67 The lack of agreement and increased variability were
not present when investigators tested these patients before LASIK. In addition, two of
these studies described a bias in autorefraction measurements in post-LASIK patients,
reporting autorefraction measurements that were more myopic than subjective refraction
measurements.63, 66
The purpose of the present study was to evaluate refractive error measurements
and axial length measurements in patients after LASIK. First, we sought to evaluate
autorefraction measurements in terms of repeatability and agreement with subjective
refraction measurements, using readily available autorefractors with two different design
principles. Second, we evaluated the repeatability of axial length measurements made
with the IOLMaster.
5.1 MATERIALS AND METHODS
Subjects
Subjects were recruited between January and March of 2002 through
advertisements in the local newspaper. All subjects were over the age of 18 years, were
55
previously myopic, and had had LASIK or any LASIK enhancements at least three
months prior to enrollment in the study. The purpose and procedures of the study were
explained, and written informed consent was obtained from all subjects. The study
protocol was approved by the Biomedical Sciences Institutional Review Board at The
Ohio State University. Measurements were made on both eyes of all 40 subjects on two
different occasions separated by approximately two weeks (range 11 days to 32 days).
Measurements were made by one observer for all subjects on both occasions. For
cycloplegic measurements, one drop of 0.5% proparacaine was instilled followed by two
drops of 1% tropicamide, five minutes apart. Cycloplegic measurements were made at
least 25 minutes after instillation of the last drop.
Refractive error measurements
All refractive error measurements, including subjective refraction, autorefraction
with the Binocular Auto Refractor/Keratometer WR-5100K (Grand Seiko Co., Ltd,
Hiroshima, Japan) and autorefraction with the Humphrey® Automatic Refractor
Keratometer (HARK) Model 599 (Carl Zeiss Meditec, Dublin, CA), were made under
non-cycloplegic and cycloplegic conditions. One subjective refraction was obtained on
each eye followed by five measurements with the Grand Seiko autorefractor and five
measurements with the Humphrey autorefractor. Grand Seiko and Humphrey
autorefraction measurements included in data analysis are the mean of the five
measurements. Subjective refraction was always obtained first, followed by
autorefraction measurements with the Grand Seiko and then the Humphrey autorefractor
to avoid biasing the examiner and therefore the results of the subjective refraction. The
56
instrument was re-aligned prior to each autorefraction measurement, and all five
measurements were made on the subject’s right eye prior to measuring the subject’s left
eye.
During examination with the Grand Seiko autorefractor, the eye that was not
tested was occluded with a patch. Measurements were made with the Grand Seiko as
described previously for the Canon R-1 in the Collaborative Longitudinal Evaluation of
Ethnicity and Refractive Error (CLEERE) Study.60 Refractive error measurements were
made with the Humphrey autorefractor in manual mode. Subjects were instructed to look
at the target, a star pattern followed by a letter chart, inside the instrument. The image of
the subject’s pupil was aligned by the examiner within the center of the white target box
shown on the examiner’s screen before each reading.
IOLMaster measurements
Axial length measurements were made following procedures recommended by the
manufacturer with the IOLMaster (Carl Zeiss Meditec, Dublin CA), a non-contact
method for measuring axial length, under non-cycloplegic conditions. Axial length
measurements were made until five axial length measurements of high confidence were
obtained. High-confidence measurements were defined as those with a signal to noise
ratio > 2.0. The mean of the five high-confidence measurements was used in data
analysis.
57
Analysis of refractive error data- conversion to power vector representation
We have chosen to convert refractive error data in this study to power vector
representations (M, J0, and J45) from the conventional method of representing refractive
error data (sphere, cylinder, and axis) based on work presented by Thibos and co-
workers.68, 69 The overall blurring strength of the subject’s refractive error, B, as
( )
described by Thibos et al., was also calculated B = M 2 + J 02 + J 452 . There are two
reasons for this decision. First, when conventional statistical analyses are applied to
refractive error data represented in the traditional form of sphere, cylinder, and axis,
misleading results are often obtained, as these statistical methods are not appropriate for
analyzing directional data, i.e. astigmatic axes. Second, the power vector method
represents spherocylindrical refractive error in three vectors (M, J0, and J45) that are
analogous to the three second-degree Zernike polynomials (Z02, Z22, and Z–22) that
describe defocus (Z02 = M), with-the-rule astigmatism and against-the-rule astigmatism
(Z22 = J0), and astigmatism with axes at 45 degrees and 135 degrees (Z–22 = J45). The
refractive surgery community has become increasingly comfortable with the use of
Zernike polynomials to describe all of the wavefront error of the eye. Using the power
vector method to represent spherocylindrical refractive errors when evaluating these
autorefractors allowed the data to be analyzed with conventional statistical methods and
also provided the data in a format that can be readily compared to other studies that
measure refractive error with aberrometers.
58
Statistical methods
All analyses were performed using SAS software (version 8.2) and performed
separately for each eye. To simplify presentation, results are shown only for the right eye,
although conclusions for the left eye were similar. To assess the repeatability of each
measurement, the method described by Bland and Altman was used.70 The mean
difference between the repeated measurements, i.e., the difference between visits,
characterizes the bias of the method. This mean difference was compared to zero using a
one-sample t-test to determine if the bias was statistically significant. The mean
difference and its standard deviation were used to construct 95% limits of agreement
[mean ± (1.96 * standard deviation)]. The limits of agreement characterize the expected
differences between repeated measurements.
The validity of each autorefractor was assessed by comparing the value obtained
from the instrument to the gold standard of subjective refraction. A statistical comparison
of the mean difference between the autorefractor and subjective refraction was performed
using a one-sample t-test. Methods of Bland and Altman were also used for these
comparisons, and we again checked to determine if the difference was related to the
mean. The limits of agreement were not calculated if the difference between
measurements was related to the mean because in this case, the measurements would be
different at every value, making the limits irrelevant. A scatter plot of the difference
between measurements versus the mean of the measurements for each subject was
constructed and used to determine if the bias was related to the underlying measurement.
59
Measurement of autorefractor infrared beam
The Humphrey autorefractor operates on a knife-edge principle and includes an
aperture associated with an internal beam splitter that is about 8 mm in diameter,
allowing near infrared light from the patient’s entire pupil to be included in the refractive
error calculation.71 Conversely, the Grand Seiko autorefractor operates on an image-size
principle.71 In addition to evaluating the clinical performance of this instrument, we
sought to measure the diameter of the infrared beam used by the Grand Seiko
autorefractor in measurements. We used an infrared video camera to record the image of
the infrared beam when it was projected onto a white millimeter ruler as the Grand Seiko
autorefractor was activated. The ruler was placed in the approximate location of a
patient’s corneal plane, and the autorefractor was positioned so that the scale on the
millimeter ruler was in sharp focus prior to activation of the autorefractor. The image of
the infrared beam and the scale on the millimeter ruler were recorded simultaneously by
the infrared camera. This allowed the scale of the ruler to be used for calibration purposes
when we measured the diameter of the image of the infrared beam. This analysis was
completed using Adobe Photoshop version 5.0 software (Adobe Systems Incorporated,
San Jose, CA).
60
5.2 RESULTS
The mean ± SD age of the 40 post-LASIK subjects was 30.3 ± 8.5 years. Twenty-two
(55%) of the subjects were female. The mean and standard deviation of ten measurements
of the diameter of the infrared beam for the Grand Seiko autorefractor was 1.97 ± 0.05
mm. Values for the left eye were similar to those values reported for right eyes; therefore,
all values reported here are for right eyes only.
Repeatability
Repeatability of M (spherical equivalent), J0, and J45, was good regardless of the
method or presence/absence of cycloplegia (Table 5.1). The coefficient of variation
(standard deviation of the difference * 2) for M, J0, and J45 indicated a normal variation in
autorefraction measurements between visits of no more than 0.76 D under any of the
reported conditions.
There was very little bias, i.e. the difference between visit 1 and visit 2, in M
(spherical equivalent) measurements, as almost all differences were appreciably zero
(mean = bias column, Table 5.1), and not statistically significant after adjusting for
multiple comparisons (all p-values > 0.002). None of the six values for the difference
between visits in J0 reported in Table 5.1 was statistically significant (all p-values >
0.29.) Similarly, none of the six values for the difference between visits for J45 was
statistically significant after adjusting for multiple comparisons (all p-values > 0.002).
61
Measure* Mean = bias SD 95% limits of agreement
Non-cycloplegic subjective refraction
M 0.01 0.37 –0.71, 0.73
J0 –0.00 0.30 –0.60, 0.59
J45 0.01 0.19 –0.37, 0.38
B –0.05 0.34 –0.72, 0.62
Non-cycloplegic Grand Seiko autorefraction
M –0.16 0.33 –0.80, 0.48
J0 –0.06 0.33 –0.69, 0.58
J45 0.00 0.27 –0.53, 0.54
B 0.09 0.32 –0.53, 0.71
Non-cycloplegic Humphrey autorefraction
M –0.03 0.30 –0.63, 0.57
J0 0.02 0.23 –0.44, 0.47
J45 –0.02 0.21 –0.44, 0.40
B –0.01 0.30 –0.60, 0.59
Cycloplegic subjective refraction
M –0.01 0.38 –0.76, 0.74
J0 –0.01 0.24 –0.47, 0.46
J45 0.07 0.22 –0.37, 0.51
B 0.04 0.28 –0.51, 0.59
Cycloplegic Grand Seiko autorefraction
M –0.01 0.24 –0.49, 0.47
J0 –0.00 0.38 –0.75, 0.75
J45 –0.02 0.22 –0.45, 0.42
B –0.03 0.26 –0.55, 0.49
Cycloplegic Humphrey autorefraction
M –0.07 0.21 –0.49, 0.34
J0 0.03 0.23 –0.42, 0.49
J45 –0.04 0.28 –0.59, 0.51
B 0.07 0.21 –0.33, 0.48
Axial length 0.008 0.04 –0.07, 0.09
Data are for right eyes only. No statistically significant differences between visits
were found (p > 0.002) after adjusting for multiple comparisons.
Bias = difference between visits (visit 2 – visit 1)
* Refractive error measurements as calculated by Thibos et al., (1997). M =
spherical equivalent, Z02; J0 = astigmatism with axes at 0 and 180 degrees,
Z22; J45 = astigmatism with axes at 45 and 135 degrees, Z–22.
Table 5.1 Repeatability for each measure of refractive error and axial length
62
Measure* Mean SD Range
Non-cycloplegic subjective refraction
M –0.40 0.60 –1.75, +1.25
J0 –0.02 0.17 –0.48, +0.36
J45 +0.01 0.17 –0.25, +0.57
B +0.59 0.48 0.00, +1.75
Non-cycloplegic Grand Seiko autorefraction
M –0.35 0.68 –2.30, +1.16
J0 –0.02 0.24 –0.60, +0.49
J45 –0.02 0.18 –0.40, +0.28
B +0.66 0.48 +0.12, +2.38
Non-cycloplegic Humphrey autorefraction
M –1.30** 0.94 –3.56, +0.74
J0 –0.03 0.12 –0.32, +0.24
J45 –0.01 0.17 –0.57, +0.33
B +1.40 0.82 +0.31, +3.57
Cycloplegic subjective refraction
M –0.33 0.52 –1.75, +1.00
J0 –0.01 0.17 –0.60, +0.36
J45 –0.02 0.14 –0.38, +0.36
B +0.49 0.44 +0.00, +1.75
Cycloplegic Grand Seiko autorefraction
M –0.16 0.63 –1.86, +1.29
J0 +0.03 0.30 –1.15, +0.52
J45 –0.02 0.17 –0.36, +0.28
B +0.59 0.43 +0.11, +1.93
Cycloplegic Humphrey autorefraction
M –2.38‡ 1.31 –6.72, –0.63
J0 –0.05 0.19 –0.43, +0.29
J45 +0.01 0.17 –0.31, +0.31
B +2.40 1.30 +0.67, +6.72
Data are from visit one and include right eyes only.
** Significantly different from non-cycloplegic subjective refraction, p < 0.0001
‡
Significantly different from cycloplegic subjective refraction, p < 0.0001
Table 5.2 Mean and standard deviation of refractive error measurements
63
95% Limits of
†
Measure* Difference (D) Agreement p-value
Non-cycloplegic Grand Seiko vs. subjective refraction
M +0.05 –0.92, +0.78 0.52
J0 +0.00 –0.55, +0.55 0.99
J45 –0.03 –0.55, +0.50 0.48
B +0.08 –0.86, +1.02 0.31
Cycloplegic Grand Seiko vs. subjective refraction
M +0.17 –0.73, +1.07 0.03
J0 +0.04 –0.70, +0.78 0.47
J45 +0.01 –0.44, +0.46 0.84
B +0.10 –0.84, +1.04 0.20
Non-cycloplegic Humphrey autorefraction vs. subjective refraction
M –0.90 –2.35, +0.55 <0.0001
J0 0.00 –0.43, +0.43 0.89
J45 –0.02 –0.55, +0.51 0.69
B +0.81 –0.66, +2.28 <0.0001
Cycloplegic Humphrey autorefraction vs. subjective refraction
M –2.05 –4.32, +0.22 <0.0001
J0 –0.03 –0.52, +0.46 0.40
J45 +0.03 –0.42, +0.48 0.36
B +1.91 –0.42, +4.24 <0.0001
Data are from visit one and include right eyes only.
†
Difference = autorefraction measurement – subjective refraction
measurement
Table 5.3 Agreement between subjective refraction measurements and
autorefraction measurements
64
Agreement between subjective refraction and autorefraction
Mean values for subjective refraction and each autorefraction for right eyes at
visit 1 are shown in Table 5.2. Results for agreement were similar between visit 1 and
visit 2, so only results for visit 1 are reported. The mean for M (spherical equivalent)
obtained using the Grand Seiko was nearly identical to the mean for M from subjective
refraction when no cycloplegia was used (mean difference = 0.05 ± 0.53 D, p = 0.52).The
difference between mean values for the Grand Seiko autorefractor and subjective
refraction with cycloplegia (mean difference = 0.17 ± 0.46 D) was not statistically
significant (p = 0.03) after adjusting for multiple comparisons. For both cycloplegic and
non-cycloplegic conditions, the Humphrey autorefraction measurements were
significantly more myopic than the subjective refractions (mean difference = –0.90 ± 0.74
D and mean difference = –2.05 ± 1.16 D, respectively, p < 0.0001 for both conditions).
For J0 and J45, no significant differences were found when comparing the result obtained
from either autorefractor with subjective refraction, with or without cycloplegia (all p-
values > 0.36).
Figures 5.1 through 5.4 show the agreement between autorefraction and
subjective refraction for spherical equivalent (M). Note that the difference between
Humphrey autorefraction and subjective refraction is dependent on the mean spherical
equivalent with and without cycloplegia (Figures 5.3 and 5.4). Although the scatter of
data points in Figures 5.1 and 5.2 appears to be random, when examining the scatter of
data points in Figure 5.3 and 5.4, a distinct pattern emerged. We fitted a regression model
to the data and a reasonably strong R2 value was found showing a relationship between
the mean of the spherical equivalent measurements and the difference between
65
2.00
1.50
1.00
0.50
Difference in spherical equivalent (M):
0.00
Grand Seiko-subjective (Diopters)
-0.50
-1.00
-1.50
-2.00
-2.50
-3.00
66
-3.50
-4.00
-4.50
-5.00
-5.50
-6.00
-4.00 -3.50 -3.00 -2.50 -2.00 -1.50 -1.00 -0.50 0.00 0.50 1.00 1.50 2.00
Mean Spherical Equivalent (D)
Figure 5.1 Difference versus mean plot for the agreement in spherical equivalent between Grand Seiko
autorefraction and subjective refraction without cycloplegia.

2.00
1.50
1.00
0.50
0.00
Grand Seiko-subjective (Diopters)
-0.50
-1.00
-1.50
-2.00
-2.50
-3.00
-3.50
67
-4.00
-4.50
-5.00
-5.50
-6.00
-4.00 -3.50 -3.00 -2.50 -2.00 -1.50 -1.00 -0.50 0.00 0.50 1.00 1.50 2.00
Figure 5.2 Difference versus mean plot for the agreement in spherical equivalent between Grand Seiko
autorefraction and subjective refraction with cycloplegia.

2.00
1.50
1.00
0.50
0.00
Humphrey-subjective (Diopters)
-0.50
-1.00
-1.50
-2.00
-2.50
-3.00
Modeling difference in methods as a

68
-3.50
function of the mean refraction:
-4.00
R2 = 0.38, p<0.0001
-4.50
Difference = -0.33 + 0.75 * M
-5.00
-5.50
-6.00
-4.00 -3.50 -3.00 -2.50 -2.00 -1.50 -1.00 -0.50 0.00 0.50 1.00 1.50 2.00
Figure 5.3 Difference versus mean plot for the agreement in spherical equivalent between Humphrey autorefraction
and subjective refraction without cycloplegia.

2.00
1.50
1.00
0.50
0.00
Humphrey-Subjective (Diopters)
-0.50
-1.00
-1.50
-2.00
-2.50
-3.00
69
-3.50 Modeling difference in methods as a

-4.00 function of the mean refraction:
2
R = 0.67, p<0.0001
-4.50
-5.00 Difference = -0.47 + 1.20 * M
-5.50
-6.00
-4.00 -3.50 -3.00 -2.50 -2.00 -1.50 -1.00 -0.50 0.00 0.50 1.00 1.50 2.00
Figure 5.4 Difference versus mean plot for the agreement in spherical equivalent between Humphrey autorefraction
and subjective refraction with cycloplegia.

measurements. The Bland Altman 95% limits of agreement are not a valid metric of
agreement in this comparison. Because the difference between subjective and automated
refraction is a function of the mean spherical equivalent, the limits of agreement will vary
in proportion to the mean spherical equivalent.
Axial length
The repeatability of axial length measurement was excellent for both eyes. Axial
length data are reported for the right eye only (Table 5.1). The mean difference between
visits (bias) was near zero (p = 0.19), and the 95% limits of agreement indicated an
expected difference of less than 0.10 mm with repeated measurements.
5.3 DISCUSSION
The first objective of this study was to determine the repeatability of refractive
error measurements and axial length measurements in patients after LASIK. Axial length
measurements were highly repeatable, and these results are very similar to a report by
Sheng and co-workers on the inter-visit repeatability of the IOLMaster in non-LASIK
subjects.62 Sheng and co-workers found 95% LoA ranging from –0.11 to 0.07 in non-
LASIK patients and we found a 95% LoA ranging from –0.07 to 0.09 in LASIK patients.
The vast majority of refractive error measurements, i.e., subjective refraction,
autorefraction with the Grand Seiko autorefractor and autorefraction with the Humphrey
autorefractor, were highly repeatable with and without cycloplegia (Table 1). Still, some
differences between visits approached statistical significance, most notably the difference
between visits for non-cycloplegic Grand Seiko spherical equivalent measurements, and
70
one could debate the clinical significance of the difference, –0.16D. This difference
between visits is perplexing for several reasons.
Because there was not a clinically meaningful or statistically significant
difference between visits when subjects were measured with the Grand Seiko
autorefractor under cycloplegic conditions, it appears that subjects in this study
accommodated a mean of 0.16 D more at visit 2 than visit 1 when measured with the
Grand Seiko autorefractor under non-cycloplegic conditions. This did not occur,
however, when the same subjects were measured with the Humphrey autorefractor at
these same visits. Future studies should investigate the impact of various vision targets on
the stimulation/control of accommodation during measurement with the Grand Seiko
autorefractor under non-cycloplegic conditions. In addition, future studies should
investigate the impact of using a mean of more than five measurements with the Grand
Seiko autorefractor. The instrument is capable of taking measurements very rapidly, and
the use of ten measurements instead of the five used in this study may improve inter-visit
repeatability.
In one of the original papers by Bland and Altman that describes the comparison
of measurement techniques, the authors state “Repeatability is relevant to the study of
method of comparison because the repeatabilities of the two methods of measurement
limit the amount of agreement which is possible. If one method has poor repeatability…
the agreement between the two methods is bound to be poor.” Because we found good
agreement between non-cycloplegic subjective refraction and Grand Seiko autorefraction,
as discussed below, we believe that the difference between visits is an anomaly and not
clinically important.
71
The second objective of this study was to determine if measurements made with
the two autorefractors agreed with the gold standard of subjective refraction. There were
no clinically relevant differences between the M, J0, and J45 components of refractive
error measurements made with the Grand Seiko autorefractor and the corresponding
components of subjective refraction measurements with or without cycloplegia. This was
also the case for the astigmatic components of refractive error measurements, J0 and J45,
made with the Humphrey autorefractor. Conversely, the spherical equivalent (M)
measurements made with the Humphrey autorefractor were significantly more myopic
than subjective refraction measurements with and without cycloplegia (–0.90 D and –
2.05 D, respectively). Both of these differences would be clinically meaningful. This bias
was dependent on the mean spherical equivalent refractive error (Figures 5.3 and 5.4), so
no single correction factor can be used to adjust for the differences in spherical equivalent
between the Humphrey autorefractor and subjective refraction.
Comparisons between the results of this study and previous studies are somewhat
problematic because most previous studies of autorefraction in LASIK patients did not
consider the pitfalls of using conventional statistics with the traditional, directional
representation of refractive error measurements, i.e., sphere, cylinder and axis. Still,
statistical differences aside, our finding that the M component (spherical equivalent) of
measurements made with the Humphrey autorefractor was more myopic than subjective
refraction results is similar to previous studies of autorefraction after corneal refractive
surgery.63, 65-67 On the other hand, we found that refractive error measurements with the
Grand Seiko autorefractor agreed well with subjective refraction; this is the first report of
its kind.
72
Our findings related to the Grand Seiko autorefractor are comparable to results of
previous autorefraction studies in non-LASIK patients. In a study of non-LASIK subjects
tested with the Canon R-1 autorefractor, Zadnik and co-workers suggested a “criterion of
0.50 D of change by autorefraction with cycloplegia should be applied in studies of
refractive error variation over time, because changes of less than 0.50 D may be
attributable to measurement variability.”61 Based on the standard deviation of 0.24 D for
the M component of the Grand Seiko cycloplegic autorefraction, and therefore a
coefficient of repeatability of 0.48D, found in the present study, a criterion of 0.50 D of
change in refractive error is also acceptable in post-LASIK patients measured with the
Grand Seiko autorefractor under cycloplegia. In addition, the differences between visits
for the refractive error components measured with the Grand Seiko autorefractor are
similar to differences in non-LASIK subjects.72
Given that so many previous studies have found problems with autorefraction
measurements in patients after LASIK, one might ask what is different about the Grand
Seiko autorefractor that enables this instrument to accurately and reliably assess
refractive error in patients after LASIK. Although this study was not specifically
designed to answer this question, we hypothesize that the answer is related to the
diameter of the infrared beam used by the Grand Seiko autorefractor. The two
autorefractors used in this study have different measurement principles and dramatically
different infrared beam sizes; the infrared beam of the Humphrey autorefractor fills the
patient’s entire pupil, whereas the infrared beam of the Grand Seiko autorefractor is only
about 2 mm in diameter. We hypothesize that any autorefractor that has an infrared beam
large enough to include either transition zones of the ablated cornea and/or untreated
73
peripheral areas of the cornea will provide falsely myopic autorefraction readings in post-
corneal refractive surgery patients. This may be due to one of two related hypotheses:
either the falsely myopic readings are due to an increase in spherical aberration in the
peripheral areas of the cornea, or the falsely myopic readings are caused by an averaging
of the central, emmetropic refractive error of the eye with the peripheral, more myopic
refractive error of the eye. Future studies should investigate both of these hypotheses.
74
CHAPTER 6
CLINICAL EVALUATION OF THE NEI-RQL-42 IN LASIK PATIENTS
Multiple questionnaires have been developed for evaluating satisfaction, subjective
quality of vision, and/or quality of life in patients who have had refractive surgery
procedures.22, 41-44 These procedures include photorefractive keratectomy (PRK) or laser in
situ keratomileusis (LASIK). Nonetheless, there is very little information available in the
published literature on how these questionnaires perform, in terms of validity and
repeatability, when used to evaluate LASIK patients. The one questionnaire that has been
validated in LASIK patients, the Refractive Status and Vision Profile (RSVP), is a lengthy
questionnaire, and the developers have found it a difficult task to convince patients to
complete and return the questionnaire (51% response rate), thereby making the usefulness
of the RSVP questionable.22
Two other questionnaires show promise for effectively evaluating LASIK results,
the Quality of Vision Questionnaire (QOV) and the National Eye Institute Refractive Error
Quality of Life Questionnaire (NEI-RQL-42). Both questionnaires were shown to be
acceptably repeatable and valid in studies by their developers.41-43 Nonetheless, there are no
published data on the use of the NEI-RQL-42 or the QOV in an exclusively LASIK patient
population.
75
Studies validating the QOV were completed with PRK patients,41 and studies
validating the NEI-RQL included patients with several different types of refractive error
correction.43 It is also important to note that the version of the QOV that was validated in
PRK patients is designed for postoperative administration only. The developers created, but
did not validate, the exclusively preoperative and postoperative versions of the
questionnaire. Additional studies demonstrating the ability of the NEI-RQL-42 to detect
differences in quality of life for patients with different types of refractive error correction
have been published. One study found differences between presbyopic and non-presbyopic
patients for several subscales,73 and another found differences between groups of patients
with various modes of refractive error correction.74
In order to properly assess which of these instruments is best suited to evaluate
outcomes of LASIK, however, the validity and the repeatability of the instruments must be
determined in LASIK patients. The developers of the NEI-RQL-42 have reported on the
responsiveness of the instrument when used in keratorefractive surgery patients but not
when used in an exclusively LASIK patient sample.44 The repeatability of the NEI-RQL-42
in post-keratorefractive surgery patients was not reported. In addition, this study
administered the NEI-RQL-42 preoperatively and then once more within three months after
LASIK. The developers did not make an assessment of the performance of the NEI-RQL-
42 when administered on multiple occasions, such as over the course of the first year after
surgery.
The purpose of the present study is to compare the results of two questionnaires that
measure symptoms and satisfaction after LASIK, the NEI-RQL-42 and the QOV. First, we
will determine if the NEI-RQL-42 is repeatable when administered to post-LASIK patients

76
at two separate visits. Second, the performance of these questionnaires when used in a
longitudinal study is evaluated. By testing the two instruments simultaneously, we will be
able to compare the results across questionnaires for an additional assessment of the
validity of the two instruments.
6.1 METHODS
Study 1: Repeatability of the NEI-RQL-42 in LASIK patients
Subjects completed the NEI-RQL-42 for the purposes of determining the
repeatability of this instrument on two separate occasions, separated by approximately
two weeks. During these two visits we also assessed the repeatability and validity of
refractive error measurements (Chapter 6). For a description of subject recruitment
procedures and exclusion criteria, see Chapter 6. The questionnaire was self-administered
and the instructions given to subjects were “Please answer every question (unless you are
asked to skip questions because they don’t apply to you),” and “If you are unsure of how
to answer a question, please give the best answer you can.” No interpretation on the
meaning or purpose of questions was supplied to the subjects, even if requested.
To assess the repeatability of the NEI-RQL-42, the difference between visits was
calculated for the scores of each of the 13 subscales. We then tested for a significant bias
in the measurements by comparing the difference between visits to zero with a one-
sample t-test. The 95% limits of agreement (LoA) were computed for the difference
between visits for each subscale.
77
Study 2: Comparison of the NEI-RQL-42 and QOV
Subjects and informed consent

Subjects were recruited from patients who presented to LASIK Vision Institutes
in Columbus, Ohio or Dallas/Fort Worth, Texas for a preoperative evaluation for LASIK
surgery. Only myopic patients who actually had LASIK surgery were enrolled. All
subjects were over the age of 21 years.
When LASIK patients presented to the surgery center for their initial examination,
the study was described to patients by an ophthalmic technician. If the patient indicated
that he or she was interested in participating in the study, then a letter with a more
detailed explanation of the study was given to the subject along with an informed consent
document and a packet of the preoperative questionnaires. The study protocol was
approved by the Biomedical Sciences Institutional Review Board at The Ohio State
University. All subjects completed the informed consent document or they were not
enrolled in the study.
Questionnaires
The study included comparisons of two quality of life questionnaires, the NEI-
RQL-42, and the QOV. The subjects completed the questionnaires at each of four time
points: pre-LASIK, and one month, three months, and six months after LASIK. Both
questionnaires were administered to all subjects at each time point and the order in which
the subjects completed the questionnaires was randomly assigned. All questionnaires
were self-administered and returned by mail.
78
Any subject who did not complete and return the questionnaire as scheduled was
contacted by telephone or email and reminded to return the questionnaire at least twice.
Subjects were contacted up to three times if the first two calls/emails were not answered
on one of the previous two occasions and no voicemail/answering machine was available
to leave a message. If required, additional questionnaires were mailed to the subject.
Data analysis
For the longitudinal assessment of the NEI-RQL-42 and QOV, the internal
reliability of each scale, i.e. the consistency of questions within a scale, was evaluated by
computing Cronbach’s Alpha for each subscale at the preoperative administration of the
questionnaire. Finally, the change in subscale scores across visits was determined by
comparisons with Repeated Measures Analysis of Variance (ANOVA) with Tukey
adjusted post hoc analyses.
6.2 RESULTS
Repeatability of the NEI-RQL-42
The subscale scores for both visits, the difference between visits (bias), the 95%
LoA for the difference between visits, and the p-value for t-tests comparing the bias value
to zero are shown in Table 6.1. The bias was not significantly different for any of the
subscales after adjusting for multiple comparisons (all p-values > 0.004). Nonetheless,
the 95% LoA are fairly large in several cases. For example, the 95% LoA for the
Expectations subscale range from –41.80 to 61.80. This means that the coefficient of
79
95% Limits of
Subscale Score Agreement
Standard
Deviation
of the Bias
Subscale Visit 1 Visit 2 Difference* Difference Lower Upper p-value
Clarity of Vision 88.4 91.3 2.9 8.5 –13.7 19.4 0.04
Expectations 66.9 76.9 10.0 26.4 –41.8 61.8 0.02
Near Vision 96.1 96.4 0.3 5.2 –9.8 10.5 0.70
Far Vision 88.8 90.9 2.1 6.9 –11.4 15.7 0.06
Diurnal 81.5 87.3 5.8 16.4 –26.3 38.0 0.03

Fluctuations
Activity 96.9 96.4 –0.5 7.8 –15.8 14.8 0.71

Limitations
Glare 74.4 77.2 2.8 12.8 –22.3 27.9 0.17
Symptoms 86.6 87.7 1.1 7.4 –13.3 15.5 0.36
Dependence 95.1 93.3 –1.8 8.2 –17.8 14.2 0.18

on Correction
Worry 72.2 76.3 4.1 11.8 –19.1 27.2 0.04
Suboptimal 96.3 97.2 0.9 10.4 –19.4 21.3 0.57

Correction
Appearance 92.7 92.5 –0.2 9.5 –18.8 18.4 0.91
Satisfaction 90.0 93.5 3.5 10.0 –16.1 23.1 0.03

with Correction
*Calculated as Visit 2 score – Visit 1 score. No values for the difference between
visits were significantly different from zero after adjusting for multiple comparisons (all
p-values > 0.003).
80
repeatability (standard deviation of the differences * 2) would be 52.86 points, or more
than half of the total value of the subscale (100 points). Also, six of the preoperative
subscale scores and eight of the postoperative subscale scores were in the nineties. This
indicates that many of the subjects scored at or near the maximum score.
Comparison of the NEI-RQL-42 and the QOV questionnaires
All measures of internal reliability, Cronbach’s Alpha, for the individual items
that made up each subscale (preoperative visit) were above 0.4, and the vast majority of
subscales were above 0.6 (Figure 6.1). Note that some NEI-RQL-42 subscales, Worry,
Suboptimal Correction, and Appearance, did not have equivalent subscales on the QOV.
The Satisfaction with Correction subscale only had one question on the NEI-RQL-42;
thus, the internal reliability cannot be calculated. In general both instruments performed
well, but the QOV performed as well if not consistently a little better than the NEI-RQL-
42 (Figure 6.1); however, many of the subscales for the QOV have many more questions
than the subscales of the NEI-RQL-42, which may have artificially inflated the
Cronbach’s alpha scores through redundancy of questions. The exception was the
Symptoms subscale where the internal consistency for this subscale on the NEI-RQL-42
performed more consistently than the Symptoms subscale on the QOV.
Tables 6.2 and 6.3 summarize the performance of the NEI-RQL-42 and QOV
across all visits. Table 6.2 gives the mean value for each subscale at each visit for the
NEI-RQL-42, and Table 6.3 gives the mean value for each subscale at each visit for the
QOV. The mean (standard error) difference between the preoperative visit and the six-
month postoperative visit for each subscale was also calculated for each subscale (Tables
81
6.2 and 6.3). Overall, there was a significant effect of visit for nine of the 13 subscales for
the NEI-RQL-42, and all of the subscales for the QOV. All of the differences between the
preoperative and six-month postoperative visit were positive; a positive score indicates an
improvement in quality of life. Tukey adjusted post hoc analyses show that the significant
effect of visit was primarily driven by differences between preoperative scores and each
of the postoperative scores, with very few significant differences in subscale scores
between each of the postoperative visits (Table 6.4).
82
Cronbach's Alpha
0 0.2 0.4 0.6 0.8 1
Clarity of Vision
Expectations
Near Vision
Far Vision
Diurnal Fluctuations
Activity Limitations
Glare
Symptoms
Dependence on Correction
Worry
Suboptimal Correction
Appearance
Satisfaction with Correction
QOV NEI-RQL-42
Figure 6.1 Internal reliaibility, Cronbach’s Alpha, of the individual items for
each subscale of the NEI-RQL-42 and QOV at the preoperative visit
83
Mean (± standard error) Score
One Three Six
Subscale Preoperative Month Month Month Change* p-value†
Clarity of Vision 67.2 67.1 73.3 79.2 +12.1 0.01
(3.1) (3.5) (3.7) (3.6) (3.9)
Expectations 14.3 46.2 58.6 61.6 +47.3 <0.0001

(4.6) (5.2) (5.4) (5.2) (5.6)
Near Vision 80.7 86.8 86.2 87.2 +6.5 0.03

(2.1) (2.3) (2.4) (2.4) (2.3)
Far Vision 76.6 85.0 87.3 86.6 +10.1 0.0001

(2.0) (2.2) (2.3) (2.3) (2.3)
Diurnal 70.3 74.6 75.4 78.7 +8.3 0.10

Fluctuations (2.7) (3.1) (3.2) (3.1) (3.6)
Activity 65.3 91.8 95.9 94.2 +29.0 <0.0001

Limitations (2.7) (3.1) (3.3) (3.1) (3.3)
Glare 62.1 58.7 65.3 68.9 +6.8 0.32

(3.3) (3.7) (3.9) (3.7) (3.9)
Symptoms 71.5 76.0 82.5 80.5 +9.05 0.0001

(2.0) (2.2) (2.3) (2.2)
Dependence on 51.5 87.7 89.6 90.7 +39.1 <0.0001

Correction (23.0) (3.4) (3.6) (3.5) (3.9)
Worry 43.1 52.6 61.0 67.1 +24.0 <0.0001

(2.8) (3.2) (3.4) (3.2) (3.6)
Suboptimal 82.3 94.0 97.0 92.0 +9.7 0.04

Correction (2.6) (3.0) (3.2) (3.0) (3.7)
Appearance 46.7 76.3 81.5 78.0 +31.4 <0.0001

(2.6) (3.0) (3.2) (3.1) (3.8)
Satisfaction 57.1 80.7 81.7 83.2 +26.1 <0.0001

with Correction (2.6) (3.0) (3.1) (3.0) (3.3)
Bold = statistically significant after adjusting for multiple comparisons (p < 0.004)
* Change = six month postoperative subscale score – preoperative subscale score
†
Repeated Measures Analysis of Variance (ANOVA)
Table 6.2 Mean (± standard error) scores for each of the NEI-RQL-42 subscales
84
Mean (± standard error) Score
One Three Six
Subscale Preoperative Month Month Month Change* p-value†
Clarity of Vision 60.8 72.9 76.1 79.8 19.1 <0.0001
(2.6 ) (3.0) (3.1) (3.0) (3.4)
Expectations NA 84.6 83.5 89.7 NA NA

(3.2) (3.3) (3.2)
Near Vision 60.5 67.9 66.5 73.3 12.9 0.01

(3.0) (3.5) (3.7) (3.5) (4.2)
Far Vision 55.7 75.3 76.2 79.5 23.8 <0.0001

(1.9) (2.2) (2.3) (2.2) (2.5)
Diurnal 13.0 65.4 63.9 75.7 62.8 <0.0001

Fluctuations (3.4) (4.0) (4.2) (4.0) (4.7)
Activity 44.2 70.5 71.1 75.6 31.4 <0.0001

Limitations (2.2) (2.6) (2.7) (2.6) (2.9)
Glare 61.0 64.8 65.4 69.2 8.2 0.003

(2.0) (2.2) (2.3) (2.3) (2.3)
Symptoms 77.1 83.1 85.1 85.9 8.8 0.0004

(2.2) (2.4) (2.4) (2.4) (2.1)
Dependence on 25.8 66.8 66.7 69.0 43.2 <0.0001

Correction (3.1) (3.5) (3.7) (3.6) (4.2)
Worry NA NA NA NA NA NA
Suboptimal NA NA NA NA NA NA
Correction
Appearance NA NA NA NA NA NA
Satisfaction with NA 84.5 83.7 89.6 NA NA

Correction (3.2) (3.3) (3.2)
Bold = statistically significant after adjusting for multiple comparisons (p < 0.006)
* Change = six month postoperative subscale score – preoperative subscale score
†
Repeated Measures Analysis of Variance (ANOVA)
NA = subscale not included in the questionnaire administered at this visit and/or unable to
calculate change because the scale is not included in the questionnaire administered in
the preoperative visit
Table 6.3 Mean (± standard error) scores for each of the QOV subscales
85
QOV RQL
0 0 0 1 1 3 0 0 0 1 1 3
Visits Compared*
1 3 6 3 6 6 1 3 6 3 6 6
Clarity of Vision X X X X X
Expectations NA NA NA X X X X X
Near Vision X X X X
Far Vision X X X X X X
Diurnal X X X X X X
Fluctuations
Activity X X X X X X
Limitations
Glare X X
Symptoms X X X X X X X X
Dependence on X X X X X X
Correction
Worry NA NA NA NA NA NA X X X X
Suboptimal NA NA NA NA NA NA X X X
Correction
Appearance NA NA NA NA NA NA X X X
Satisfaction with NA NA NA X X X
Correction
* Visit 0 = preoperative visit
†
Scales not found in QOV
X = significant differences between the pair of visits (p < 0.05)
NA = subscale not included in the questionnaire administered during one or both
visits
Table 6.4 Tukey adjusted post hoc comparisons of the Repeated Measures
Analysis of Variance (ANOVA) tests for each subscale
86
6.3 DISCUSSION
When assessing the usefulness of a given questionnaire for patient-based research
or even for use in clinical practice, it is desirable to have an instrument that is repeatable
and can detect clinically relevant changes in patient symptoms and quality of life. The
test-retest repeatability of the NEI-RQL-42 was assessed as a part of this study. We found
that most of the subscale scores were consistent across visits and that none of the
differences were statistically significant; however, the bias values reported here were
somewhat higher than those reported in previous studies. Both the developers of the NEI-
RQL-42 and Nichols and co-workers reported excellent repeatability for this
instrument.43, 74 Nonetheless, when we compare the 95% LoA for the test-retest
repeatability of subscale scores obtained in this study to those obtained by Nichols and
co-workers (Table 6.5), the range is actually smaller for each subscale in this study of
LASIK patients, even though there is a significant bias for some subscales (differences
between visits are significantly different from zero).
It is also important to note that when the difference between visits was
significantly different from zero for a given subscale, the difference was fairly small and
amounted to a less than 10% difference. Any biases we noted as differences between
visits are also very small when compared to differences between preoperative and
postoperative visits. When we administered the NEI-RQL-42 to LASIK patients over the
course of six months, the difference between preoperative and six-month postoperative
scores was always several points larger for a given subscale than the bias found in the
repeatability study. In addition, it is likely that most future clinical studies using the NEI-
RQL-42 would attempt to find at least a 10% difference in subscale

87
Current Study Nichols, et al.
Difference Difference
Between 95% Limits of Between 95% Limits of
Subscales Visits Agreement Visits Agreement
Clarity of Vision 2.9 –13.7, 19.4 0.05 –22.3, 22.4
Expectations 10.0 –41.8, 61.8 0.60 –34.5, 35.7
Near Vision 0.3 –9.8, 10.5 –0.50 –11.9, 10.9
Far Vision 2.1 –11.4, 15.7 –0.80 –16.3, 14.7
Diurnal Fluctuations 5.8 –26.3, 38.0 1.10 –26.1, 28.3
Activity Limitations –0.5 –15.8, 14.8 0.40 –19.8, 20.6
Glare 2.8 –22.3, 27.9 0.30 –35.7, 36.2
Symptoms 1.1 –13.3, 15.5 0.30 –19.3, 19.9
Dependence on –1.8 –17.8, 14.2 0.70 –35.4, 36.8
Correction
Worry 4.1 –19.1, 27.2 1.10 –30.7, 32.9
Suboptimal 0.9 –19.4, 21.3 –2.30 –25.4, 20.8
Correction
Appearance –0.2 –18.8, 18.4 2.20 –34.8, 39.2
Satisfaction with 3.5 –16.1, 23.1 3.30 –22.4, 29.0
Correction
Table 6.5 Comparison of the repeatability of the National Eye Institute Refractive
Error Quality of Life (NEI-RQL-42) Questionnaire across studies
88
scores between groups, or a 10% difference after changing modes of refractive error
correction. In this case, the repeatability of the instrument would be acceptable.
In the case of the Expectations subscale, the difference between visits was 10
points, and the coefficient of repeatability was 52.86. The coefficient of repeatability, an
indicator of the amount of variation that can be attributed to measurement error, is more
than half of the total score for the Expectations subscale and represents an unacceptable
level of repeatability. We anecdotally noted during data collection that we were
frequently asked to clarify one of the two questions in this subscale. “If you had perfect
vision without glasses, contact lenses, or any other type of vision correction, how
different would your life be?” Patients seemed to be confused by “or any other type of
vision correction.” Patients often asked if that meant “if I had not had LASIK.” Our study
protocol did not allow for the clarification of the meaning of questions to intentionally
uncover ambiguity in questions during the test-retest process; thus, the confusion
associated with the interpretation of this question may be a cause for the 10 point
difference between visits.
Our test of the internal reliability for the NEI-RQL-42 and QOV demonstrated an
acceptable level of consistency among the questions within each subscale for both
questionnaires. Although, the internal reliability was higher in previous studies for both
the Worry and Appearance subscales when this aspect of the questionnaire design was
evaluated.43, 74 The internal reliability for other subscale scores reported in this study was
remarkably similar to these previous studies.
Our study was different from previous evaluations of questionnaires in that we
evaluated the performance of the NEI-RQL-42 and the QOV in a LASIK-only patient
89
population over the course of time. We found a significant effect of visit for nine of the
13 NEI-RQL-42 subscales. The main differences were, as expected, between the
preoperative visit and each of the postoperative visits. Differences were also noted across
the postoperative visits for Clarity of Vision, Expectations, Glare, Symptoms, and Worry
subscales. For these subscales, quality of life improved from the preoperative visit
through the postoperative visits and reached a plateau after the three-month visit. All
NEI-RQL-42 subscale scores stabilized between the three- and six-month visits. This
finding is similar to the picture of postoperative recovery that most clinicians experience
with LASIK patients, i.e. by six months the vast majority of patients symptoms have
subsided, or are at least stable.
Our preoperative-to-postoperative comparison of scores on the NEI-RQL-42
subscales is remarkably similar to that reported by McDonnell and co-workers.44 If one
compares the preoperative-to-postoperative differences in subscale scores reported in the
two studies, differences of only a few points are found in most cases (Table 6.6). There
are exceptions to this rule. The Clarity of Vision, Dependence on Correction, and Worry
subscales all had larger changes between preoperative and postoperative visits in this
study than those changes reported by McDonnell and co-workers. The postoperative
scores for these subscales, however, were similar between studies; the difference seems
to be in the preoperative scores. In the case of all subscales where we found a larger
90
Current Study McDonnell et al.
6-Month 5-Month
Preoperative Postoperative Preoperative Postoperative
Subscale Score Score Score Score
Clarity of Vision 67.16 79.24 80.91 82.29
Expectations 14.28 61.61 14.05 55.81
Near Vision 80.69 87.20 79.47 87.52
Far Vision 76.56 86.63 79.12 88.46
Diurnal 70.30 78.65 72.21 76.62
Fluctuations
Activity Limitations 65.28 94.24 66.70 93.12
Glare 62.10 68.86 74.73 67.09
Symptoms 71.49 80.54 77.24 84.90
Dependence on 51.52 90.66 26.08 83.85
Correction
Worry 43.06 67.08 51.08 65.68
Suboptimal 82.34 92.04 86.21 96.55
Correction
Appearance 46.69 78.04 64.28 91.79
Satisfaction with 57.14 83.19 56.41 82.61
Correction
Table 6.6 Comparison of preoperative and postoperative NEI-RQL-42 scores
across studies
91
preoperative-to-postoperative difference than McDonnell and co-workers, our sample had
an average preoperative subscale score that was lower (poorer quality of life) than in the
sample studied by McDonnell and co-workers. Thus, there was more room for
improvement in quality of life in our sample of patients.
The most surprising finding in our study was the non-significant difference
between preoperative and postoperative Glare subscale scores. When we compared our
results to that of McDonnell and co-workers who found a decrease in quality of life for
the Glare subscale, we again find a preoperative difference between the two samples of
patients. Our patients had a lower mean score on the Glare subscale (indicating more
difficulties with glare conditions) prior to having LASIK, but the mean postoperative
scores were similar. This indicates that the NEI-RQL-42 performed similarly in the two
studies; the samples of the two studies just had difference baseline characteristics that led
to differences in the preoperative-to-postoperative comparisons.
The NEI-RQL-42 is unique in that it was developed to provide an instrument that
could be used to assess changes in quality of life that occur when one changes from one
type of refractive error correction to another, as well as to compare quality of life across
various types of refractive error correction. Based on the results of this study, it appears
that the NEI-RQL-42 is acceptably sensitive to changes that occur when one has
refractive surgery and repeatable when used to assess outcomes of LASIK. Nonetheless,
future studies should investigate the wording of questions contained within the NEI-
RQL-42 to see if minor changes might clarify the meaning of certain questions within
subscales with poorer repeatability and lead to improvement in the test-retest
performance of the NEI-RQL-42 in LASIK patients.

92
CHAPTER 7
ASSESSMENT OF VISUAL FUNCTION IN PATIENTS AFTER LASIK
As we have reported and discussed previously (Chapter 1), multiple investigators
have found a consistent decrease in low-contrast visual acuity, contrast sensitivity, or
both following corneal refractive surgery under.10, 16, 19, 20 The effects are more
pronounced at low luminance or with large pupil diameters. We have also discussed the
inconsistencies across studies using other levels of luminance and inconsistencies in
determining if these changes persist over time. One potential explanation for these
inconsistencies could be differences in the type and quality of the instruments used to
collect these data. A recent report suggests that two commonly used contrast sensitivity
tests have unacceptable ceiling and floor effects when used in normal and post-LASIK
patients and that the measurements lack test-retest repeatability.75
Nonetheless, contrast sensitivity measurements have demonstrated important
changes in visual function that are induced by LASIK. Table 7.1 is the work of Lee, Hu,
and Wang who demonstrated a decrease in contrast sensitivity in post-LASIK eyes.76 The
table was modified to include a column showing the difference between contrast
sensitivity scores of non-LASIK corneas and post-LASIK corneas. If one follows the
differences in contrast sensitivity down the columns for moderate myopes and then again
for high myopes, it appears that a trend for increasing differences is present as the spatial
93
frequency increases. There is also a trend for increasing differences if one compares the
three luminance conditions, comparing normal luminance to dim luminance and then dim
luminance to dim luminance with glare.
Although these data make a compelling case for the use of contrast sensitivity
measurements to evaluate results of corneal refractive surgery, the lack of repeatability
associated with contrast sensitivity measurements makes the use of contrast sensitivity
measurements in clinical research problematic.75 Visual acuity measurements, however,
are generally very repeatable when the measurements are standardized and logMAR
visual acuity charts are used.77 Visual acuity measurements also carry the added
advantage of being a task that is familiar to patients.
The purpose of this study is to determine if a low-contrast visual acuity chart
and/or a series of low-contrast visual acuity charts used under various lighting conditions
would provide a repeatable and sensitive measurement of visual function in refractive
surgery patients.
94
Test Moderate Myopia (–3.50 to –7.00) High Myopia (–7.00 to –10.00)
Post- Post-
Normal LASIK p- Normal LASIK p-
Cornea cornea Difference value Cornea Cornea Difference value
BSCVA, day 1.10 1.11 –0.01 0.61 1.08 1.07 0.01 0.58
BSCVA, night 1.02 1.01 0.01 0.48 0.99 0.97 0.02 0.46
PCVA, day 1.05 0.95 0.10 0.001 0.98 0.9 0.08 0.07
PCVA, night 0.90 0.75 0.15 <0.001 0.86 0.8 0.06 0.11
CSD (1.5) 1.56 1.41 0.15 0.001 1.50 1.37 0.13 0.001
CSD(3) 1.73 1.60 0.13 0.001 1.71 1.6 0.11 0.003
CSD (6) 1.78 1.59 0.19 <0.001 1.75 1.61 0.14 0.007
CSD (12) 1.56 1.31 0.25 <0.001 1.45 1.35 0.10 0.15
CSD (18) 1.19 1.02 0.17 0.008 1.01 0.96 0.05 0.30
CSN (1.5) 1.48 1.35 0.13 0.001 1.48 1.34 0.14 0.003
CSN (3) 1.66 1.56 0.10 0.001 1.67 1.55 0.12 0.002
CSN (6) 1.63 1.46 0.17 0.001 1.56 1.42 0.14 0.03
CSN (12) 1.20 0.96 0.24 <0.001 1.05 0.94 0.11 0.06
CSN (18) 0.74 0.48 0.26 0.001 0.60 0.48 0.12 0.15
CSNG (1.5) 1.41 1.19 0.22 <0.001 1.40 1.19 0.21 0.007
CSNG (3) 1.58 1.40 0.18 0.002 1.60 1.40 0.20 0.004
CSNG (6) 1.50 1.28 0.22 0.001 1.48 1.25 0.23 0.006
CSNG (12) 1.01 0.71 0.30 0.002 0.95 0.69 0.26 0.004
CSNG (18) 0.60 0.29 0.31 0.003 0.52 0.28 0.24 0.015
*Table is a modified version of a table of data in a report by Lee et al. The column with
differences between the normal cornea and the post-LASIK cornea was added to demonstrate
the trend for increasing differences. BSCVA = best-spectacle-corrected visual acuity, PCVA =
predicted corneal visual acuity, CSD = contrast sensitivity in daylight, CSN = contrast sensitivity in
dim illumination, CSNG = contrast sensitivity in dim illumination with glare
Table 7.1 Contrast sensitivity from a report by Lee, Hu and Wang (2003).*
95
7.1 METHODS
Subjects
Subjects were recruited between January and March of 2003 through
advertisements in the campus newspaper, The Lantern. All subjects were over the age of
18 years, were previously myopic, and had had LASIK or any LASIK enhancements at
least three months prior to enrollment in the study. Following enrollment of each LASIK
subject, a non-LASIK subject was enrolled to serve as a control for comparative analyses.
For each LASIK subject enrolled in the study, a non-LASIK subject who was of similar
age (± 3.0 years) and axial length (± 0.5 mm) was recruited and enrolled as a control. We
controlled for axial length to account for differences in retinal image magnification and
retinal stretching. Thirty-five LASIK patients and 35 control patients were enrolled in the
study. The purpose and procedures of the study were explained to all subjects, and
written informed consent was obtained from all subjects. The study protocol was
approved by the Biomedical Sciences Institutional Review Board at The Ohio State
University. All measurements, except autorefraction, subjective refraction and stereopsis,
were made on the right eyes only of all LASIK subjects and all control subjects on two
different occasions, separated by approximately two weeks. Refraction measurements
were made by one observer for all subjects on both occasions.
Refractive error measurements
Two refractive error measurements, subjective refraction and autorefraction with
the Binocular Auto Refractor/Keratometer WR-5100K (Grand Seiko Co., Ltd,
Hiroshima, Japan), were made under non-cycloplegic conditions for all subjects. The
96
results of the final subjective refraction were placed in a trial frame for all patients for
visual acuity testing.
IOLMaster measurements
Axial length measurements were made following procedures recommended by the
manufacturer with the IOLMaster (Carl Zeiss Meditec, Dublin CA), a non-contact
method for measuring axial length, under non-cycloplegic conditions. Axial length
measurements were made until five axial length measurements of high confidence were
obtained. High-confidence measurements were defined as those with a signal to noise
ratio > 2.0. The mean of the five high-confidence measurements was used in data
analysis.
Visual acuity measurements
Visual acuity measurements were made under best-spectacle-corrected conditions
with logMAR visual acuity charts at multiple levels of contrast (Michelson: 2.5%, 5%,
10%, and 100%) under normal (450 lux), dim (5.0 lux), and dim with glare lighting
conditions. The order of contrast and lighting conditions was randomized across subjects
to minimize learning and fatigue effects. Subjects adapted to each lighting condition for
two minutes prior to visual acuity measurements. A brightness acuity tester (BAT) on the
highest setting was used as a glare source for the dim with glare condition. Visual acuity
testing was performed with the best spectacle correction in a trial frame. Testing was
performed at 6 meters with the visual acuity charts front illuminated to daylight/photopic
conditions (96 - 108 cd/m2) under all conditions. Testing was completed by an
97
experienced examiner, Examiner 1, at both Visit 1 and Visit 2. A second visual acuity
testing session was completed at Visit 2 by an inexperienced examiner, Examiner 2, after
a short break.
Subjects were required to read each letter on each row, beginning with the first
line at the top of the chart. Subjects were encouraged to guess when errors were made
with smaller letters, near the bottom of the acuity chart. Subjects were allowed to stop
reading letters when they missed three or more letters on a line after reading the entire
line. A visual acuity score was assigned based on the number of letters read correctly.
Each row progressively changes by 0.1 log minutes of arc in size and there are five letters
in each row, therefore, each letter has a value of 0.02 logMAR. All visual acuity scores
were corrected for spectacle magnification prior to data analysis.
7.2 RESULTS
Thirty-five LASIK and 35 non-LASIK subjects of similar axial length and age
underwent testing (Table 7.2). A high contrast (100%) measurement under dim with glare
illumination was not available for one LASIK subject for testing with Examiner 2 due to
a visual acuity testing protocol violation. Visual acuity scores at 2.5% contrast under the
dim with glare lighting conditions were not available for one LASIK subject because the
subject was unable to read more than two letters on the top line of the chart at any
distance.
98
LASIK patients Control patients
Factor
Mean (SD) age in years 31.5 (8.7) 30.7 (8.3)
% Male 54.3 42.9
% White 91.4 94.3
Mean (SD) time since surgery in months
OD 30.7 (17.7) NA
OS 30.3 (17.9) NA
Mean (SD) axial length in mm 25.4 (1.1) 25.32 (1.1)
Mean (SD) manifest spherical equivalent RE in diopters
OD -0.25 (0.5) -4.43 (2.2)
OS 0.00 (0.5) -4.48 (2.2)
% stereopsis > 40” 17.1 22.6
Table 7.2 Descriptive statistics by group
Repeatability
Mean (±SD) scores for visual acuity testing under all conditions, for both
examiners, and for both visits are found in Table 7.3. No significant bias was found for
within-examiner (Table 7.4) nor for between-examiner repeatability (Table 7.5) for the
LASIK patients; however, a small but significant bias was found for control patients
between visits for several of the low-contrast visual acuity measurements (Table 7.4). All
significant differences amounted to a difference of four letters or less.
99
Examiner 1, Visit 1 Examiner 1, Visit 2 Examiner 2, Visit 2
Lighting Contrast Mean (SD) Range Mean (SD) Range Mean (SD) Range
LASIK PATIENTS
Normal 100% –0.17 (0.06) –0.30, 0.07 –0.17 (0.06) –0.29, 0.02 –0.17 (0.08) –0.34, 0.00
10% 0.02 (0.09) –0.20, 0.28 0.01 (0.09) –0.18, 0.19 –0.00 (0.11) –0.14, 0.28
5% 0.16 (0.11) –0.06, 0.40 0.14 (0.12) –0.05, 0.38 0.13 (0.11) –0.04, 0.37
2.5% 0.30 (0.12) 0.10, 0.64 0.28 (0.96) 0.14, 0.47 0.28 (0.12) 0.09, 0.65
Dim 100% –0.16 (0.06) –0.24, –0.04 –0.16 (0.07) –0.29, 0.00 –0.18 (0.08) –0.30, 0.00
10% 0.05 (0.10) –0.20, 0.28 0.06 (0.09) –0.10, 0.36 0.04 (0.09) –0.12, 0.34
5% 0.19 (0.12) –0.04, 0.40 0.20 (0.12) 0.00, 0.43 0.19 (0.11) –0.07, 0.46
2.5% 0.35 (0.12) 0.08, 0.62 0.66 (0.12) 0.36, 0.12 0.35 (0.12) 0.09, 0.68
Dim + Glare 100% –0.14 (0.06) –0.24, 0.02 –0.14 (0.08) –0.28, 0.06 –0.12 (0.14) –0.30, 0.44
10% 0.12 (0.10) –0.04, 0.32 0.10 (0.12) –0.07, 0.38 0.11 (0.16) –0.09, 0.80
5% 0.30 (0.13) 0.02, 0.62 0.26 (0.15) –0.01, 0.80 0.26 (0.16) 0.03, 0.82
2.5% 0.51 (0.22) 0.24, 1.28 0.47 (0.15) 0.24, 0.84 0.43 (0.16) 0.10, 0.84
Examiner 1, Visit 1 Examiner 1, Visit 2 Examiner 2, Visit 2
Lighting Contrast Mean (SD) Range Mean (SD) Range Mean (SD) Range
100
CONTROL PATIENTS
Normal 100% –0.18 (0.06) –0.31, –0.04 –0.19 (0.07) –0.36, 0.01 –0.20 (0.08) –0.41, –0.06
10% –0.00 (0.09) –0.21, 0.23 –0.24 (0.07) –0.17, 0.14 –0.04 (0.07) –0.18, 0.10
5% 0.12 (0.09) –0.07, 0.32 0.08 (0.09) –0.05, 0.36 0.07 (0.08) –0.08, 0.26
2.5% 0.26 (0.11) 0.02, 0.46 0.21 (0.11) 0.06, 0.63 0.22 (0.86) 0.10, 0.50
Dim 100% –0.16 (0.07) –0.33, –0.03 –0.18 (0.08) –0.36, 0.01 –0.19 (0.07) –0.33, –0.08
10% 0.04 (0.09) –0.19, 0.26 0.01 (0.09) –0.17, 0.19 –0.00 (0.08) –0.16, 0.24
5% 0.16 (0.10) –0.05, 0.42 0.13 (0.11) –0.03, 0.40 0.12 (0.09) –0.04, 0.34
2.5% 0.30 (0.10) 0.11, 0.52 0.28 (0.10) 0.12 , 0.59 0.27 (0.12) 0.02, 0.65
Dim + Glare 100% –0.15 (0.08) –0.34, –0.01 –0.17 (0.08) –0.34, –0.02 –0.18 (0.06) –0.35, –0.74
10% 0.08 (0.11) –0.07, 0.38 0.06 (0.09) –0.10, 0.28 0.04 (0.10) –0.15, 0.29
5% 0.25 (0.09) 0.10, 0.43 0.21 (0.10) 0.00, 0.42 0.19 (0.10) –0.00, 0.36
2.5% 0.48 (0.19) 0.27, 1.07 0.38 (0.11) 0.16, 0.69 0.35 (0.10) 0.17, 0.60
Table 7.3 Visual acuity scores for each level of contrast and lighting level by group, examiner, and visit
Contrast Bias p–value 95% LoA Bias p–value 95% LoA
Normal Lighting
100% –0.00 0.83 –0.12, 0.12 –0.01 0.27 –0.15, 0.12
10% –0.01 0.65 –0.20, 0.19 –0.02 0.10 –0.18, 0.13
5% –0.02 0.46 –0.27, 0.24 –0.05 0.02 –0.26, 0.17
2.5% –0.02 0.50 –0.27, 0.24 –0.01 0.02 –0.24, 0.23
Dim Lighting
100% 0.00 0.77 –0.13, 0.14 –0.02 0.23 –0.15, 0.12
10% 0.01 0.71 –0.25, 0.26 –0.03 0.07 –0.23, 0.17
5% 0.00 0.91 –0.27, 0.28 –0.04 0.02 –0.21, 0.14
2.5% 0.01 0.71 –0.27, 0.28 –0.02 0.34 –0.21, 0.18
Dim + Glare Lighting
100% 0.02 0.70 –0.13, 0.14 –0.02 0.13 –0.18, 0.14
10% –0.03 0.17 –0.22, 0.17 –0.03 0.09 –0.20, 0.15
5% –0.04 0.16 –0.33, 0.26 –0.04 0.04 –0.23, 0.16
2.5% –0.02 0.47 –0.33, 0.30 –0.09 0.00 NA*
Bias = Visit 2 – Visit 1
*
For Dim + Glare lighting at 2.5% contrast control patients the difference was related
to the mean: Difference = 0.22 – 0.74*Mean
Table 7.4 Within examiner test-retest repeatability
101
Contrast Bias p–value 95% LoA Bias p–value 95% LoA
Normal Lighting
100% –0.00 0.92 NA* –0.00 0.71 –0.12, 0.11
10% –0.01 0.34 –0.17, 0.14 –0.02 0.18 –0.15, 0.12
5% –0.02 0.29 –0.17, 0.14 –0.01 0.30 –0.13, 0.11
2.5% –0.00 0.78 –0.16, 0.15 0.01 0.34 –0.13, 0.15
Dim Lighting
100% –0.02 0.04 –0.12, 0.08 –0.01 0.19 –0.13, 0.10
10% –0.02 0.16 –0.15, 0.12 –0.01 0.46 –0.13, 0.11
5% –0.01 0.68 –0.18, 0.17 –0.01 0.55 –0.14, 0.13
2.5% –0.01 0.33 –0.17, 0.14 –0.02 0.32 –0.19, 0.16
Dim + Glare Lighting
100% 0.02 0.44 NA† –0.01 0.30 NA†
10% 0.01 0.65 NA‡ –0.01 0.39 –0.19, 0.16
5% –0.00 0.95 –0.20, 0.20 –0.02 0.31 –0.24, 0.20
2.5% –0.04 0.09 –0.29, 0.22 –0.03 0.16 –0.29, 0.22
Bias = Examiner 2 – Examiner 1
*
Normal lighting at 100% contrast: LASIK patients (Difference = 0.061 +
0.360*Mean)
†
Dim+Glare lighting at 100% contrast: LASIK patients (Difference = 0.094 +
0.597*Mean), Control patients (Difference = –0.080 – 0.386*Mean)
‡
Dim+Glare lighting at 10% contrast: LASIK patients (Difference = –0.023 +
0.301*Mean)
Table 7.5 Between examiner test-retest repeatability
102
Comparisons of visual acuity between groups
When comparing the visual acuity scores for LASIK subjects to the visual acuity
scores for non-LASIK subjects, no statistically significant differences were found (Table
7.6). Under dim and dim with glare lighting conditions, differences between groups of
approximately two letters were found for some low-contrast conditions, but these differences
were not statistically significant. No significant differences between groups for low-contrast
visual acuity were found when controlling for visual acuity at 100% contrast.
Mean Score Observed Adjusteda

Level of LASIK Non-LASIK Differenceb p- Differenceb p-
contrast value* value†
Normal lighting
100% –0.17 (0.06) –0.18 (0.06) –0.01 0.55 --- ---
10% 0.02 (0.09) –0.00 (0.09) –0.02 0.36 –0.02 0.37
5% 0.16 (0.11) 0.12 (0.09) –0.03 0.17 –0.03 0.18
2.5% 0.30 (0.12) 0.26 (0.11) –0.03 0.25 –0.03 0.25
Dim lighting
100% –0.16 (0.06) –0.16 (0.07) –0.00 0.73 --- ---
10% 0.05 (0.10) 0.04 (0.09) –0.01 0.59 0.01 0.59
5% 0.19 (0.12) 0.16 (0.10) –0.03 0.25 –0.03 0.26
2.5% 0.35 (0.12) 0.30 (0.10) –0.05 0.10 –0.05 0.10
Dim + Glare lighting
100% –0.14 (0.06) –0.15 (0.08) –0.01 0.56 --- ---
10% 0.12 (0.10) 0.08 (0.11) –0.04 0.10 –0.04 0.09
5% 0.30 (0.13) 0.25 (0.09) –0.05 0.06 –0.05 0.07
2.5% 0.51 (0.22) 0.48 (0.19) –0.03 0.54 –0.03 0.54
a
Adjusted for mean visual acuity at 100% contrast for pair
b
Difference = Control value – LASIK value
* Paired t-tests comparing LASIK and non-LASIK scores
†
Linear regression testing the mean difference against zero while controlling for mean
visual acuity at 100% contrast
Table 7.6 Difference between LASIK and Non-LASIK patient visual acuity
103
2
Contrast β p-value R β p-value R2
100% –0.01 0.32 0.03 –0.02 0.08 0.10
10% 0.02 0.19 0.05 –0.01 0.57 0.01
5% 0.02 0.11 0.08 –0.01 0.43 0.02
2.5% 0.03 0.06 0.11 –0.01 0.37 0.03
Table 7.7 Relationship between mean visual acuity scores (across light levels)
and axial length
The relationship between axial length and logMAR visual acuity was not a function
of group at the 100% (p=0.344), 10% (p=0.208), or 5% (p=0.101) contrast level (mean visual
acuity score across all three lighting conditions. There was, however, a significant effect of
group at the 2.5% contrast level (p=0.047) (Table 7.7). Given this relationship and previous
literature showing differences in visual function based on preoperative refractive error, we
further assessed the relationship between visual acuity differences and axial length. Figures
7.1 through 7.4 are plots representing the difference between visual acuity scores for LASIK
and control patient pairs as a function of axial length. For 5% contrast and dim illumination,
there was a significant relationship between axial length and the difference between LASIK
and non-LASIK visual acuity scores (R2 = 0.12, p = 0.04).
104
0.6
Dim lighting at 100% contrast
R2 = 0.037, p = 0.265
0.4
Difference in logMAR visual acuity
0.2
-0.2
-0.4
-0.6
23 23.5 24 24.5 25 25.5 26 26.5 27 27.5 28
Mean axial length
Figure 7.1 Mean axial length versus difference between pairs at 100% contrast in
dim lighting conditions
105
0.6
R2 = 0.030, p = 0.320
0.4
0.2
-0.2
-0.4
-0.6
23 23.5 24 24.5 25 25.5 26 26.5 27 27.5 28
Mean axial length
Figure 7.2 Mean axial length versus difference between pairs at 10% contrast in dim
lighting conditions
106
0.6

difference = 1.253 - 0.051*axial length
0.4 R2 = 0.120, p = 0.041
0.2
-0.2
-0.4
-0.6
23 23.5 24 24.5 25 25.5 26 26.5 27 27.5 28
Mean axial length
Figure 7.3 Mean axial length versus difference between pairs at 5% contrast in dim
lighting conditions
107
0.6
Dim lighting at 2.5% contrast

R2 = 0.055, p = 0.177
0.4
0.2
-0.2
-0.4
-0.6
23 23.5 24 24.5 25 25.5 26 26.5 27 27.5 28
Mean axial length
Figure 7.4 Mean axial length versus difference between pairs at 2.5% contrast in
dim lighting conditions
108
7.3 DISCUSSION
As with choosing any measurement in clinical research, measurements of visual
function should be sensitive to true differences between groups and possess an acceptable
level of test-retest repeatability so that researchers can detect clinically meaningful changes
in visual function. For the visual acuity measurements assessed as a part of this study, the
difference between visits and the differences between examiners was almost always
equivalent to one letter or less. A statistically significant level of bias between visits was
detected for the control (non-LASIK) patients under several conditions. While the differences
between the two visits rarely approached a level of clinical significance, it is curious that this
only occurred for the control group. This finding may be the result of the population from
which we recruited the control patients. All control patients were recruited from the faculty,
staff, and students at The Ohio State University College of Optometry. Because these
individuals might have some prior knowledge of the testing procedures, it is possible that a
portion of these subjects learned which letters were and were not included in the visual acuity
charts or other aspects of the test that would lead to a one or two letter improvement in score.
The improvement in visual acuity scores, or practice effect, did not occur for the testing
between examiners (the two testing sessions that occurred during the second visit), so one
might expect the practice effect we observed with control (non-lay) subjects to reach a
plateau and/or be eliminated with two administrations of visual acuity testing. Regardless, the
improvement we noted between the two visits is not necessarily clinically meaningful.
In addition to considering the repeatability of the tests, one must also consider how
sensitive a test is to clinically important distinctions between groups of patients. Multiple
previous studies have shown that low-contrast visual acuity measurements are sensitive to
109
changes after PRK and LASIK,7, 10, 16, 17 but many different levels of low-contrast visual
acuity charts are available. No previous studies have distinguished one chart as being more
sensitive to visual function after refractive surgery than another. When strictly considering
differences in visual acuity scores between LASIK and non-LASIK patients in this study, it
appeared that larger differences were present with the 5% and 2.5% contrast charts, although
none of the differences was statistically significant or clinically meaningful.
Given that the within-examiner 95% LoA for the 2.5% and 5% contrast charts were
equivalent to several lines under the dim with glare lighting condition, but were more narrow
under the dim condition, we chose to further analyze the differences in visual acuity under
dim illumination between LASIK and non-LASIK patients by determining if the difference
was related to axial length. Previous studies have shown that losses of low-contrast visual
acuity can be larger for higher myopes.18 In this study, the differences between LASIK and
non-LASIK visual acuity scores on the 5% contrast charts were significantly associated with
axial length (a surrogate measure for preoperative myopia). This relationship was not present
for other levels of contrast (100%, 10%, and 2.5%). While future studies comparing visual
acuity scores in the same patient before and after refractive surgery might provide additional
evidence of sensitivity, 5% low-contrast visual acuity appears to be the most sensitive to
visual function differences between LASIK and non-LASIK patients in this study. This is
based on the fact that we would expect LASIK patients to have poorer low-contrast visual
acuity than non-LASIK patients as the amount of preoperative myopia, i.e. axial length,
increases.
In summary, measuring 5% contrast visual acuity under dim illumination conditions
appears to be an acceptably sensitive and repeatable measurement of visual function in

110
LASIK patients. In addition, this is a task that is very familiar to patients, and we noted no
floor or ceiling effects for this measurement, i.e. no patients scored the maximum score and
all patients were able to achieve at least a minimum score on the test. While previous studies
of contrast sensitivity measurements after LASIK have shown that this measurement is
sensitive to differences in visual function,76 two commercially available contrast sensitivity
tests have been shown to lack acceptable test-retest repeatability and have the added
disadvantage of floor and ceiling effects. Based on these findings, we would recommend 5%
contrast visual acuity measurements in dim illumination in refractive surgery studies in
addition to the traditional, high-contrast, logMAR visual acuity testing.
111
CHAPTER 8
CONCLUSIONS
Several recent publications have reported the long-term outcomes of corneal laser
refractive surgery.26, 78-80 The safety of the procedure that was demonstrated in the initial
FDA studies, including very minor losses of best-spectacle-corrected visual acuity, seem
to be stable in the long-term. When evaluating efficacy, these studies all noted at least
minor amounts of what was termed “regression.” Two studies with 24 months of
postoperative follow-up found significant increases in myopia over the course of the
postoperative period in patients with higher levels of myopia.78, 80 In one of these studies,
only 28.6% of the higher myopes were within 1.00 D of the intended correction at the 24-
month visit.78 This same study also followed changes in corneal curvature and found a
significant increase in corneal curvature over the course of two years,78 indicating that all
of the change in refractive error is not due to axial length changes.
Retrospective studies of six and eight years of follow-up have also found
increases in myopia with long-term follow-up. Sekundo and co-workers found a shift
from –0.25 D to –0.88 D on average between one and six years after LASIK.26 This study
also indicated that 29% of the patients were unhappy with their uncorrected visual acuity.
While 52% of patients were within 1.00 D of the intended correction at one year, only
112
46% of patients were within 1.00 D of the intended correction at 6 years. Changes in
corneal curvature were not reported except to report that there was no evidence of
keratectasia.
In order to plan studies to evaluate the long-term results of LASIK and determine
if certain factors would predict problems like regression, pilot data in several areas would
be required. Conducting studies in a field of rapidly changing technology is difficult,
because researchers can end up studying a moving target. Based on the results of our
evaluation of FDA studies, it appears that changes in LASIK outcomes have stabilized, at
least for lower myopes.
It is also important in clinical research to fully evaluate the outcomes of a
procedure from the patient and the eye care practitioner’s point of view. Based on our
survey of postoperative LASIK patients and the focus groups discussions with eye care
practitioners we were able to determine that patients are primarily interested in
decreasing their dependence on optical aids. We also learned that eye care practitioners
are concerned about the impact of patient personality factors on LASIK outcomes.
Previous studies have rarely had spectacle/contact lens dependence as an outcome
measure, and we are aware of no studies that have specifically looked at the impact of
patient personality and/or mental illness such as depression and anxiety disorders on the
outcomes of elective, refractive surgical procedures. Future studies should evaluate both
of these areas.
In order to conduct these studies, we also needed to determine how to measure
refractive surgery outcomes. Refractive error measurements, visual acuity measurements
and measurements of quality of life need to provide results that are both repeatable and
113
sensitive to clinically meaningful changes. In the course of collecting these pilot data, we
were able to identify refractive error, visual function, and quality of life instruments that
should provide us with the tools we need to address the important unanswered questions
in corneal refractive surgery. We determined that cycloplegic autorefraction
measurements with the Grand Seiko autorefractor, high- and low-contrast visual acuity
measurements, and measurements with the vast majority of subscales of the NEI-RQL-42
can be made in a repeatable and valid manner in post-LASIK patients.
114
LIST OF REFERENCES
1. Gimbel HV, van Westenbrugge JA, Penno EE, Ferensowicz M, Feinerman GA,
Chen R. Simultaneous bilateral laser in situ keratomileusis: safety and efficacy.
Ophthalmology. 1999;106(8):1461-1467; discussion 1467-1468.
2. Marinho A, Pinto MC, Pinto R, Vaz F, Neves MC. LASIK for high myopia: one
year experience. Ophthalmic Surg Lasers. 1996;27(5 Suppl):S517-520.
3. McGhee CN, Craig JP, Sachdev N, Weed KH, Brown AD. Functional,
psychological, and satisfaction outcomes of laser in situ keratomileusis for high
myopia. J Cataract Refract Surg. 2000;26(4):497-509.
4. McDonald MB, Carr JD, Frantz JM, et al. Laser in situ keratomileusis for myopia
up to -11 diopters with up to - 5 diopters of astigmatism with the summit
autonomous LADARVision excimer laser system. Ophthalmology.
2001;108(2):309-316.
5. Knorz MC, Wiesinger B, Liermann A, Seiberth V, Liesenhoff H. Laser in situ

keratomileusis for moderate and high myopia and myopic astigmatism.
Ophthalmology. 1998;105(5):932-940.
6. Verdon W, Bullimore M, Maloney RK. Visual performance after photorefractive

keratectomy. A prospective study. Arch Ophthalmol. 1996;114(12):1465-1472.
7. Nakamura K, Bissen-Miyajima H, Toda I, Hori Y, Tsubota K. Effect of laser in

situ keratomileusis correction on contrast visual acuity. J Cataract Refract Surg.
2001;27(3):357-361.
8. Mutyala S, McDonald MB, Scheinblum KA, Ostrick MD, Brint SF, Thompson H.
Contrast sensitivity evaluation after laser in situ keratomileusis. Ophthalmology.
2000;107(10):1864-1867.
9. Perez-Santonja JJ, Sakla HF, Alio JL. Contrast sensitivity after laser in situ
keratomileusis. J Cataract Refract Surg. 1998;24(2):183-189.
115
10. Holladay JT, Dudeja DR, Chang J. Functional vision and corneal changes after
laser in situ keratomileusis determined by contrast sensitivity, glare testing, and
corneal topography. J Cataract Refract Surg. 1999;25(5):663-669.
11. Marcos S. Aberrations and visual performance following standard laser vision
correction. J Refract Surg. 2001;17(5):S596-601.
12. Mrochen M, Kaemmerer M, Mierdel P, Seiler T. Increased higher-order optical

aberrations after laser refractive surgery. A problem of subclinical decentration. J
Cataract Refract Surg. 2001;27(3):362-369.
13. Seiler T, Kaemmerer M, Mierdel P, Krinke HE. Ocular optical aberrations after
photorefractive keratectomy for myopia and myopic astigmatism. Arch
Ophthalmol. 2000;118(1):17-21.
14. Kiely PM, Smith G, Carney LG. The mean shape of the human cornea. Optica
Acta. 1982;29(8):1027-1040.
15. Applegate RA, Hilmantel G, Howland HC, Tu EY, Starck T, Zayac EJ. Corneal
first surface optical aberrations and visual performance. J Refract Surg.
2000;16(5):507-514.
16. Gauthier CA, Holden BA, Epstein D, Tengroth B, Fagerholm P, Hamberg-

Nystrom H. Assessment of high and low contrast visual acuity after
photorefractive keratectomy for myopia. Optom Vis Sci. 1998;75(8):585-590.
17. Bullimore MA, Olson MD, Maloney RK. Visual performance after
photorefractive keratectomy with a 6-mm ablation zone. Am J Ophthalmol.
1999;128(1):1-7.
18. Bailey MD, Olson MD, Bullimore MA, Jones LA, Maloney RK. The effect of
LASIK on best-corrected high and low contrast visual acuity. Optom Vis Sci.
2004;81(5):362-368.
19. Montes-Mico R, Charman WN. Mesopic contrast sensitivity function after

excimer laser photorefractive keratectomy. J Refract Surg. 2002;18(1):9-13.
20. Anera RG, Jimenez JR, Jimenez del Barco L, Bermudez J, Hita E. Changes in
corneal asphericity after laser in situ keratomileusis. J Cataract Refract Surg. Apr
2003;29(4):762-768.
116
21. el Danasoury MA. Prospective bilateral study of night glare after laser in situ
keratomileusis with single zone and transition zone ablation. J Refract Surg.
1998;14(5):512-516.
22. Schein OD, Vitale S, Cassard SD, Steinberg EP. Patient outcomes of refractive
surgery. The refractive status and vision profile. J Cataract Refract Surg.
2001;27(5):665-673.
23. Bailey MD, Mitchell GL, Dhaliwal DK, Boxer Wachler BS, Zadnik K. Patient
satisfaction and visual symptoms after laser in situ keratomileusis.
Ophthalmology. Jul 2003;110(7):1371-1378.
24. Chalita MR, Chavala S, Xu M, Krueger RR. Wavefront analysis in post-LASIK

eyes and its correlation with visual symptoms, refraction, and topography.
Ophthalmology. Mar 2004;111(3):447-453.
25. Chalita MR, Xu M, Krueger RR. Correlation of aberrations with visual symptoms
using wavefront analysis in eyes after laser in situ keratomileusis. J Refract Surg.
Nov-Dec 2003;19(6):S682-686.
26. Sekundo W, Bonicke K, Mattausch P, Wiegand W. Six-year follow-up of laser in

situ keratomileusis for moderate and extreme myopia using a first-generation
excimer laser and microkeratome. J Cataract Refract Surg. Jun 2003;29(6):1152-
1158.
27. Stern ME, Beuerman RW, Fox RI, Gao J, Mircheff AK, Pflugfelder SC. The
pathology of dry eye: the interaction between the ocular surface and lacrimal
glands. Cornea. 1998;17(6):584-589.
28. Battat L, Macri A, Dursun D, Pflugfelder SC. Effects of laser in situ

keratomileusis on tear production, clearance, and the ocular surface.
Ophthalmology. 2001;108(7):1230-1235.
29. Kim WS, Kim JS. Change in corneal sensitivity following laser in situ
30. Matsui H, Kumano Y, Zushi I, Yamada T, Matsui T, Nishida T. Corneal sensation

after correction of myopia by photorefractive keratectomy and laser in situ
117
31. Toda I, Asano-Kato N, Hori-Komai Y, Tsubota K. Laser-assisted in situ
keratomileusis for patients with dry eye. Arch Ophthalmol. 2002;120(8):1024-
1028.
32. Hovanesian JA, Shah SS, Maloney RK. Symptoms of dry eye and recurrent
erosion syndrome after refractive surgery. J Cataract Refract Surg.
2001;27(4):577-584.
33. Huang F, Sung-Huei T, Shih M, Chen FK. Effect of artificial tears on corneal
surface regularity, contrast sensitivity, and glare disability in dry eyes.
Ophthalmology. 2002;109:1934-1940.
34. Leaming DV. Practice styles and preferences of ASCRS members-2002 survey. J
Cataract Refract Surg. Jul 2003;29(7):1412-1420.
35. Duffey RJ, Leaming D. US trends in refractive surgery: 2002 ISRS survey. J
Refract Surg. May-Jun 2003;19(3):357-363.
36. Twa MD, Bullimore MA, roberts C. Patient reported quality of life after laser
assisted in-situ keratomileusis [aao abstract]. Optom Vis Sci. 2003;80:s12.
37. Merchea M, Tokarewicz A. Comparison of laser in situ keratomileusis outcomes

with the Nidek EC- 5000 and LSX LaserSight excimer lasers. J Refract Surg.
2001;17(2 Suppl):S246-249.
38. Seward MS, Oral D, Bowman RW, El-Agha MS, Cavanagh HD, McCulley JP.
Comparison of LADARVision and Visx Star S3 laser in situ keratomileusis
outcomes in myopia and hyperopia. J Cataract Refract Surg. Dec
2003;29(12):2351-2357.
39. Nuijts RM, Nabar VA, Hament WJ, Eggink FA. Wavefront-guided versus
standard laser in situ keratomileusis to correct low to moderate myopia. J
Cataract Refract Surg. Nov 2002;28(11):1907-1913.
40. Agresti A, NetLibrary Inc. Categorical data analysis. New York: Wiley; 1990.
41. Brunette I, Gresset J, Boivin JF, et al. Functional outcome and satisfaction after
photorefractive keratectomy. Part 2: survey of 690 patients. Ophthalmology.
2000;107(9):1790-1796.
118
42. Brunette I, Gresset J, Boivin JF, Boisjoly H, Makni H. Functional outcome and
satisfaction after photorefractive keratectomy. Part 1: development and validation
of a survey questionnaire. Ophthalmology. 2000;107(9):1783-1789.
43. Hays RD, Mangione CM, Ellwein L, Lindblad AS, Spritzer KL, McDonnell PJ.
Psychometric properties of the National Eye Institute-Refractive Error Quality of
Life instrument. Ophthalmology. Dec 2003;110(12):2292-2301.
44. McDonnell PJ, Mangione C, Lee P, et al. Responsiveness of the National Eye
Institute Refractive Error Quality of Life instrument to surgical correction of
refractive error. Ophthalmology. Dec 2003;110(12):2302-2309.
45. Anderson NJ, Edelhauser HF, Sharara N, et al. Histologic and ultrastructural
findings in human corneas after successful laser in situ keratomileusis. Arch
Ophthalmol. 2002;120(3):288-293.
46. Balazsi G, Mullie M, Lasswell L, Lee PA, Duh YJ. Laser in situ keratomileusis
with a scanning excimer laser for the correction of low to moderate myopia with
and without astigmatism. J Cataract Refract Surg. 2001;27(12):1942-1951.
47. Miller AE, McCulley JP, Bowman RW, Cavanagh HD, Wang XH. Patient
satisfaction after LASIK for myopia. Clao J. Apr 2001;27(2):84-88.
48. Hill JC. An informal satisfaction survey of 200 patients after laser in situ
keratomileusis. J Refract Surg. 2002;18(4):454-459.
49. Pop M, Payette Y. Photorefractive keratectomy versus laser in situ keratomileusis:

a control-matched study. Ophthalmology. 2000;107(2):251-257.
50. Bailey MD, Mitchell GL, Dhaliwal DK, Boxer Wachler BS, Zadnik K. Patient
satisfaction and visual symptoms after laser in situ keratomileusis.
Ophthalmology. 2003;110:1371-1378.
51. Miller MD, Ferris DG. Measurement of subjective phenomena in primary care
research: the Visual Analogue Scale. Fam Pract Res J. 1993;13(1):15-24.
52. Khan-Lim D, Craig JP, McGhee CN. Defining the content of patient
questionnaires: reasons for seeking laser in situ keratomileusis for myopia. J
Cataract Refract Surg. 2002;28(5):788-794.
119
53. Festinger L. A theory of cognitive dissonance. 1st ed. Stanford: Stanford
University Press; 1968.
54. Cobb LA, Thomas GI, Dillard DH, Merendino KA, Bruce RA. An evaluation of
internal-mammary-artery ligation by a double-blind technic. New Eng J Med.
1959;260(22):1115-1118.
55. Homer JJ, Sheard CE, Jones NS. Cognitive dissonance, the placebo effect and the
evaluation of surgical results. Clin Otolaryngol. 2000;25(3):195-199.
56. Sweeney JC, Hausknecht D, Soutar GN. Cognitive dissonance after a purchase: a
multidimensional scale. Psychology & Marketing. 2000;17(5):369-385.
57. Bullimore MA, Jones LA, Moeschberger ML, Zadnik K, Payor RE. A
retrospective study of myopia progression in adult contact lens wearers. Invest
Ophthalmol Vis Sci. Jul 2002;43(7):2110-2113.
58. Gwiazda J, Hyman L, Hussein M, et al. A randomized clinical trial of progressive

addition lenses versus single vision lenses on the progression of myopia in
children. Invest Ophthalmol Vis Sci. Apr 2003;44(4):1492-1500.
59. Walline JJ, Mutti DO, Jones LA, et al. The contact lens and myopia progression
(CLAMP) study: design and baseline data. Optom Vis Sci. Apr 2001;78(4):223-
233.
60. Zadnik K, Manny RE, Yu JA, et al. Ocular component data in schoolchildren as a
function of age and gender. Optom Vis Sci. Mar 2003;80(3):226-236.
61. Zadnik K, Mutti DO, Adams AJ. The repeatability of measurement of the ocular
components. Invest Ophthalmol Vis Sci. Jun 1992;33(7):2325-2333.
62. Sheng H, Bottjer CA, Bullimore MA. Ocular component measurement using the
Zeiss IOLMaster. Optom Vis Sci. 2003;In press.
63. Siganos DS, Popescu C, Bessis N, Papastergiou G. Autorefractometry after laser

in situ keratomileusis. J Cataract Refract Surg. Jan 2003;29(1):133-137.
64. Salchow DJ, Zirm ME, Stieldorf C, Parisi A. Comparison of objective and
subjective refraction before and after laser in situ keratomileusis. J Cataract
Refract Surg. 1999;25(6):827-835.
120
65. Rosa N, Iura A, Romano M, Verolino G, Romano A. Correlation between
automated and subjective refraction before and after photorefractive keratectomy.
J Refract Surg. 2002;18(4):449-453.
66. Oyo-Szerenyi KD, Wienecke L, Businger U, Schipper I.

Autorefraction/autokeratometry and subjective refraction in untreated and
photorefractive keratectomy-treated eyes. Arch Ophthalmol. 1997;115(2):157-
164.
67. Russell GE, Bergmanson JP, Barbeito R, Cross WD. Differences between
objective and subjective refractions after radial keratotomy. Refract Corneal Surg.
Jul-Aug 1992;8(4):290-295.
68. Thibos LN, Wheeler W, Horner D. Power vectors: an application of Fourier

analysis to the description and statistical analysis of refractive error. Optom Vis
Sci. Jun 1997;74(6):367-375.
69. Thibos LN, Horner D. Power vector analysis of the optical outcome of refractive
surgery. J Cataract Refract Surg. Jan 2001;27(1):80-85.
70. Bland JM, Altman DG. Statistical methods for assessing agreement between two
methods of clinical measurement. Lancet. 1986;1(8476):307-310.
71. Benjamin WJ, Borish IM. Borish's clinical refraction. Philadelphia: W.B.
Saunders; 1998.
72. Davies LN, Mallen EA, Wolffsohn JS, Gilmartin B. Clinical evaluation of the
Shin-Nippon NVision-K 5001/Grand Seiko WR-5100K autorefractor. Optom Vis
Sci. Apr 2003;80(4):320-324.
73. McDonnell PJ, Lee P, Spritzer K, Lindblad AS, Hays RD. Associations of
presbyopia with vision-targeted health-related quality of life. Arch Ophthalmol.
Nov 2003;121(11):1577-1581.
74. Nichols JJ, Mitchell GL, Saracino M, Zadnik K. Reliability and validity of
refractive error-specific quality-of-life instruments. Arch Ophthalmol. Sep
2003;121(9):1289-1296.
121
75. Pesudovs K, Hazel CA, Doran RM, Elliott DB. The usefulness of Vistech and
FACT contrast sensitivity charts for cataract and refractive surgery outcomes
research. Br J Ophthalmol. Jan 2004;88(1):11-16.
76. Lee YC, Hu FR, Wang IJ. Quality of vision after laser in situ keratomileusis. J
Cataract Refract Surg. 2003;29:769-777.
77. Klein R, Klein BE, Moss SE, DeMets D. Inter-observer variation in refraction and
visual acuity measurement using a standardized protocol. Ophthalmology. Nov
1983;90(11):1357-1359.
78. Han HS, Song JS, Kim HM. Long-term results of laser in situ keratomileusis for
high myopia. Korean J Ophthalmol. Jun 2000;14(1):1-6.
79. Pietila J, Makinen P, Pajari T, et al. Eight-year follow-up of photorefractive

keratectomy for myopia. J Refract Surg. Mar-Apr 2004;20(2):110-115.
80. Magallanes R, Shah S, Zadok D, et al. Stability after laser in situ keratomileusis in
moderately and extremely myopic eyes. J Cataract Refract Surg.
2001;27(7):1007-1012.
122

Eximer BL Osu1086104689

Uploaded by

Copyright:

Available Formats

Eximer BL Osu1086104689

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Eximer BL Osu1086104689

Uploaded by

Copyright:

Available Formats

THE ASSESSMENT OF POSTOPERATIVE REFRACTIVE SURGERY

PATIENTS IN CLINICAL RESEARCH

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

Melissa Diane Bailey, OD, MS

The Ohio State University

Professor Joseph A. Barr

Associate Professor Mark A. Bullimore

occurred with changes in laser technology. The repeatability of autorefraction

Error Quality of Life (NEI-RQL-42) Questionnaire was evaluated in laser in situ

keratomileusis (LASIK) patients. We also compared the performance of the NEI-RQL-42

when subjects completed the surveys simultaneously. Finally, we evaluated patient

generating hypotheses for future research studies.

We found that cycloplegic autorefraction measurements with the Grand Seiko

acceptable levels of repeatability, and 5% contrast visual acuity appears to be sensitive to

dry eye symptoms.

patient’s view of a LASIK procedure?

I would like to acknowledge my co-authors of the studies contained in this dissertation:

Karla Zadnik, OD, PhD Deepinder K. Dhaliwal, MD

Michael D. Twa, OD, MS Michael D. Olson, OD, PhD

Mark A. Bullimore, MCOptom, PhD Joseph Shovlin, OD

G. Lynn Mitchell, MAS Stephen Pascucci, MD

Lisa A. Jones, PhD Brian Boxer-Wachler, MD

NIH/NEI grant #T32-EY013359 and #R24-EY14792, American Optometric Foundation

March 1, 1975 Born – Gillette, Wyoming

1997 B.S. Neuroscience, Muskingum College

2001 O.D. The Ohio State University

2001 M.S. Vision Science, The Ohio State University

2001 Graduate Teaching Associate, The Ohio State University

2001- present Postdoctoral Fellow, The Ohio State University

Major Field: Vision Science

1.3 Refractive surgery recommendations: 2002 ASCRS Survey .....................................9

1.4 Refractive surgery recommendations: 2002 ISRS Survey........................................10

2.4 Proportion of low myopes in each category of laser with postoperative

4.1 Reasons for recommending LASIK to a friend ........................................................50

5.2 Mean and standard deviation of refractive error measurements...............................63

5.3 Agreement between subjective refraction measurements and autorefraction

6.1 Repeatability of the NEI-RQL-42 subscales across two visits .................................78

6.6 Comparison of preoperative and postoperative NEI-RQL-42 scores across

7.1 Contrast sensitivity from a report by Lee, Hu and Wang (2003)..............................92

7.2 Descriptive statistics by group ..................................................................................96

7.4 Within examiner test-retest repeatability ..................................................................98

7.5 Between examiner test-retest repeatability ...............................................................99

1.1 Associations of visual function after LASIK..............................................................2

3.1 Photograph of flip chart for participants’ comments during discussion of a

experienced an explosion of technological advances in the last few decades. So many

design of future studies of LASIK outcomes.

1.1 Common unsuccessful LASIK outcomes

Previous studies overwhelmingly support two conclusions about the LASIK

procedure when it is used to correct myopia:

and best-corrected, high-contrast visual acuity after LASIK.1

• Patient reported satisfaction is generally high, ranging from 82 to 98%.2-5

Nonetheless, some patients have subjective complaints of decreased vision after

Consequently, studies of both (photorefractive keratectomy) PRK and LASIK have

high-contrast visual acuity.6, 7 Figure 1 is an outline of a series of relationships. These

relationships summarize studies of several outcomes after LASIK-induced changes in

LASIK have not found similar 2 ?

results. Some studies have Night Vision