Full text of DOI:10.1007/s12009-001-0047-4

A B S T R A C T
A prospective, nonrandomized, unmasked, 1-year clinical trial was
performed in 2482 eyes to evaluate the safety and efficacy of laser
in situ keratomileusis (LASIK) for myopia and myopic astigmatism
using a new excimer laser system. The study found that LASIK was
a safe and effective method for correcting myopia and myopic
astigmatism in patients with sphere to –15 diopters (D) and cylin-
der to 5 D.
O R I G I N A L A R T I C L E
Prospective LASIK Trial for Myopia
and Myopic Astigmatism: 1-Year Results
Frederic B. Kremer, MD, George Pronesti, MD,
John Sloat, OD, & Peter Calla, OD
n North America alone, myopia affects an estimated
I
75 million people, and the prevalence is increas-
ing.1-5 The recent population-based survey in Beaver
Dam, Wis, showed that 14.4% of phakic adults older
than 75 years had myopia, but 42.9% of phakic adults
aged 43 to 54 years had myopia greater than or equal
to 0.5 diopters (D).6 Several other population-based
studies suggest that this rise in myopia prevalence is
largely due to an adaptive lengthening of the eye that
results from a significant increase in close-up work.1,7,8
The shape of the lens may also change as the tone of
the ciliary muscle and zonular fibers relaxes with age,
and this may also affect refraction.9
More than 90% of myopes have low to moderate
myopia (to –6 D), and about two-thirds also have some
degree of astigmatism. In 1993, we began offering
laser in situ keratomileusis (LASIK) for myopes and
individuals with myopic astigmatism who desired sur-
gical correction. We used a new excimer laser and per-
formed the procedures initially under Institutional
Review Board (IRB) approval and subsequently under
a slightly modified protocol approved for an investiga-
tional device exemption (IDE). We studied this proce-
dure in an effort to validate the theory that LASIK
promotes rapid visual recovery with minimal or no
postoperative pain, haze, or regression. We also
assessed the incidence and clinical significance of
complications, especially those unique to the LASIK
Reprints:
Frederic B. Kremer, MD, Kremer Laser Eye Center, 200 Mall Blvd., King of Prussia, PA
19406.
The authors are from the Kremer Laser Eye Center, King of Prussia, Pa.
Dr. Kremer has a financial interest in the excimer laser system used in this study.
Acknowledgment
The authors thank Mike Dayton, Mike Blair, and Maureen Lyden, MS, for assistance
with data analysis and Dianne Herrin for assistance with manuscript preparation.
ANN OPHTHALMOL. 2001;33(4):315–322
315

procedure, such as lost flaps, astigmatism caused by
flap folds or irregularities, and “melted” flaps result-
ing from an infiltration of white blood cells into the
stroma-flap interface.
This laser has a computer-controlled aperture
assembly that shapes the laser beam as it passes
through tissue. This aperture assembly consists of an
adjustable iris for the spherical (myopic) correction
and a slot with adjustable width and angle for the
oblong (astigmatic) correction, which affects just one
meridian of the eye. This laser differs from other
lasers of its type in several ways. Its operating micro-
scope has a fully automated foot control that enables
the surgeon to zoom in and out during cap manipula-
tion and ablation, the software was developed specifi-
cally for LASIK, and the laser allows astigmatic
ablations within a full 6-mm optical zone.
Following ablation, the operating surgeon irrigated
the bed and interface and replaced the flap without any
sutures. A clear shield was then placed over the eye. To
prevent trauma to the eyes and potential cap displace-
ment, the patient was instructed to wear the shield and
keep the eye closed for the first day. Thereafter,
patients wore the eye shield for 1 week while sleeping.
We also prescribed 1% prednisolone acetate sterile oph-
thalmic solution twice a day and a broad-spectrum top-
ical antibiotic 4 times a day, both for 4 days.
Patients & Methods
A prospective, nonrandomized, unmasked, clinical
trial was conducted initially under IRB approval
(cohort 1) and subsequently, starting July 1, 1996,
under a slightly modified IDE-approved protocol
(cohort 2). Patients who came to one of our laser eye
centers beginning May 1, 1993, seeking surgical cor-
rection of myopia were eligible for the study if they
were 18 years old or older, had primary myopia
between –1.0 D and –15.0 D with or without astigma-
tism up to 5.0 D, and had stable refraction defined as
less than 0.5 D shift for the 30 days before surgery.
Subjects also needed to want to be free of spectacles or
contact lenses and be willing to comply with all post-
operative follow-up visits. Patients refrained from soft
contact lens wear for 2 weeks before the preoperative
evaluation or hard contact lens wear for 3 weeks
before the preoperative evaluation.
Patients not meeting these criteria, as well as
patients with any of the following conditions, were
excluded from the study: active ocular or systemic
infection, severe dry eye syndrome, Fuchs dystrophy,
anterior basement membrane dystrophy, keratoconus
associated with thinning, and/or central corneal scars
that affected visual acuity. Patients receiving chronic
topical steroid therapy and those with corneas too thin
to permit the desired correction also were excluded.
Preoperatively, complete ocular and medical his-
tories were taken. Best-corrected visual acuity
(BCVA) and uncorrected visual acuity (UCVA) mea-
surements and manifest and cycloplegic refractions
were obtained. Dilated fundus examinations, bright-
ness acuity testing, pupillary examinations, ker-
atometry, tonometry, pachymetry, and topography
were performed. Refractive and ocular stability was
documented.
On every operative day, we calibrated the laser and
tested beam quality. Each patient was prepared with a
povidone-iodine (Betadine) eyelid scrub before surgi-
cal draping. We then administered approximately 5
tetracaine drops, placed a lid speculum, and simulta-
neously vacuumed and irrigated the eye with bal-
anced salt solution. After placing the keratome
suction ring on the eye, we used a microkeratome
(Chiron Automated Corneal Shaper, Bausch & Lomb
Surgical, Claremont, Calif) to create a nasally hinged
corneal flap 160 µm thick and approximately 8 mm in
diameter. The incision began at the temporal side and
stopped 90% of the way across. After swinging the cap
nasally to expose the stromal bed, the patient fixated
on a light coaxial with the laser beam, and the stromal
bed was ablated with a 193-nm wavelength excimer
laser (Kremer Excimer Laser, Kremer Laser Center,
King of Prussia, Pa), using a typical fluence of about
140 mJ/cm.2
Patients were evaluated on postoperative day 1,
week 1, and months 1, 3, 6, and 12 at the Kremer Laser
Eye Centers and qualified comanaging sites. Qualified
sites performed standardized postoperative evalua-
tions and completed standardized follow-up forms.
Postoperative evaluations included keratometry,
uncorrected and best spectacle-corrected visual
acuities, manifest refraction, and a thorough slit-lamp
examination for assessment of corneal clarity and the
anterior chamber and lens status. All patient com-
plaints, complications, and adverse reactions were
recorded. Cycloplegic refractions were performed at
12 months. All measurements were taken by
optometrists or ophthalmologists other than the
authors and were entered into a computer database.
The database was closed for purposes of this analysis
on November 20, 1997.
We evaluated the effectiveness of the procedure
based on the absolute manifest refraction over time
(sphere and cylinder); improvement in UCVA; reduc-
tion in the spherical equivalent refraction, including
independent analyses of both the spherical and cylin-
drical components; and accuracy of the achieved
spherical equivalent. We evaluated the predictability
of the procedure and device by assessing the propor-
tion of eyes experiencing deviations from the intend-
ed correction within 0.5 D, 1.0 D, and 2.0 D.
Stability was defined as a change of 1.0 D or more for
2 consecutive visits spaced 3 months apart.
We evaluated safety in terms of a loss in BCVA of
greater than 2 lines, a BCVA worse than 20/40, BCVA
worse than 20/25 in eyes that were 20/20 or better pre-
operatively, and all symptoms and complications. We
asked patients to describe their symptoms as bother-
some or not bothersome, and we also recorded this
information. We recorded the results of all eyes re-
treated for undercorrection. Per instruction from the
316
ANN OPHTHALMOL. 2001;33(4)

Food and Drug Administration (FDA), all eyes treated
for overcorrection were excluded from the analysis.
The protocol for cohort 1 differed from that of
cohort 2 in 2 primary ways. Initially, the intended cor-
rection typically was calculated by subtracting 20%
from the theoretical correction, but with time and
experience we reduced and ultimately eliminated this
safety factor. All patients in cohort 2 were treated
using an intended correction equivalent to the ideal
correction. Also, a second investigator (G.P.) joined the
study during the IDE phase (cohort 2).
T A B L E 1
Key Demographic and Visual Characteristics
Cohort 1
(n=616)
Cohort 2
Characteristic
Male, No. (%)
(n=704)
359 (51.0)
36.3
329 (53.4)
38.1
Mean age, y
Myopia, No. (%) of eyes
487/1140
(42.7)
630/1342
(46.9)
Myopic astigmatism,
No. (%) of eyes
653/1140
(57.3)
712/1342
(53.1)
To determine whether the protocol, degree of cor-
rection (less than 7 D preoperative spherical equiva-
lent v greater than or equal to 7 D spherical
equivalent), or type of correction (myopia only v
myopic astigmatism) influenced the visual outcome,
we performed statistical comparisons of the key safe-
ty and efficacy outcome measures at the 0.05 signifi-
cance level at the point of stability (the 6-month
interval). Depending on the analysis, we used either
the χ2 or Cochran-Mantel-Haenszel statistic, control-
ling for protocol or preoperative refraction when
appropriate. To determine whether the visual results
remained stable over time, key 6- and 12-month out-
come measures were also compared using the
Cochran-Mantel-Haenszel statistic, controlling for
protocol. So that we could independently assess both
the position (axis) and magnitude of any residual
astigmatism, we performed vector analysis using the
widely accepted Holladay method.10 Finally, we also
compared the accuracy of the surgical correction in a
series of eyes with both manifest and cycloplegic
refractive data at 12 months using the McNemar test.
This enabled us to determine whether we induced any
significant refractive changes in the cornea that were
masked by the patient’s natural accommodation.
Monovision eyes were excluded from all analyses.
In patients with monovision, the goal of LASIK is to
achieve slightly different visual acuities in each eye—
1 for near (reading) and 1 for far vision—to accommo-
date for presbyopia.
Fig 1.—Mean manifest sphere over time, before LASIK and 1, 3, 6, and 12 months
after the procedure. Parenthetical values are the SD.
Results
Demographics. A total of 2482 eyes were treated
under both protocols. We treated slightly more males
than females and more eyes for myopic astigmatism
than for myopia only. Table 1 summarizes the key
demographic characteristics, which were similar for
both cohorts. There were 1402 eyes with 6 months of
follow-up and 957 eyes with 12 months of follow-up.
Preoperatively, 58.5% and 72% of eyes in cohorts 1
and 2, respectively, had myopia (sphere) less than
–7 D, and 96% and 94% of eyes in these cohorts also
had astigmatism (cylinder) less than 3 D. Less than 1%
of all eyes combined had UCVA of 20/40 or better.
Efficacy. Mean sphere went from –6.03 D preopera-
tively to –0.40 D at 6 months; mean cylinder went from
0.87 D to 0.53 D, respectively. Figures 1 and 2 show the
absolute sphere and cylinder ( SD) at all intervals.
Six months after LASIK, 88% of all eyes had UCVA
Fig 2.—Mean manifest cylinder over time, before LASIK and 1, 3, 6, and 12 months
after the procedure. Parenthetical values are the SD.
of 20/40 or better. The UCVA was statistically better in
eyes from cohort 2 (Table 2) because cohort 1 eyes were
intentionally undercorrected. The manifest refraction
spherical equivalent was accurate to within 1.0 D of
the intended correction in 86.7% of all eyes (Table 2).
Table 2 shows that refractive outcomes remained sta-
ANN OPHTHALMOL. 2001;33(4)
317

T A B L E 2
Key Efficacy Results 6 and 12 Months after LASIK in All Eyes*
Cohort 1, No. (%)
Cohort 2, No. (%)
P†
Efficacy
Measure
Cohort
Comparison
6- and 12-mo
Comparison
6 mo
12 mo
6 mo
12 mo
UCVA: ‡
≥ 20/20
212/668 (31.7)
579/668 (86.7)
202/612 (32.7)
535/612 (87.4)
199/491 (40.5)
446/491 (90.3)
87/218 (39.9)
201/218 (92.2)
.002
.029
.840
.523
≥ 20/40
MRSE: §
0.5 D
494/737 (67.0)
631/737 (85.6)
710/737 (96.3)
431/688 (62.6)
588/688 (85.5)
669/688 (97.2)
472/665 (71.0)
584/665 (87.8)
650/665 (97.7)
199/269 (74.0)
249/269 (92.6)
266/269 (98.9)
.111
.226
.123
.401
.231
.091
1 D
2 D
*Includes re-treated eyes.
2
†Comparison between cohorts at the point of stability (6 months) used the χ statistic, and comparison
between 6- and 12-month results used the Cochran-Mantel-Haenszel statistic, controlling for protocol.
‡All eyes treated for monovision were excluded from the analysis for uncorrected visual acuity (UCVA), since
undercorrection is intentional in this subpopulation.
§MRSE indicates manifest refraction spherical equivalent and indicates percentage of eyes corrected to with-
in 0.5 diopters (D), 1.0 D, and 2.0 D of the intended correction.
ble over time, as there were no statistically significant
differences between the 6- and 12-month key efficacy
results when controlled for protocol.
The median shift in cylinder axis in all eyes was 25°
and 35° for cohorts 2 and 1, respectively. However, the
highest shifts in astigmatic axes tended to occur in
eyes with the lowest degrees of residual cylinder;
68.8% and 70.9% of eyes treated for myopia only in
cohorts 2 and 1, respectively, had residual cylinder
from 0 to less than 1 D at 6 months. This is clinically
important because measurements of the astigmatic
axis in eyes with small amounts of astigmatism are
highly inaccurate and tend to fluctuate widely.
To determine whether surgery induced any signifi-
cant refractive changes in the cornea that were
masked by the patient’s natural accommodation, we
compared the accuracy of the spherical equivalent
based on manifest v cycloplegic refractions. As Table 4
shows, a greater proportion of eyes achieved manifest
refractions within 1.0 D than cycloplegic refractions
within 1.0 D of the intended correction. However,
there was no difference in the proportions of eyes
achieving within 0.5 D or 2.0 D of the intended cor-
rection, regardless of the method of refraction.
Safety. In this study, 2.3% of all eyes (22/957) lost 2
or more lines of BCVA 12 months after surgery. Only
1.6% of eyes (n=15) had BCVA worse than 20/40. There
were no statistically significant differences in rates of
BCVA losses greater than 2 lines or BCVA worse than
20/40 between cohorts or between the 6- and 12-month
results. As Table 5 shows, most eyes that lost BCVA
had higher refractive errors (7 D or greater) and
required a greater degree of laser correction.
The LASIK surgery was most effective in eyes with
low to moderate myopia. Both UCVA and the accuracy
of the correction were significantly better in eyes with
less than 7 D of preoperative spherical equivalent
(Table 3). When we assessed cohort 2 eyes to determine
whether the type of correction (myopia only v myopic
astigmatism) influenced the refractive and visual out-
comes, there was one statistically significant differ-
ence. Eyes with preoperative spherical equivalent of 7
D or greater treated for myopic astigmatism were less
likely to achieve 20/20 UCVA than were eyes with pre-
operative spherical equivalent greater than or equal to
7 D that were treated for myopia only (P=.057,
Cochran-Mantel-Haenszel statistic controlling for pre-
operative spherical equivalent). However, there was no
statistically significant difference between the propor-
tions of these eyes achieving 20/40 uncorrected vision
(P=.808, Cochran-Mantel-Haenszel statistic control-
ling for preoperative spherical equivalent).
When we assessed the spherical and cylindrical
components of the 6-month postoperative refraction
separately, we found that eyes treated for myopia only
had the most residual cylinder. In all eyes treated for
myopia only (both cohorts), mean cylinder overcorrec-
tion at 6 months was approximately 0.75 D (cohort 1:
0.71 D 0.68, n=657; cohort 2: 0.72 D 0.64, n=713).
By 12 months, less than 1% of eyes (4/414) treated for
myopia only had cylinder increases greater than 2 D.
As expected, symptoms were most prevalent in
318
ANN OPHTHALMOL. 2001;33(4)

T A B L E 3
Key Efficacy Results at 6 Months Stratified by Preoperative Manifest Refraction
Efficacy
Measure
Cohort 1, No. (%)
Cohort 2, No. (%)
<7 D
≥ 7 D
<7 D
≥ 7 D
P*
UCVA: †
≥ 20/20
176/447 (39.4)
407/447 (91.1)
36/221 (16.3)
172/221 (77.8)
185/371 (49.9)
358/371 (96.5)
14/120 (11.7)
88/120 (73.3)
.001
.001
≥ 20/40
MRSE: ‡
0.5 D
356/483 (73.7)
439/483 (90.9)
480/483 (99.4)
138/254 (54.3)
192/254 (75.6)
230/254 (90.6)
374/477 (78.4)
446/477 (93/5)
475/477 (99.6)
98/188 (52.1)
138/188 (73.4)
175/188 (93.1)
.001
.001
.001
1 D
2 D
*P values are identical for the comparisons between eyes with <7 diopters (D) of preoperative spherical
equivalent and ≥ 7 D eyes within each protocol population. Cochran-Mantel-Haenszel statistic comparing
outcomes by preoperative spherical equivalent, controlling protocol.
†All eyes treated for monovision were excluded from the analysis for uncorrected visual acuity (UCVA), since
undercorrection is intentional in this subpopulation.
‡MRSE indicates manifest refraction spherical equivalent and indicates percentage of eyes corrected to with-
in 0.5 D, 1.0 D, and 2.0 D of the intended correction.
T A B L E 4
Accuracy of Correction: Manifest v Cycloplegic Refraction
Spherical Equivalent
(diopters)
Manifest Refraction,
No. (%)*
Cycloplegic Refraction,
No. (%)*
P†
0.5 D
1.0 D
2.0 D
158/224 (70.5)
211/224 (94.2)
222/224 (99.1)
160/224 (71.4)
199/224 (88.8)
221/224 (98.7)
.7440
.0075
1.000
*224 eyes had both manifest and cycloplegic refractive data at 12 months.
†Derived from the McNemar test.
eyes with high spherical equivalents (7 D or greater)
and in eyes with residual refractive errors. The inci-
dence of symptoms that patients considered bother-
some ranged from 0.1% to 4.1% for both cohorts and
included glare, halos, trouble with night driving, dou-
ble vision/ghost images, foreign body sensation, anxi-
ety, and pain. The most frequent, bothersome
symptom was difficulty with right driving. However,
as with all other symptoms, this symptom was report-
ed much less frequently by cohort 2 patients than by
those in cohort 1. We also anecdotally observed that
foreign body sensations were more prevalent in
patients with prior dry eye syndrome. The study did
not measure the extent to which the symptoms may
have resolved in the presence of spectacle correction,
as we would expect.
There were 2 types of intraoperative complications:
incomplete flaps and hingeless flaps. In 5 cases, we
aborted the LASIK procedure; 4 of these cases were
due to a microkeratome gear malfunction that devel-
oped gradually. We later treated all of these eyes with-
out sequelae. In 29 cases, the microkeratome cut all
the way through and rendered the flap hingeless;
there were no associated sequelae.
We noted one malaligned cap during the first 24
postoperative hours and were able to reposition it
without sequelae. Corneal edema was present
between week 1 and month 1 in 73 eyes, and all of
these cases later resolved.
Additional complications noted at 1, 3, 6, and 12
months postoperatively are shown in Table 6. Overall,
there was a lower incidence of epithelium in the inter-
ANN OPHTHALMOL. 2001;33(4)
319

T A B L E 5
Loss of BCVA at 6 Months in Eyes with Low-to-Moderate and High Preoperative Spherical Equivalent*
Cohort 1, No. (%)
Cohort 2, No. (%)
BCVA Measure
<7 D
≥ 7 D
<7 D
≥ 7 D
P†
Loss of ≥ 2 lines
2/483 (0.4)
18/254 (7.1)
2/477 (0.4)
1/477 (0.2)
9/477 (2.0)
12/188 (6.3)
<.001
<20/40
0/483 (0)
11/254 (4.3)
3/188 (1.6)
<.001
<.001
<20/25 with ≥ 20/20 preop
8/469 (2.0)
17/183 (9.0)
17/188 (9.0)
*BCVA indicates best-corrected visual acuity; D, diopters; and preop, preoperatively.
†Cochran-Mantel-Haenszel statistic was used to compare preoperative spherical equivalent groups, control-
ling for protocol.
T A B L E 6
Incidence of Postoperative Complications
Cohort 2, No. (%)
Cohort 1, No. (%)
Complication
Mo 1
Mo 3
Mo 6
Mo 12
Mo 1
Mo 3
Mo 6
Mo 12
Corneal edema ≥ 1 mo
2/873
(0.2)
1/814
(0.1)
0/657
(0)
0/308
(0)
14/804
(1.7)
4/910
(0.4)
0/788
(0)
0/752
(0)
Corneal infiltrate
0/873
(0)
0/814
(0)
0/657
(0)
0/308
(0)
0/804
(0)
0/910
(0)
1/788
(0.1)
0/752
(0)
Epithelium in interface,
central
1/873
(0.1)
1/814
(0.1)
0/657
(0)
0/308
(0)
2/741
(0.3)
0/859
(0)
1/750
(0.1)
0/687
(0)
Epithelium in interface,
peripheral
1/873
(0.1)
0/814
(0)
1/657
(0.2)
0/308
(0)
8/741
(1.1)
12/859
(1.4)
15/750
(2)
10/687
(1.5)
Epithelial defect, central
Epithelial defect, peripheral
Interface foreign bodies
Cap striae
1/873
(0.1)
1/814
(0.1)
0/657
(0)
0/308
(0)
2/741
(0.3)
1/859
(0.1)
2/750
(0.3)
0/687
(0)
1/873
(0.1)
0/814
(0)
1/657
(0.2)
0/308
(0)
1/741
(0.1)
1/859
(0.1)
0/750
(0)
2/687
(0.3)
126/873 128/814 95/657
(14)
27/308
(9)
71/741
(9.6)
78/859
(9.1)
50/750
(6.7)
43/687
(6.3)
(14)
(15)
9/873
(1.0)
6/814
(0.7)
5/657
(0.8)
1/308
(0.3)
3/741
(0.4)
7/859
(0.8)
5/750
(0.7)
3/687
(0.4)
Pain
4/873
(0.5)
4/814
(0.5)
3/657
(0.5)
2/308
(0.7)
4/741
(0.5)
1/859
(0.1)
3/750
(0.4)
2/687
(0.3)
Retinal detachment
0/873
(0)
0/814
(0)
0/657
(0)
1/308
(0.3)
0/804
(0)
0/910
(0)
0/788
(0)
1/752
(0.1)
face of the stromal bed and the underside of the flap
in cohort 2 eyes, as compared with cohort 1 eyes. We
believe this was due to our increased surgical experi-
ence. When the epithelium was in the periphery of the
interface, it did not influence vision and we did not
remove it. When the epithelium was central in the
interface, we removed it. The foreign bodies in the
interface were observed under slit-lamp examination
320
ANN OPHTHALMOL. 2001;33(4)

and had no clinical sequelae. In some instances, cap
striae (at least 1% incidence at all intervals) was asso-
ciated with small BCVA decreases. Two retinal detach-
ments occurred at 11 months and more than 1 year
after surgery, suggesting that they were unrelated to
surgery. Several patients also had early cataracts, but
these cataracts were noted before surgery and were
therefore not iatrogenic.
Re-treatments. The re-treatment rate for undercor-
rections in the cohort 2 population was 3.5% (47/1342).
This re-treatment rate was higher in cohort 1 eyes due
to the intentional undercorrection planned under this
earlier protocol. In cohort 1, 14.2% of eyes (162/1140)
were re-treated for undercorrection. We also re-treated
19 eyes in cohort 2 and 25 eyes in cohort 1 for overcor-
rection, although as stated earlier, these eyes were
excluded from the data analysis.
erative myopia was –1.5 D 0.42 and –0.88 D 0.70 in
the PRK and LASIK groups, respectively. In addition,
3 of the trials14-16 showed that efficacy outcomes are
generally similar between PRK and LASIK after 6
months and up to 12 months, although additional con-
trolled comparisons are needed to determine the long-
term outcomes of the 2 procedures.
In this series, there were no vision-threatening
complications; there were no corneal infections, lost or
melted caps, anterior chamber perforations, or retinal
vascular accidents. Nevertheless, it is important to
realize that the potential for serious complications
exists, and each stage of the LASIK procedure must be
performed with care and control. We learned that one
of the critical and sometimes limiting factors is micro-
keratome function. In addition, lack of fixation during
ablation can lead to decentered ablations, which can
induce astigmatism and ultimately cause glare, ghost
images, and diplopia. With time, we improved our cen-
tering strategy, and this may be the reason why the
incidence of glare, halos, and problems with night dri-
ving were lower in the second cohort.
It should be noted that, in this study population,
there was a loss of patients with time. At 6 months,
1402 eyes were evaluated, representing an account-
ability of 81.1% for cohort 2 and a somewhat lower
accountability for cohort 1. To determine whether this
affected the validity of the data, we performed 24 addi-
tional intra-data-set statistical comparisons. We com-
pared safety and efficacy variables resulting from
assessments at 6 months, at the last visit for eyes of
patients who did not return for follow-up before the 6-
month visit, and at the last visit for eyes not yet due
for a 6-month evaluation. Similar comparisons were
performed at the 12-month interval. These analyses
demonstrated that the data are reliable.
We did observe that, at the upper end of the refrac-
tive range (–10 to –13 D), refraction took longer to sta-
bilize. We have since improved the homogeneity of the
excimer laser beam and decreased the tolerance of
some of the laser calibrations to improve results in
these eyes. In addition, while we used a 200-µm
postablation corneal thickness as a minimum for this
study, we observed that some eyes with ablations to
200 µm had slower visual recoveries. We have since
increased this required minimum. For certain patients
with thin corneas and with spherical equivalents
greater than –10 D, we now consider alternative treat-
ments including phakic lens implants.
Discussion
The results of this study demonstrate that LASIK is
an effective and safe alternative for correcting myopia
with or without astigmatism. Based on these data, the
FDA granted the first approval of an excimer laser sys-
tem for performing the LASIK procedure.
With LASIK, the surgeon can avoid the need to
invade the corneal epithelium, Bowman membrane,
and the basal nerve plexus. In our study, LASIK
appeared to benefit patients in a number of ways.
First, patients in this study did not require extensive
pain management. Second, none of the patients in this
study required extension of their short-term steroid
regimen, because the corrections were accurate and
stable. Even though we conservatively determined
that stability (defined as a change in spherical equiva-
lent less than l D from visit to visit) occurred between
the 3- and 6-month follow-up intervals, stability was
achieved in 90.6% of cohort 2 eyes between the 1- and
3-month visits. Although anecdotal, most of our
patients were able to see well enough to pass the dri-
ver’s test on the first or second postoperative day.
Third, although the eyes in this study with higher pre-
operative refractive errors (7 D or greater) were some-
what less likely to achieve UCVA of 20/40 or better and
were slightly more likely to experience BCVA loss and
complications, the procedure is effective and pre-
dictable enough to warrant FDA approval for correct-
ing myopia as high as –15 D and astigmatism to 5 D.
A number of recent studies directly comparing
LASIK with PRK in eyes with moderate to high
myopia found that LASIK produced superior early
results.11-16 In all of these studies, LASIK-treated eyes
had better early UCVA and were associated with less
postoperative pain. In several of the studies, LASIK
was also associated with a more rapid recovery of
BCVA (due to lack of haze) and caused less glare,
halos, and diplopia than did PRK.11-16 One 80-eye trial
of eyes with average myopia (–9.25 D)11 showed that
some grade of haze developed in more than one-fourth
of PRK-treated patients, whereas no LASIK-treated
patient experienced haze. In that study, mean postop-
In sum, these data demonstrate that, for most
patients with myopia or myopic astigmatism who seek
and qualify for surgical correction, LASIK is both safe
and effective.
References
1. Angle J, Wissman DA. Age, reading, and myopia. Am J Optom
Physiol Opt. 1978;55:302–308.
2. Sperduto RD, Seigel D, Roberts L Rowland M. Prevalence of
myopia in the United States. Arch Ophthalmol. 1993;101:405–407.
3. Working Group on Myopia Prevalence and Progression. Myopia:
Prevalence and Progression. Washington, DC: National Academy
Press; 1989.
ANN OPHTHALMOL. 2001;33(4)
321

4. Young FA, Leary GA, Baldwin WR, et al. The transmission of
refractive errors within Eskimo families. Am J Optom Physiol
Opt. 1969;46:676–685.
5. Framingham Offspring Eye Study Group. Familial aggregation
and prevalence of myopia in the Framingham Offspring Eye
Study. Arch Ophthalmol. 1996;114:326–332.
6. Wang Q, Klein BE, Klein, et al. Refractive status in the Beaver
Dam Eye Study. Invest Ophthalmol Vis Sci. 1994;35:4344–4377.
7. Saw S-M, Katz J, Schein OD, Chew S-J, Chan T-K. Epidemiology
of myopia. Epidemiol Rev. 1996;18:175–187.
8. Bear JC, Richler A, Burke G. Nearwork and familial resemblances
in ocular refraction: a population study of Newfoundland. Clin
Genet. 1981;19:462–472.
12. Goldstein BHS, Hersh PS, Brint SF, et al. PRK versus LASIK for
moderate to high myopia: optical side effects [scientific poster].
In: Abstracts of the Association for Research in Vision and Oph-
thalmology Annual Meeting; 1998; Ft Lauderdale, Fla. Abstract
3441.
13. Pineda AA, Jaramillo J, Jaramillo M. PRK vs. LASIK for low and
moderate myopia [scientific poster]. In: Abstracts of the Associa-
tion for Research in Vision and Ophthalmology Annual Meeting;
1998; Ft Lauderdale, Fla. Abstract 3440.
14. Hersh PS, Brint SF, Maloney RK, et al. Photorefractive keratecto-
my versus laser in situ keratomileusis for moderate to high
myopia: a randomized prospective study. Ophthalmology. 1998;
105:1512–1523.
9. Kaufman PL. Accommodation and presbyopia: neuromuscular
and biophysical aspects. In: Hart WM Jr, ed. Adler’s Physiology of
the Eye: Clinical Application. St Louis, Mo: Mosby-Year Book;
1992:390–411.
10. Holladay JT, Craby TV, Koch DD. Calculating the surgically
induced refractive change following ocular surgery. J Cataract
Refract Surg. 1992;18:429–443.
11. Restivo L, Almazar D, Moreno L, et al. A comparative study of
PRK and LASIK techniques for the correction of moderate
myopia [scientific poster]. In: Abstracts of the Association for
Research in Vision and Ophthalmology Annual Meeting; 1998; Ft
Lauderdale, Fla. Abstract 3439.
15. El Danasoury MA, el Maghraby A, Klyce SD, Mehrez K. Compari-
son of photorefractive keratectomy with excimer laser in situ ker-
atomileusis in correcting low myopia (from –2.00 to –5.50
diopters): a randomized study. Ophthalmology. 1999;106:411–420.
16. Steinert RF, Hersh PS, for the Summit Technology PRK-LASIK
Study Group. Spherical and aspherical photorefractive keratecto-
my and laser in-situ keratomileusis for moderate to high myopia:
two prospective, randomized clinical trials. Trans Am Ophthalmol
Soc. 1998;96:197–221.
322
ANN OPHTHALMOL. 2001;33(4)