- Email: [email protected]
Results of bilateral photorefractive keratectomy
Results of Bilateral Photorefractive Keratectomy Mihai Pop, MD, Yves Payette, MSc Objective: To study the refraction and potential risks of bilateral photorefractive keratectomy for myopia. Design: Retrospective, noncomparative case series. Participants: Four hundred eighteen consecutive patients (836 eyes) with myopia from ⫺18 to ⫺0.50 diopters (D) had bilateral simultaneous refractive surgery. Intervention: Refractive surgery was performed with the Nidek EC-5000 slit-scanning excimer laser (Nidek Co., Tokyo, Japan). Main Outcome Measures: Manifest refraction, Snellen best spectacle-corrected visual acuity (BSCVA) and uncorrected visual acuity, haze, and halos were evaluated for 12 months after surgery. Both eyes of each patient were examined at the same follow-up after initial treatment. Results: At 12 months after surgery, 95% of eyes were within ⫾1 D of emmetropia, and 0.4% of patients had residual anisometropia more than 2 D. Thirty-three eyes (3.9%) were retreated, whereas 5 patients (1.2%) had retreatments in both eyes. Odds ratios between unilateral versus bilateral postoperative events were evaluated for the retreatments (1:0.22; P ⫽ 0.001), undercorrections more than 1 D (1:0.17; P ⬍ 0.0001), overcorrections more than 1 D (1:0.23; P ⬍ 0.0001), loss of 1 line of BSCVA (1:0.44; P ⱕ 0.002), and loss of 2 lines of BSCVA (1:0.18; P ⫽ 0.013). At 12 months after surgery, one eye with a loss of 2 lines of BSCVA was treated for haze more than 1; the fellow eye had haze 0.75. No patient had bilateral haze more than 1. At 12 months after surgery, no patient lost 2 lines of BSCVA in both eyes. There was a higher incidence of halos in both eyes rather than in one eye only (odds ratio: 4.17; P ⬍ 0.0001). No postoperative infections occurred. Conclusions: Bilateral events occurred approximately 2.5 to 4 times less often than unilateral events, except for the incidence of halos. The calculated odds ratio did not show an increased risk for the fellow eye while performing bilateral surgery, although not exempted from the risk of infection. It may be difficult to predict a complication in the fellow eye based on the results of the first operated eye. Ophthalmology 2000;107:472– 479 © 2000 by the American Academy of Ophthalmology. Photorefractive keratectomy (PRK) for the treatment of myopia is a well-known procedure that is documented to be safe and effective.1– 6 Bilateral simultaneous surgery is commonly performed, although the potential for bilateral infection is rare and not well documented. Anisometropia can also result from a unilateral or bilateral refractive surgery and can be significant if the refraction between both eyes differs by 1 diopter (D).7 Many refractive surgeons prefer to wait up to 3 months before operating on the fellow eye8 to analyze the outcome of the first eye, during which time patients may experience anisometropia. Bilateral surgery is perceived by some as a way of increasing the cost:benefit ratios for the refractive surgeon.9
Originally received: November 10, 1998. Accepted: November 2, 1999. Manuscript no. 98599. The Michel Pop Clinics, Montreal and Hull, Quebec, Canada. Presented in part at the American Academy of Ophthalmology annual meeting, New Orleans, Louisiana, November 1998. The authors have no proprietary interest in any of the materials used in this study. Reprint requests to Mihai Pop, MD, Michel Pop Clinics, 9001 de l’Acadie Boulevard N., Suite 900, Montreal, PQ, Canada H4N 3H5.
472
© 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
However, benefits for the patients in terms of visual recovery of bilateral vision and risk to the fellow eye have not been studied extensively. Previous studies reported that patients did not experience anisometropia after bilateral surgery and that no visionthreatening complications occurred.10 –13 To our knowledge, no other study has shown the relative risk for the fellow eye after bilateral surgery. The main objective of the present study was not to compare results of sequential and bilateral simultaneous surgery; it is expected that consecutive sequential and bilateral surgeries from the same surgeon using the same excimer laser during the same time frame could generate equal refractive outcomes.11 Zaldivar et al10 reported a statistically insignificant increase of 6.5% of eyes within ⫾0.5 D of emmetropia after bilateral simultaneous surgery, but this difference was associated with the experience gained by the surgeon after performing sequential surgery only for the first part of the study. The main objectives of the present study were to determine the postoperative refractive relation between both eyes after bilateral refractive surgery and to establish the frequency of occurrence and relative odds ratios of different postoperative events for one or both eyes. Few studies have evaluated the risks to both eyes after refractive surgery,12,13 ISSN 0161-6420/00/$–see front matter PII S0161-6420(99)00129-3
Pop and Payette 䡠 Results of Bilateral PRK Table 1. Eyes Before Surgery No. Eyes
BSCVA 20/20 or Better (%)
BSCVA 20/40 or Better (%)
836
705 (84.3)
828 (99.0)
Refraction (D) Sphere ⫾ SD
Sphere Equivalent ⫾ SD
⫺5.18 ⫾ 2.77 (⫺18 to ⫺0.50)
⫺5.61 ⫾ 2.86 (⫺18.5 to ⫺1)
Pupil Size ⴞ SD (mm)
Age ⴞ SD (yrs)
7.0 ⫾ 1.0
35 ⫾ 10
BSCVA ⫽ best spectacle-corrected visual acuity; D ⫽ diopter; SD ⫽ standard deviation.
and some authors have expressed the wish to study such risks.14,15
Patients and Methods Patients The Nidek EC-5000 slit-scanning excimer laser (Nidek Co., Tokyo, Japan) was used to perform PRK on 836 eyes (418 patients) by the same surgeon (MP). The patients were treated for myopia from ⫺0.50 to ⫺18 D with astigmatism from 0 to ⫺5 D. All patients gave their informed consent for the bilateral surgery.
Examination Examinations included subjective refraction, Snellen best spectacle-corrected visual acuity (BSCVA) and uncorrected visual acuity (UCVA), corneal topography, slit-lamp examination, haze and halos measure. Eye dominance was evaluated before surgery. The pupil size was evaluated in scotopic conditions using the Rosenbaum pupil gauge (Graham-Field Surgical Co., New York, NY). Haze was graded from 0 to 4 by increments of 0.25, as proposed by several authors.16 –18 Patients’ reports of halos were evaluated on a subjective scale using the terms none, mild, moderate, and disturbing. Follow-up examinations were performed at 72 hours and 1, 2, 3, 6, and 12 months after surgery. Both eyes of a patient were examined at the same follow-up.
Surgical Procedure Topical anesthesia, proparacaine 0.5% drops (Alcaine; Allergan, Montreal, Canada), was instilled and manual debridement of the epithelium was performed using a blunt no. 67 blade. The laser was set at 40 Hz, removing 0.6 m per scan, with a 6-mm optical zone and a transition zone of 8 mm. A multipass technique was also used: the total amount of correction was separated into multiple smaller treatments of equal values of sphere and cylinder. The passes were set so that the operating time at each pass was 20 to 30 seconds. The sum of these smaller treatments was equal to the actual intended correction. The laser was stopped between each pass for approximately 15 seconds. Both eyes of a patient had consecutive refractive surgery on the same day, beginning with the right eye. The standardized PRK procedure used is explained in detail elsewhere.19 For all patients more than 40 years old, a residual myopia of ⫺0.75 D was intended for the nondominant eye; this procedure was called monovision. For all other patients, emmetropia was the intended correction. The first eye to be operated on was the right eye.
Medication After PRK, a soft contact lens was placed on the eye. Nonsteroidal anti-inflammatory drugs, such as ketorolac tromethamine (Acular;
Allergan, Montreal, Canada), and corticosteroid drops, such as 0.1% fluorometholone (FML; Allergan, Montreal, Canada), were given every 4 hours for the first day and then thrice daily for the next 48 hours. Antibiotic drops, ofloxacin 0.3% (Ocuflox; Allergan, Montreal, Canada), were given every 4 hours for the first 72 hours and thrice daily for the following week. For patients with myopia more than ⫺6 D, corticosteroids were given twice daily for the first postoperative month, four times daily for the second month, thrice daily for the third month, twice daily for the fourth month, and once daily for the fifth month. For patients with myopia less than ⫺6 D, corticosteroids (0.1% fluorometholone) were given thrice daily during the first postoperative week only.
Study Design Four hundred eighteen consecutive patients who underwent refractive surgery for both eyes on the same day were analyzed retrospectively. All surgeries were performed by the same surgeon (MP), and follow-up examinations for both eyes of the patients were performed at the appropriate intervals on the same follow-up date. Retreated eyes were also included in the study. The results of the initial procedure for all eyes were included in the study. Retreatments were analyzed separately at 12 months after surgery. The postoperative residual anisometropia between both eyes was analyzed with the patients who had a postoperative examination for both eyes at the same time. For the evaluation of the residual anisometropia, the patients with an intended undercorrection of ⫺0.75 D in the nondominant eye (monovision) were separated from the patients in whom emmetropia was the intended correction.
Data Analysis Statistical analysis was performed using Foxpro 2.6 (Microsoft, Seattle, WA) for compilation of data. Statistical analysis was performed with Simstat 1.22 (Provalis Research, Montreal, Canada) for the different Student’s t and chi-square tests. Odds ratio were determined using the logrank test, as described by DawsonSaunders and Trapp,20 or using the Fisher exact test in the case of 2 ⫻ 2 contingency tables. A P value of 0.05 or less was considered statistically significant.
Results Patients Forty-nine eyes had initial sphere between ⫺10 and ⫺18 D, 202 eyes had initial sphere between ⫺6 and ⫺9.5 D, and 585 eyes had initial sphere between ⫺0.50 and ⫺6 D. Mean sphere was ⫺5.19 ⫾ 2.77 D with astigmatism up to ⫺5 D (Table 1). Patients’ mean age was 35 ⫾ 10 years. The mean pupil size was 7 ⫾ 1 mm. Seventy-two percent of eyes required no astigmatic correction. Eight eyes did not have BSCVA of 20/40 or better before the initial surgery, but had BSCVA of 20/70 and were considered fit
473
Ophthalmology Volume 107, Number 3, March 2000 Table 2. Postoperative Refractive Results for All Eyes and for Eyes That Were Retreated Months after Surgery 1 2 3 6 12
Patients*
No. Eyes
Sphere ⴞ0.5 D (%)
Sphere 0.5–1 D (%)
Sphere 1–2 D (%)
Sphere >2 D (%)
Initial Retreated Initial Retreated Initial Retreated Initial Retreated Initial Retreated
686 28 431 11 578 19 647 26 646 22
72.2 64.3 71.7 72.7 76.5 94.7 80.0 92.3 83.5 81.9
19.9 25.0 20.4 18.2 16.8 5.3 14.7 7.7 12.0 13.6
6.7 3.6 6.5 9.1 5.7 0 4.8 0 4.2 4.5
1.2 7.1 1.4 0 1.0 0 0.5 0 0.3 0
D ⫽ diopter. *Results for all eyes were taken after the initial treatment. Results for eyes that had a retreatment were taken after the retreatment.
for surgery. One hundred eleven patients (27%) had an intended monovision correction on their nondominant eye. Thirty-three eyes (3.9%) of 28 patients were retreated. No infections occurred after surgery. Three hundred seventeen patients (634 eyes) had a bilateral examination at 12 months after surgery.
Refraction At 12 months after surgery (Table 2), 83% of eyes were within ⫾0.5 D of emmetropia, and 12% of eyes were between 0.5 and 1 D, for a total of 95% of eyes within ⫾1 D. Two eyes (0.3%) were not within ⫾2 D of emmetropia. After retreatment, 95% of eyes were also within ⫾1 D of emmetropia.
Internal Consistency The first eye to have surgery during a bilateral procedure could be considered a unilateral procedure if the fellow eye did not undergo surgery afterward. Because surgery was performed first on the right eye of each patient, statistical tests were performed to eval-
uate differences induced on the fellow eye (left) after the initial procedure. Routine t tests showed that the postoperative mean spheres of the right and left eyes were not statistically different at 1 month (P ⫽ 0.57), 2 months (P ⫽ 0.57), 3 months (P ⫽ 0.79), 6 months (P ⫽ 0.48), and 12 months (P ⫽ 0.89) after surgery. The postoperative likelihood ratios of right and left eyes within ⫾1 D of emmetropia at all follow-up months was also not statistically different (P ⫽ 0.99) using a chi-square test. The postoperative likelihood ratios of eyes with BSCVA of 20/20 or better and with UCVA of 20/20 or better at all follow-up months were not statistically different between the right and left eyes (P ⫽ 0.99 and P ⫽ 0.94, respectively). The odds ratio between right and left eyes was not statistically different for the number of eyes with loss of 2 lines of BSCVA or more (P ⫽ 0.06), for the number of eyes with loss of 1 line of BSCVA (P ⫽ 0.45), and for the number of retreatments (P ⬎ 0.99).
Residual Anisometropia Table 3 shows the residual anisometropia of the manifest refractions found between both eyes. At 12 months after surgery, for patients in whom emmetropia was the intended correction (Table 3), residual anisometropia more than 2 D occurred in 1 patient (0.4%); 7 patients (3.0%) had residual anisometropia of 1 to 2 D; 27 patients (11.5%) had between 0.50 and 1 D of residual anisometropia; and 200 patients (85.1%) had less than 0.50 D of residual anisometropia. At 12 months after surgery, 2 patients (2.4%) with an intended undercorrection in the nondominant eye (monovision) had more than 2 D of residual anisometropia (Table 3). The postoperative proportion of residual anisometropia was always lower than the preoperative existing proportion of anisometropia, although this was expected because the patients underwent bilateral surgery. No anisometropia more than 1 D was found at 12 months after surgery for patients who were retreated. Figure 1 shows the difference between the manifest refraction of the left and the right eyes per level of initial myopia. No significant correlation was found (r ⫽ 0.06; P ⫽ 0.26). The three eyes with residual anisometropia more than 2 D (one eye in the emmetropia group and two eyes in the monovision group) all had initial myopia more than 10 D.
Table 3. Postoperative Residual Anisometropia of Manifest Refraction Residual Anisometropia for Patients with Intended Correction Equal to Emmetropia Patients Before Surgery All patients 1 2 3 6 12 Retreated patients 12 months after retreatment†
Residual Anisometropia for Patients with Monovision
No. Patients*
n
⫾0.5 D (%)
0.5–1 D (%)
1–2 D (%)
⬎2 D (%)
n
⫾0.5 D (%)
0.5–1 D (%)
1–2 D (%)
⬎2 D (%)
418
307
63.8
20.9
10.4
4.9
111
58.6
18.9
15.3
7.2
342 215 285 319 317
251 152 211 236 235
75.3 82.2 84.4 85.6 85.1
18.7 10.5 10.9 10.6 11.5
5.2 6.6 3.8 3.4 3.0
0.8 0.7 0.9 0.4 0.4
91 63 74 83 82
57.1 54.0 51.4 55.4 56.2
27.5 34.9 31.1 28.9 28.0
14.3 9.5 16.2 14.5 13.4
1.1 1.6 1.3 1.2 2.4
4
3
66.7
33.3
0
0
0
0
1
100
0
D ⫽ diopters. *Patients who had an examination for both eyes at the same time after initial treatment. †Results for all patients were taken after the initial treatment. Results for patients who had a retreatment were taken 12 months after retreatment, because the postoperative follow-up dates were different between both eyes of the patients.
474
Pop and Payette 䡠 Results of Bilateral PRK
Figure 1. Twelve-month postoperative difference of the manifest refraction between the left and the right eye per initial degree of myopia of the left eye of a patient.
Regression and Astigmatism The mean individual regression between two time intervals was calculated. The 1- to 12-month regression rate was 0.16 ⫾ 0.54 D after the initial treatment, and was 0.58 ⫾ 1.04 D for retreated eyes. At 12 months after surgery, a maximum of 3.9% of eyes regressed more than 1 D after the initial treatment. At 12 months after surgery, 86% of eyes had final astigmatism within 0.5 D, and 98% of eyes had final astigmatism within 1 D. The average remaining cylinder at 12 months after surgery was ⫺0.29 ⫾ 0.34 D; the maximum remaining cylinder was ⫺2.25 D.
Visual Acuity At 12 months after surgery, 98% of eyes had UCVA of 20/40 or better and 87% of eyes had UCVA of 20/20 or better (Table 4). Sixty-eight percent of patients experienced no difference in UCVA between both eyes; 23% of patients had one UCVA line difference between both eyes; 5% of patient had a difference of two UCVA lines between both eyes; and 4% of eyes had a difference of more than two UCVA lines. The differences in BSCVA and UCVA lines
between both eyes decreased gradually over time (Table 4). Eight eyes (0.9%) of 7 patients had BSCVA more than 20/40, and 13 eyes (1.6%) of 11 patients had BSCVA equal to 20/40 before surgery. Although the percent of eyes with BSCVA more than 20/40 was 0.9% and 1.4% at 1 and 2 months after surgery, respectively, the percent of these eyes dropped to 0.5% of eyes at 12 months after surgery (Table 4). All the patients who had at least one eye with BSCVA more than 20/40 at 6 and 12 months after surgery also had initial BSCVA of 20/40 or more. A maximum loss of one Snellen line was seen in eyes with BSCVA of 20/40 or more at all times after surgery. Table 5 shows the number of patients who lost lines of BSCVA during the first postoperative year. All eyes were included in Table 5, even those that did not have a bilateral examination (fellow eye was retreated). The maximum number of eyes that lost lines of BSCVA occurred at the first and second postoperative month. No patient lost two lines of BSCVA in both eyes after the second postoperative month and no patient lost two lines of BSCVA in both eyes at 12 months after surgery. No patient lost any lines of BSCVA after retreatment at 2, 3, 6, or 12 months after surgery. All eyes, except one, that lost two or more Snellen lines of BSCVA at 1, 2, or 3 months after surgery were carefully monitored and regained at least one line at 6 or 12 months after surgery or after retreatment, for a final maximum loss of one line of BSCVA. The eye that lost two lines of BSCVA at 3, 6, and 12 months after surgery had a final outcome of 20/40 BSCVA with haze 1. Initially, this eye was treated for myopia of ⫺12.50 D. The fellow eye of this patient also experienced haze of 0.75 at 12 months after surgery. Both eyes were treated with corticosteroids for late haze. Fewer patients lost BSCVA lines bilaterally than those who lost lines of BSCVA in one eye only at all postoperative examinations. The odds ratio of patients who lost one line of BSCVA in one eye versus in both eyes was 0.44 (P ⫽ 0.002), and the odds ratio of patients who lost two lines of BSCVA in one eye versus in both eyes was 0.18 (P ⫽ 0.013). The odds ratio of patients who lost two lines of BSCVA in one eye versus the patients who lost two lines in one eye plus one line in the fellow eye was 0.27 (P ⫽ 0.032). Statistical tests were performed with low statistical power (20%– 40%).
Table 4. Postoperative Visual Acuities
Before Surgery All patients 1 2 3 6 12 Retreated patients 12 months after retreatment§
BSCVA Differences (Snellen Lines)‡
UCVA Differences (Snellen Lines)‡
No. Eyes
UCVA 20/20 or Better (%)*
UCVA 20/40 or Better (%)*
BSCVA >20/40 (%)*
No. Patients†
0 (%)
1 (%)
2 (%)
⬎2 (%)
0 (%)
1 (%)
2 (%)
⬎2 (%)
836
NA
NA
0.9
418
86.8
9.1
2.6
1.4
NA
NA
NA
NA
686 431 578 647 646
67.5 67.5 72.3 81.5 86.7
95.6 95.6 96.5 97.4 98.0
0.9 1.4 0.8 0.6 0.5
342 215 285 319 317
80.8 84.6 79.2 84.9 83.9
16.2 11.2 18.3 14.2 15.1
2.4 3.7 2.1 0.6 0.6
0.6 0.5 0.4 0.3 0.3
53.3 59.9 56.4 63.4 68.3
28.1 26.4 29.1 24.5 22.8
12.4 8.5 8.2 6.0 4.8
6.2 5.2 6.4 7.2 4.2
22
86.4
4
75.0
25.0
0
0
25.0
75.0
0
100
0
0
BSCVA ⫽ Snellen best spectacle-corrected visual acuity; NA ⫽ not available; UCVA ⫽ Snellen uncorrected visual acuity. *Percent of all eyes. †
Number of patients who had an examination for both eyes at the same time. Percent of patients who had an examination for both eyes at the same time after initial treatment. § Results for patients who had a retreatment were taken 12 months after retreatment, because the postoperative follow-up dates were different between both eyes of the patients. ‡
475
Ophthalmology Volume 107, Number 3, March 2000 Table 5. Loss of BSCVA Snellen Lines after Initial Treatment Months after Surgery
No. Patients
No. Eyes
1 Line (%)
ⱖ2 Lines (%)
One Eye
Both Eyes
One Eye
Both Eyes
1 2 3 6 12
344 216 293 328 329
686 431 578 647 646
38 (5.5) 13 (3.0) 22 (3.8) 9 (1.4) 3 (0.5)
8 (1.2) 6 (1.4) 2 (0.3) 1 (0.2) 1 (0.2)*
20 7 12 3 1
8 3 5 3 0
4 4 2 1 0
1 1 0 0 0
Total Cases
Odds ratio (probability) 95% Confidence interval
>2 Lines Lost
1 Line Lost
1:0.21 (⬍0.0001) 0.11–0.40
1:0.44 (0.002) 0.22–0.92
1:0.18 (0.013) 0.02–1.17
1 and 2 Lines in Both Eyes 2 0 0 0 1† 1:0.27 (0.032)* 0.05–1.42
BSCVA ⫽ best spectacle-corrected visual acuity. *Statistical test performed with frequency of eyes that lost two lines in one eye. †All eyes that lost two lines of BSCVA at 1, 2, or 3 months after surgery regained at least one line of BSCVA at 6 and 12 months after surgery except for one eye. The right eye of this patient had initial myopia of ⫺12.50 diopters and lost two lines of BSCVA at 3, 6, and 12 months after surgery, with a final BSCVA of 20/40 and was treated for late haze of 1.
Haze
Odds Ratio
Mean haze values were below 1 at all postoperative follow-ups (Table
Table 8 lists the maximum occurrence of refractive events that may occur during the first year after surgery. All 418 patients were evaluated after the initial procedure. These results should be carefully analyzed and compared with Tables 4, 5, 6, and 7 for correct interpretation of the final results. An odds ratio equal to 1 would mean that odds are even; the P value (Table 8) is the probability for the odds of being even. The odds ratios of Table 8 are between 0.18 to 0.44, except for the occurrence of halos, meaning that the proportion of bilateral events were 2.5 to 4 times lower than unilateral events. All odds ratios had significant probabilities of being different than 1 (equal odds), meaning that there was not a one-to-one relation between unilateral and bilateral events. Odds ratios could not be calculated for haze of 1 or more because only one eye had haze equal to 1 at 6 and 12 months after surgery. Table 9 is a recapitulation of events and risks to the fellow eye during the first postoperative year. The occurrence of refractive events for all eyes are included; risks to the fellow eye are calculated from all bilateral examinations taken from the previous tables. Table 9 is intended to portray a global picture of the risks generated by bilateral surgery. The risks to the fellow eye reported in Table 9 are the odds ratios calculated in Table 8. The product of
6), on the 0 to 4 scale used for the study.16 –18 All eyes were included in Table 6, even those that did not undergo a bilateral examination (fellow eye was retreated). All eyes had a difference of less than 0.5 of haze between both eyes at 6 and 12 months after surgery. No haze more than 1 occurred after all retreatments. One eye had haze equal to 1 in one eye and haze of 0.75 in the fellow eye; as reported previously, this eye lost two lines of BSCVA and was treated for late haze. The number of patients who had 0.5 to 0.75 haze in both eyes was higher than the number of patients who had low haze in only one eye (odds ratio, 1.724; P ⫽ 0.0008).
Halos Moderate and disturbing halos were considered problematic postoperative issues, although no patient reported disturbing halos at any time (Table 7). Mild halos were considered occasionally problematic. At 12 months after surgery, 13 patients (23 eyes) had moderate halos. The same patients complained about halos in both eyes, so that the percent of halos per eyes (Table 7) approximated the percent of halos per patients. The odds ratio of moderate halos in one eye versus in both eyes was 4.17 (P ⬍ 0.0001). All eyes were included in Table 7, even those eyes that did not have a bilateral examination (fellow eye was retreated).
Table 6. Postoperative Haze after Initial Treatment Total Cases Months after Surgery
No. Patients
No. Eyes
1 2 3 6 12
344 216 293 328 329
686 431 578 647 646
Mean Haze*
Haze ⫽ 1 (%)
One Eye
Both Eyes
One Eye
Both Eyes
Haze 0.5–0.75 and Haze of 1 in Both Eyes
0.22 ⫾ 0.15 0.22 ⫾ 0.18 0.16 ⫾ 0.17 0.09 ⫾ 0.16 0.04 ⫾ 0.11
70 (10.2) 81 (18.8) 52 (9.0) 41 (6.3) 14 (2.2)
0 (0) 0 (0) 0 (0) 1 (0.2) 1 (0.2)
16 7 10 17 8
27 37 21 12 3
0 0 0 0 0
0 0 0 0 0
0 0 0 1 1
Odds ratio (probability) 95% Confidence interval NA ⫽ not available. *Haze was assessed using a 0 to 4 scale.
476
Haze ⴝ 1
Haze 0.5–0.75
Haze 0.5–0.75 (%)
1:0.008 (⬍0.0001) 0.001–0.032
1:1.724 (0.0008) 1:10–2.70
NA NA
NA NA
Pop and Payette 䡠 Results of Bilateral PRK Table 7. Halos after Initial Treatment Total Cases Months after Surgery
No. Patients
No. Eyes
1 2 3 6 12
344 216 293 328 329
686 431 578 647 646
Mild Halos
Moderate Halos
Mild Halos (%)
Moderate Halos (%)
One Eye
Both Eyes
One Eye
Both Eyes
Mild and Moderate Halos in Both Eyes
155 (22.5) 79 (18.3) 133 (23.0) 144 (22.2) 113 (17.4)
53 (7.8) 23 (5.3) 39 (6.7) 38 (5.9) 23 (3.6)
9 10 5 10 6
72 34 63 66 53
7 2 5 2 2
22 10 16 17 10
2 1 2 2 1
Odds ratio (probability) 95% Confidence interval
1:0.28 (⬍0.0001) 0.23–0.35
the maximum occurrence (% of eyes) by the risk to the fellow eye should indicate the maximum occurrence of bilateral events. The maximum loss of two lines of BSCVA or more occurred in 1.4% of eyes during the first postoperative year with odds ratio to the fellow eye of 0.18 (Table 5). Moderate halos had a higher bilateral than unilateral incidence, occurring in 7.8% of eyes. A total of 3.9% of patients had retreatments with an odds ratio to the fellow eye of 0.22 (Table 8); overcorrections more than 1 D occurred in 6.9% of eyes, and undercorrections occurred in 4.7% of eyes, with odds ratios below 0.25 for both events (Table 8).
Discussion The results of this bilateral study were not expected to be better than the results of sequential surgery. However, the
1:7.20 (⬍0.0001) 4.85–10.69
1:4.17 (⬍0.0001) 2.17–8.03
1:0.44 (0.05) 0.14–1.38
present study showed the amount of postoperative residual anisometropia and odds ratios of different postoperative events after bilateral surgery. Similar to recent studies, the refractive outcomes of this study, as analyzed per eyes, are approximately the same or higher at the different follow-up examinations.3– 6 However, the results of this study were dependent on the initial amount of myopia per eye. It was estimated, because these patients were consecutively selected, that the distribution of patient’s myopia of this study may represent the average population of a surgeon with a private practice in refractive surgery. The literature has numerous references to evaluate the results of PRK for low, moderate, and high myopia.3,4,5,17,21,22 A study comparing sequential and bilateral procedures
Table 8. Statistics for All 418 Patients during the First Postoperative Year Events (Maximum Occurrence)
No. Patients
Odds Ratio (Probability)
95% Confidence Interval
One eye retreated Both eyes retreated
23 5
0.22 (0.001)
(0.08–0.58)
One eye with undercorrection more than 1 D Both eyes with undercorrection more than 1 D
29 5
0.17 (⬍0.0001)
(0.07–0.45)
One eye with overcorrection more than 1 D Both eyes with overcorrection more than 1 D
40 9
0.23 (⬍0.0001)
(0.11–0.47)
One eye with loss of BSCVA of one Snellen line Both eyes with loss of BSCVA of one Snellen line
20 8
0.44 (0.002)
(0.22–0.92)
0.27 (⬍0.032)† 0.18 (0.013)†
(0.05–1.42)† (0.02–1.17)†
One eye with loss of BSCVA of one Snellen line and one eye with loss of BSCVA of two lines or more in the fellow eye* One eye with loss of BSCVA of two Snellen lines or more Both eyes with loss of BSCVA of two Snellen lines or more*
2 4 1
One eye with haze more than 1 One eye with haze more than 1 and fellow eye with haze 0.5/0.75 Both eyes with haze more than 1
0 1 0
One eye with moderate halos Both eyes with moderate halos
7 22
NA 4.17 (⬍0.0001)
NA (2.17–8.03)
BSCVA ⫽ best spectacle-corrected visual acuity; D ⫽ diopters; NA ⫽ could not be statistically evaluated. *All eyes that lost two lines of BSCVA at 1, 2, or 3 months after surgery regained at least one line of BSCVA at 6 and 12 months except for one eye. This eye lost two lines of BSCVA at 3, 6, and 12 months after surgery and ended with BSCVA of 20/40 with haze of 1. †Statistical test was performed with very low power.
477
Ophthalmology Volume 107, Number 3, March 2000 Table 9. Recapitulation on the Occurrence of Events for All Eyes and Risks to the Fellow Eye During the First Postoperative Year (Consult Tables 5– 8)
Events
Maximum Occurrence (% of All Eyes)
Risk to Fellow Eye (%)
Undercorrection ⬍1 D Overcorrection ⬎1 D Loss of ⱖ2 lines of BSCVA* Haze ⱖ1 Moderate halos Retreatment
4.7 6.9 1.4 0.2 7.8 3.9
⫻ 17 ⫻ 23 ⫻ 18 NA ⫻ 417 ⫻ 22
BSCVA ⫽ best spectacle-corrected visual acuity; NA ⫽ could not be statistically evaluated. *Maximum occurred 2 months after surgery. No patient lost two lines of BSCVA in both eyes after second postoperative month.
would not specifically determine bilateral patterns for some patients, such as bilateral loss of BSCVA, bilateral haze, and bilateral halos. However, in a bilateral study, the first eye to undergo surgery could be evaluated as an unilateral procedure. In this study, the first operated eye and the fellow eye were compared as part of internal consistency and found not to be statistically significantly different. Bilateral PRK resulted in predictable refractive outcomes for both eyes in approximately 95% of cases. Using the slit-scanning laser, the percent of patients who lost two lines of BSCVA in both eyes was below 1.4% during the first postoperative year. The results showed that the risk of losing lines of BSCVA in both eyes was lower than the risk for one eye. As shown in Tables 8 and 9, bilateral events occurred 2.5 to 4 times less often than unilateral events, except for the occurrence of halos and haze of 0.5 to 0.75, on the 0 to 4 scale. With a 0.2% occurrence of haze more than 1, the present study showed that haze was not problematic. One patient had haze of 0.75 in one eye and haze of 1 in the fellow eye, as previously discussed, and was treated accordingly. No patient had haze more than 1 for both eyes during the first postoperative year. Although a greater number of patients had bilateral haze of 0.5 to 0.75, this fact was of no consequence at the end of the first postoperative year and was expected because PRK commonly creates nonsignificant haze. The slit-scanning laser also contributed to low haze after surgery.22 Halos were four times as common bilaterally compared with unilaterally during the first postoperative year. The odds ratio for halos showed a distinctive pattern toward bilateral halos. Therefore, this could be interpreted as some patients being at higher risk to perceive halos in both eyes. Because reports of halos were subjective, it is likely that the perception of halos in an eye may be influenced by global bilateral vision or even by the global satisfaction of the vision. Further studies may be required to evaluate halos, because the scale used for this study should be correlated to objective measures. Satisfaction of the global vision should also be correlated with the presence of halos. Pupil size, depth of the ablation, optical zones types, and sizes are also possible variables in this equation.
478
In this study, the amount of residual anisometropia after a bilateral procedure led to interesting benefits for the patient, as for visual recovery, with low risks to the fellow eye. Patients experienced only one period of postoperative healing, pain, and visual recovery. Although other studies with greater numbers of patients and longer follow-ups may be needed, the present study showed that the possibility of an adverse event in the fellow eye is approximately 1 in 4 if a complication occurs in the first eye. No pattern could be established to predict postoperative events, except for the incidence of postoperative halos for the fellow eye when the first eye is affected. Considering the probabilities of obtaining bilateral events, the bilateral probabilities can be easily evaluated by multiplying the maximum event incidence (% of all eyes) by the risk to fellow eye (Table 9). The probability of obtaining two consecutive events is the square product of the probability of one event. For example, obtaining two consecutive “ones” in two rows with a dice would be 1/6 ⫻ 1/6 ⫽ 1/36; P ⫽ 1/36 being the bilateral probability of obtaining “ones.” Using the present study results (Table 9), the bilateral probabilities were not the square product of unilateral probabilities, meaning that bilateral risks were not equal to the same unilateral risks applied to both eyes. Bilateral probabilities of event incidence were generally higher than the square product of unilateral probabilities, but were different, and statistically lower, than unilateral probabilities. The fact that the risk to the fellow eye was higher after a problem had occurred in the first eye can lead to the hypothesis that bilateral events are not independent. These rare events may be related to unidentified healing properties of the eye, or unknown predisposing factors that may take months to observe and that influence both eyes. However, the comparison of the unilateral and bilateral probabilities confirms that event incidence cannot be predicted using the results of a patient’s first operated eye. To reveal the true difference between unilateral and bilateral probabilities, a study with greater occurrence of events may be needed to increase statistical power. Bilateral surgery is still controversial. The present study does not state that bilateral simultaneous surgery is safer than sequential surgery, nor can the opposite be also stated. Bilateral surgery is a procedure with refractive benefits, and low risks, that should not be rejected by the back of the hand because of general usage and custom, and it has become a valid option for patients. None of the complications faced in the present study could have led a surgeon not to operate in the fellow eye. One of the risks not assessed in this study was the infection rate. Although no infection occurred during this study, risks of infection should be thoroughly explained to every patient before bilateral surgery. The present study does not offer a conclusion on the risk of bilateral infection. However, one may consider these risks to be extremely rare. By cumulating the study results of 9375 eyes, Maloney (see Discussion)23 showed that the probability of a catastrophic complication, including infections, after PRK or LASIK could be statistically estimated to be less than 1 in 3129. Using a ratio of 1:3 for bilateral events, he estimated the bilateral probability of catastrophic complication to be less
Pop and Payette 䡠 Results of Bilateral PRK than 1 in 10,000, and concluded that as long as the patient did not have an unusual risk factor, bilateral simultaneous photorefractive surgery was sufficiently safe, and that this type of surgery may be compared with other types of bilateral surgeries.
References 1. Goggin MJ, Kenna PF, Lavery FL. Photoastigmatic refractive keratectomy for compound myopic astigmatism with a Nidek laser. J Refract Surg 1997;13:162– 6. 2. Stephenson CG, Gartry DS, O’Brart DPS, et al. Photorefractive keratectomy. A 6-year follow-up study. Ophthalmology 1998;105:273– 81. 3. Tuunanen TH, Tervo TT. Results of photorefractive keratectomy for low, moderate, and high myopia. J Refract Surg 1998;14:437– 46. 4. Kim JH, Kim MS, Hahn TW, et al. Five year results of photorefractive keratectomy for myopia. J Cataract Refract Surg 1997;23:731–5. 5. Zadok D, Haviv D, Vishnevskia-Dai V, et al. Excimer laser photoastigmatic refractive keratectomy. Eighteen-month follow-up. Ophthalmology 1998;105:620 –3. 6. Carr JD, Stulting RD, Sano Y, et al. Prospective comparison of single-zone and multizone laser in situ keratomileusis for the correction of low myopia. Ophthalmology 1998;105:1504 – 11. 7. Grosvenor T. Primary Care Optometry: Anomalies of Refraction and Binocular Vision, 3rd ed. Boston: Butterworth-Heinemann, 1996. 8. Kohl M. To wait or not to wait: opinion varies on simultaneous bilateral surgery. J Refract Surg 1997;13:611–2. 9. Chynn EM. Merits of bilateral refractive surgery are debated. Ophthalmology Times, April 1st, 1999: p. 41. 10. Zaldivar R, Oscherow S, Ricur G, Piezzi V. Bilateral simultaneous laser in situ keratomileusis. J Refract Surg 1999;15(2 Suppl):S202– 8.
11. Gimbel HV, van Westenbrugge JA, Anderson Penno EE, et al. Simultaneous bilateral laser in situ keratomileusis: safety and efficacy. Ophthalmology 1999;106:1461– 68. 12. El-Maghraby A, Salah T, Polit F, et al. Efficacy and safety of excimer laser photorefractive keratectomy and radial keratotomy for bilateral myopia. J Cataract Refract Surg 1996;22: 51– 8. 13. Loewenstein A, Lipshitz I, Ben-Sirah A, et al. Symmetry of outcome after photorefractive keratectomy for myopia. J Refract Surg 1995;11(Suppl):S268 –9. 14. Gimbel HV, Van Westenbrugge JA, Johnson WH, et al. Visual, refractive, and patient satisfaction results following bilateral photorefractive keratectomy for myopia. Refract Corneal Surg 1993;9(Suppl):S5–S10. 15. Hashim A. Eye dominance in excimer laser photorefractive keratectomy. J Refract Surg 1997;13(Suppl):S461–2. 16. Pop M, Aras M. Multizone/multipass photorefractive keratectomy: six month results. J Cataract Refract Surg 1995;21:633– 43. 17. Pop M, Aras M. Photorefractive keratectomy retreatments for regression. One-year follow-up. Ophthalmology 1996;103:1979 – 84. 18. Fantes FE, Hanna KD, Waring GO III, et al. Wound healing after excimer laser keratomileusis (photorefractive keratectomy) in monkeys. Arch Ophthalmol 1990;108:665–75. 19. Pop M. Prompt re-treatment after photorefractive keratectomy. J Cataract Refract Surg 1998;24:320 – 6. 20. Dawson-Saunders B, Trapp RG. Basic and Clinical Biostatistics, 2nd ed. Norwalk, CT: Appleton & Lange, 1994. 21. Piovella M, Camesaca FI, Fattori C. Excimer laser photorefractive keratectomy for high myopia. Four-year experience with a multiple zone technique. Ophthalmology 1997;104: 1554 – 65. 22. Pop M, Payette Y. Multipass versus single pass photorefractive keratectomy for high myopia using a scanning laser. J Refract Surg 1999;15:444 –50. 23. Maloney RK. Discussion. Ophthalmology 1999;106:1467– 68.
479
Originally received: November 10, 1998. Accepted: November 2, 1999. Manuscript no. 98599. The Michel Pop Clinics, Montreal and Hull, Quebec, Canada. Presented in part at the American Academy of Ophthalmology annual meeting, New Orleans, Louisiana, November 1998. The authors have no proprietary interest in any of the materials used in this study. Reprint requests to Mihai Pop, MD, Michel Pop Clinics, 9001 de l’Acadie Boulevard N., Suite 900, Montreal, PQ, Canada H4N 3H5.
472
© 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
However, benefits for the patients in terms of visual recovery of bilateral vision and risk to the fellow eye have not been studied extensively. Previous studies reported that patients did not experience anisometropia after bilateral surgery and that no visionthreatening complications occurred.10 –13 To our knowledge, no other study has shown the relative risk for the fellow eye after bilateral surgery. The main objective of the present study was not to compare results of sequential and bilateral simultaneous surgery; it is expected that consecutive sequential and bilateral surgeries from the same surgeon using the same excimer laser during the same time frame could generate equal refractive outcomes.11 Zaldivar et al10 reported a statistically insignificant increase of 6.5% of eyes within ⫾0.5 D of emmetropia after bilateral simultaneous surgery, but this difference was associated with the experience gained by the surgeon after performing sequential surgery only for the first part of the study. The main objectives of the present study were to determine the postoperative refractive relation between both eyes after bilateral refractive surgery and to establish the frequency of occurrence and relative odds ratios of different postoperative events for one or both eyes. Few studies have evaluated the risks to both eyes after refractive surgery,12,13 ISSN 0161-6420/00/$–see front matter PII S0161-6420(99)00129-3
Pop and Payette 䡠 Results of Bilateral PRK Table 1. Eyes Before Surgery No. Eyes
BSCVA 20/20 or Better (%)
BSCVA 20/40 or Better (%)
836
705 (84.3)
828 (99.0)
Refraction (D) Sphere ⫾ SD
Sphere Equivalent ⫾ SD
⫺5.18 ⫾ 2.77 (⫺18 to ⫺0.50)
⫺5.61 ⫾ 2.86 (⫺18.5 to ⫺1)
Pupil Size ⴞ SD (mm)
Age ⴞ SD (yrs)
7.0 ⫾ 1.0
35 ⫾ 10
BSCVA ⫽ best spectacle-corrected visual acuity; D ⫽ diopter; SD ⫽ standard deviation.
and some authors have expressed the wish to study such risks.14,15
Patients and Methods Patients The Nidek EC-5000 slit-scanning excimer laser (Nidek Co., Tokyo, Japan) was used to perform PRK on 836 eyes (418 patients) by the same surgeon (MP). The patients were treated for myopia from ⫺0.50 to ⫺18 D with astigmatism from 0 to ⫺5 D. All patients gave their informed consent for the bilateral surgery.
Examination Examinations included subjective refraction, Snellen best spectacle-corrected visual acuity (BSCVA) and uncorrected visual acuity (UCVA), corneal topography, slit-lamp examination, haze and halos measure. Eye dominance was evaluated before surgery. The pupil size was evaluated in scotopic conditions using the Rosenbaum pupil gauge (Graham-Field Surgical Co., New York, NY). Haze was graded from 0 to 4 by increments of 0.25, as proposed by several authors.16 –18 Patients’ reports of halos were evaluated on a subjective scale using the terms none, mild, moderate, and disturbing. Follow-up examinations were performed at 72 hours and 1, 2, 3, 6, and 12 months after surgery. Both eyes of a patient were examined at the same follow-up.
Surgical Procedure Topical anesthesia, proparacaine 0.5% drops (Alcaine; Allergan, Montreal, Canada), was instilled and manual debridement of the epithelium was performed using a blunt no. 67 blade. The laser was set at 40 Hz, removing 0.6 m per scan, with a 6-mm optical zone and a transition zone of 8 mm. A multipass technique was also used: the total amount of correction was separated into multiple smaller treatments of equal values of sphere and cylinder. The passes were set so that the operating time at each pass was 20 to 30 seconds. The sum of these smaller treatments was equal to the actual intended correction. The laser was stopped between each pass for approximately 15 seconds. Both eyes of a patient had consecutive refractive surgery on the same day, beginning with the right eye. The standardized PRK procedure used is explained in detail elsewhere.19 For all patients more than 40 years old, a residual myopia of ⫺0.75 D was intended for the nondominant eye; this procedure was called monovision. For all other patients, emmetropia was the intended correction. The first eye to be operated on was the right eye.
Medication After PRK, a soft contact lens was placed on the eye. Nonsteroidal anti-inflammatory drugs, such as ketorolac tromethamine (Acular;
Allergan, Montreal, Canada), and corticosteroid drops, such as 0.1% fluorometholone (FML; Allergan, Montreal, Canada), were given every 4 hours for the first day and then thrice daily for the next 48 hours. Antibiotic drops, ofloxacin 0.3% (Ocuflox; Allergan, Montreal, Canada), were given every 4 hours for the first 72 hours and thrice daily for the following week. For patients with myopia more than ⫺6 D, corticosteroids were given twice daily for the first postoperative month, four times daily for the second month, thrice daily for the third month, twice daily for the fourth month, and once daily for the fifth month. For patients with myopia less than ⫺6 D, corticosteroids (0.1% fluorometholone) were given thrice daily during the first postoperative week only.
Study Design Four hundred eighteen consecutive patients who underwent refractive surgery for both eyes on the same day were analyzed retrospectively. All surgeries were performed by the same surgeon (MP), and follow-up examinations for both eyes of the patients were performed at the appropriate intervals on the same follow-up date. Retreated eyes were also included in the study. The results of the initial procedure for all eyes were included in the study. Retreatments were analyzed separately at 12 months after surgery. The postoperative residual anisometropia between both eyes was analyzed with the patients who had a postoperative examination for both eyes at the same time. For the evaluation of the residual anisometropia, the patients with an intended undercorrection of ⫺0.75 D in the nondominant eye (monovision) were separated from the patients in whom emmetropia was the intended correction.
Data Analysis Statistical analysis was performed using Foxpro 2.6 (Microsoft, Seattle, WA) for compilation of data. Statistical analysis was performed with Simstat 1.22 (Provalis Research, Montreal, Canada) for the different Student’s t and chi-square tests. Odds ratio were determined using the logrank test, as described by DawsonSaunders and Trapp,20 or using the Fisher exact test in the case of 2 ⫻ 2 contingency tables. A P value of 0.05 or less was considered statistically significant.
Results Patients Forty-nine eyes had initial sphere between ⫺10 and ⫺18 D, 202 eyes had initial sphere between ⫺6 and ⫺9.5 D, and 585 eyes had initial sphere between ⫺0.50 and ⫺6 D. Mean sphere was ⫺5.19 ⫾ 2.77 D with astigmatism up to ⫺5 D (Table 1). Patients’ mean age was 35 ⫾ 10 years. The mean pupil size was 7 ⫾ 1 mm. Seventy-two percent of eyes required no astigmatic correction. Eight eyes did not have BSCVA of 20/40 or better before the initial surgery, but had BSCVA of 20/70 and were considered fit
473
Ophthalmology Volume 107, Number 3, March 2000 Table 2. Postoperative Refractive Results for All Eyes and for Eyes That Were Retreated Months after Surgery 1 2 3 6 12
Patients*
No. Eyes
Sphere ⴞ0.5 D (%)
Sphere 0.5–1 D (%)
Sphere 1–2 D (%)
Sphere >2 D (%)
Initial Retreated Initial Retreated Initial Retreated Initial Retreated Initial Retreated
686 28 431 11 578 19 647 26 646 22
72.2 64.3 71.7 72.7 76.5 94.7 80.0 92.3 83.5 81.9
19.9 25.0 20.4 18.2 16.8 5.3 14.7 7.7 12.0 13.6
6.7 3.6 6.5 9.1 5.7 0 4.8 0 4.2 4.5
1.2 7.1 1.4 0 1.0 0 0.5 0 0.3 0
D ⫽ diopter. *Results for all eyes were taken after the initial treatment. Results for eyes that had a retreatment were taken after the retreatment.
for surgery. One hundred eleven patients (27%) had an intended monovision correction on their nondominant eye. Thirty-three eyes (3.9%) of 28 patients were retreated. No infections occurred after surgery. Three hundred seventeen patients (634 eyes) had a bilateral examination at 12 months after surgery.
Refraction At 12 months after surgery (Table 2), 83% of eyes were within ⫾0.5 D of emmetropia, and 12% of eyes were between 0.5 and 1 D, for a total of 95% of eyes within ⫾1 D. Two eyes (0.3%) were not within ⫾2 D of emmetropia. After retreatment, 95% of eyes were also within ⫾1 D of emmetropia.
Internal Consistency The first eye to have surgery during a bilateral procedure could be considered a unilateral procedure if the fellow eye did not undergo surgery afterward. Because surgery was performed first on the right eye of each patient, statistical tests were performed to eval-
uate differences induced on the fellow eye (left) after the initial procedure. Routine t tests showed that the postoperative mean spheres of the right and left eyes were not statistically different at 1 month (P ⫽ 0.57), 2 months (P ⫽ 0.57), 3 months (P ⫽ 0.79), 6 months (P ⫽ 0.48), and 12 months (P ⫽ 0.89) after surgery. The postoperative likelihood ratios of right and left eyes within ⫾1 D of emmetropia at all follow-up months was also not statistically different (P ⫽ 0.99) using a chi-square test. The postoperative likelihood ratios of eyes with BSCVA of 20/20 or better and with UCVA of 20/20 or better at all follow-up months were not statistically different between the right and left eyes (P ⫽ 0.99 and P ⫽ 0.94, respectively). The odds ratio between right and left eyes was not statistically different for the number of eyes with loss of 2 lines of BSCVA or more (P ⫽ 0.06), for the number of eyes with loss of 1 line of BSCVA (P ⫽ 0.45), and for the number of retreatments (P ⬎ 0.99).
Residual Anisometropia Table 3 shows the residual anisometropia of the manifest refractions found between both eyes. At 12 months after surgery, for patients in whom emmetropia was the intended correction (Table 3), residual anisometropia more than 2 D occurred in 1 patient (0.4%); 7 patients (3.0%) had residual anisometropia of 1 to 2 D; 27 patients (11.5%) had between 0.50 and 1 D of residual anisometropia; and 200 patients (85.1%) had less than 0.50 D of residual anisometropia. At 12 months after surgery, 2 patients (2.4%) with an intended undercorrection in the nondominant eye (monovision) had more than 2 D of residual anisometropia (Table 3). The postoperative proportion of residual anisometropia was always lower than the preoperative existing proportion of anisometropia, although this was expected because the patients underwent bilateral surgery. No anisometropia more than 1 D was found at 12 months after surgery for patients who were retreated. Figure 1 shows the difference between the manifest refraction of the left and the right eyes per level of initial myopia. No significant correlation was found (r ⫽ 0.06; P ⫽ 0.26). The three eyes with residual anisometropia more than 2 D (one eye in the emmetropia group and two eyes in the monovision group) all had initial myopia more than 10 D.
Table 3. Postoperative Residual Anisometropia of Manifest Refraction Residual Anisometropia for Patients with Intended Correction Equal to Emmetropia Patients Before Surgery All patients 1 2 3 6 12 Retreated patients 12 months after retreatment†
Residual Anisometropia for Patients with Monovision
No. Patients*
n
⫾0.5 D (%)
0.5–1 D (%)
1–2 D (%)
⬎2 D (%)
n
⫾0.5 D (%)
0.5–1 D (%)
1–2 D (%)
⬎2 D (%)
418
307
63.8
20.9
10.4
4.9
111
58.6
18.9
15.3
7.2
342 215 285 319 317
251 152 211 236 235
75.3 82.2 84.4 85.6 85.1
18.7 10.5 10.9 10.6 11.5
5.2 6.6 3.8 3.4 3.0
0.8 0.7 0.9 0.4 0.4
91 63 74 83 82
57.1 54.0 51.4 55.4 56.2
27.5 34.9 31.1 28.9 28.0
14.3 9.5 16.2 14.5 13.4
1.1 1.6 1.3 1.2 2.4
4
3
66.7
33.3
0
0
0
0
1
100
0
D ⫽ diopters. *Patients who had an examination for both eyes at the same time after initial treatment. †Results for all patients were taken after the initial treatment. Results for patients who had a retreatment were taken 12 months after retreatment, because the postoperative follow-up dates were different between both eyes of the patients.
474
Pop and Payette 䡠 Results of Bilateral PRK
Figure 1. Twelve-month postoperative difference of the manifest refraction between the left and the right eye per initial degree of myopia of the left eye of a patient.
Regression and Astigmatism The mean individual regression between two time intervals was calculated. The 1- to 12-month regression rate was 0.16 ⫾ 0.54 D after the initial treatment, and was 0.58 ⫾ 1.04 D for retreated eyes. At 12 months after surgery, a maximum of 3.9% of eyes regressed more than 1 D after the initial treatment. At 12 months after surgery, 86% of eyes had final astigmatism within 0.5 D, and 98% of eyes had final astigmatism within 1 D. The average remaining cylinder at 12 months after surgery was ⫺0.29 ⫾ 0.34 D; the maximum remaining cylinder was ⫺2.25 D.
Visual Acuity At 12 months after surgery, 98% of eyes had UCVA of 20/40 or better and 87% of eyes had UCVA of 20/20 or better (Table 4). Sixty-eight percent of patients experienced no difference in UCVA between both eyes; 23% of patients had one UCVA line difference between both eyes; 5% of patient had a difference of two UCVA lines between both eyes; and 4% of eyes had a difference of more than two UCVA lines. The differences in BSCVA and UCVA lines
between both eyes decreased gradually over time (Table 4). Eight eyes (0.9%) of 7 patients had BSCVA more than 20/40, and 13 eyes (1.6%) of 11 patients had BSCVA equal to 20/40 before surgery. Although the percent of eyes with BSCVA more than 20/40 was 0.9% and 1.4% at 1 and 2 months after surgery, respectively, the percent of these eyes dropped to 0.5% of eyes at 12 months after surgery (Table 4). All the patients who had at least one eye with BSCVA more than 20/40 at 6 and 12 months after surgery also had initial BSCVA of 20/40 or more. A maximum loss of one Snellen line was seen in eyes with BSCVA of 20/40 or more at all times after surgery. Table 5 shows the number of patients who lost lines of BSCVA during the first postoperative year. All eyes were included in Table 5, even those that did not have a bilateral examination (fellow eye was retreated). The maximum number of eyes that lost lines of BSCVA occurred at the first and second postoperative month. No patient lost two lines of BSCVA in both eyes after the second postoperative month and no patient lost two lines of BSCVA in both eyes at 12 months after surgery. No patient lost any lines of BSCVA after retreatment at 2, 3, 6, or 12 months after surgery. All eyes, except one, that lost two or more Snellen lines of BSCVA at 1, 2, or 3 months after surgery were carefully monitored and regained at least one line at 6 or 12 months after surgery or after retreatment, for a final maximum loss of one line of BSCVA. The eye that lost two lines of BSCVA at 3, 6, and 12 months after surgery had a final outcome of 20/40 BSCVA with haze 1. Initially, this eye was treated for myopia of ⫺12.50 D. The fellow eye of this patient also experienced haze of 0.75 at 12 months after surgery. Both eyes were treated with corticosteroids for late haze. Fewer patients lost BSCVA lines bilaterally than those who lost lines of BSCVA in one eye only at all postoperative examinations. The odds ratio of patients who lost one line of BSCVA in one eye versus in both eyes was 0.44 (P ⫽ 0.002), and the odds ratio of patients who lost two lines of BSCVA in one eye versus in both eyes was 0.18 (P ⫽ 0.013). The odds ratio of patients who lost two lines of BSCVA in one eye versus the patients who lost two lines in one eye plus one line in the fellow eye was 0.27 (P ⫽ 0.032). Statistical tests were performed with low statistical power (20%– 40%).
Table 4. Postoperative Visual Acuities
Before Surgery All patients 1 2 3 6 12 Retreated patients 12 months after retreatment§
BSCVA Differences (Snellen Lines)‡
UCVA Differences (Snellen Lines)‡
No. Eyes
UCVA 20/20 or Better (%)*
UCVA 20/40 or Better (%)*
BSCVA >20/40 (%)*
No. Patients†
0 (%)
1 (%)
2 (%)
⬎2 (%)
0 (%)
1 (%)
2 (%)
⬎2 (%)
836
NA
NA
0.9
418
86.8
9.1
2.6
1.4
NA
NA
NA
NA
686 431 578 647 646
67.5 67.5 72.3 81.5 86.7
95.6 95.6 96.5 97.4 98.0
0.9 1.4 0.8 0.6 0.5
342 215 285 319 317
80.8 84.6 79.2 84.9 83.9
16.2 11.2 18.3 14.2 15.1
2.4 3.7 2.1 0.6 0.6
0.6 0.5 0.4 0.3 0.3
53.3 59.9 56.4 63.4 68.3
28.1 26.4 29.1 24.5 22.8
12.4 8.5 8.2 6.0 4.8
6.2 5.2 6.4 7.2 4.2
22
86.4
4
75.0
25.0
0
0
25.0
75.0
0
100
0
0
BSCVA ⫽ Snellen best spectacle-corrected visual acuity; NA ⫽ not available; UCVA ⫽ Snellen uncorrected visual acuity. *Percent of all eyes. †
Number of patients who had an examination for both eyes at the same time. Percent of patients who had an examination for both eyes at the same time after initial treatment. § Results for patients who had a retreatment were taken 12 months after retreatment, because the postoperative follow-up dates were different between both eyes of the patients. ‡
475
Ophthalmology Volume 107, Number 3, March 2000 Table 5. Loss of BSCVA Snellen Lines after Initial Treatment Months after Surgery
No. Patients
No. Eyes
1 Line (%)
ⱖ2 Lines (%)
One Eye
Both Eyes
One Eye
Both Eyes
1 2 3 6 12
344 216 293 328 329
686 431 578 647 646
38 (5.5) 13 (3.0) 22 (3.8) 9 (1.4) 3 (0.5)
8 (1.2) 6 (1.4) 2 (0.3) 1 (0.2) 1 (0.2)*
20 7 12 3 1
8 3 5 3 0
4 4 2 1 0
1 1 0 0 0
Total Cases
Odds ratio (probability) 95% Confidence interval
>2 Lines Lost
1 Line Lost
1:0.21 (⬍0.0001) 0.11–0.40
1:0.44 (0.002) 0.22–0.92
1:0.18 (0.013) 0.02–1.17
1 and 2 Lines in Both Eyes 2 0 0 0 1† 1:0.27 (0.032)* 0.05–1.42
BSCVA ⫽ best spectacle-corrected visual acuity. *Statistical test performed with frequency of eyes that lost two lines in one eye. †All eyes that lost two lines of BSCVA at 1, 2, or 3 months after surgery regained at least one line of BSCVA at 6 and 12 months after surgery except for one eye. The right eye of this patient had initial myopia of ⫺12.50 diopters and lost two lines of BSCVA at 3, 6, and 12 months after surgery, with a final BSCVA of 20/40 and was treated for late haze of 1.
Haze
Odds Ratio
Mean haze values were below 1 at all postoperative follow-ups (Table
Table 8 lists the maximum occurrence of refractive events that may occur during the first year after surgery. All 418 patients were evaluated after the initial procedure. These results should be carefully analyzed and compared with Tables 4, 5, 6, and 7 for correct interpretation of the final results. An odds ratio equal to 1 would mean that odds are even; the P value (Table 8) is the probability for the odds of being even. The odds ratios of Table 8 are between 0.18 to 0.44, except for the occurrence of halos, meaning that the proportion of bilateral events were 2.5 to 4 times lower than unilateral events. All odds ratios had significant probabilities of being different than 1 (equal odds), meaning that there was not a one-to-one relation between unilateral and bilateral events. Odds ratios could not be calculated for haze of 1 or more because only one eye had haze equal to 1 at 6 and 12 months after surgery. Table 9 is a recapitulation of events and risks to the fellow eye during the first postoperative year. The occurrence of refractive events for all eyes are included; risks to the fellow eye are calculated from all bilateral examinations taken from the previous tables. Table 9 is intended to portray a global picture of the risks generated by bilateral surgery. The risks to the fellow eye reported in Table 9 are the odds ratios calculated in Table 8. The product of
6), on the 0 to 4 scale used for the study.16 –18 All eyes were included in Table 6, even those that did not undergo a bilateral examination (fellow eye was retreated). All eyes had a difference of less than 0.5 of haze between both eyes at 6 and 12 months after surgery. No haze more than 1 occurred after all retreatments. One eye had haze equal to 1 in one eye and haze of 0.75 in the fellow eye; as reported previously, this eye lost two lines of BSCVA and was treated for late haze. The number of patients who had 0.5 to 0.75 haze in both eyes was higher than the number of patients who had low haze in only one eye (odds ratio, 1.724; P ⫽ 0.0008).
Halos Moderate and disturbing halos were considered problematic postoperative issues, although no patient reported disturbing halos at any time (Table 7). Mild halos were considered occasionally problematic. At 12 months after surgery, 13 patients (23 eyes) had moderate halos. The same patients complained about halos in both eyes, so that the percent of halos per eyes (Table 7) approximated the percent of halos per patients. The odds ratio of moderate halos in one eye versus in both eyes was 4.17 (P ⬍ 0.0001). All eyes were included in Table 7, even those eyes that did not have a bilateral examination (fellow eye was retreated).
Table 6. Postoperative Haze after Initial Treatment Total Cases Months after Surgery
No. Patients
No. Eyes
1 2 3 6 12
344 216 293 328 329
686 431 578 647 646
Mean Haze*
Haze ⫽ 1 (%)
One Eye
Both Eyes
One Eye
Both Eyes
Haze 0.5–0.75 and Haze of 1 in Both Eyes
0.22 ⫾ 0.15 0.22 ⫾ 0.18 0.16 ⫾ 0.17 0.09 ⫾ 0.16 0.04 ⫾ 0.11
70 (10.2) 81 (18.8) 52 (9.0) 41 (6.3) 14 (2.2)
0 (0) 0 (0) 0 (0) 1 (0.2) 1 (0.2)
16 7 10 17 8
27 37 21 12 3
0 0 0 0 0
0 0 0 0 0
0 0 0 1 1
Odds ratio (probability) 95% Confidence interval NA ⫽ not available. *Haze was assessed using a 0 to 4 scale.
476
Haze ⴝ 1
Haze 0.5–0.75
Haze 0.5–0.75 (%)
1:0.008 (⬍0.0001) 0.001–0.032
1:1.724 (0.0008) 1:10–2.70
NA NA
NA NA
Pop and Payette 䡠 Results of Bilateral PRK Table 7. Halos after Initial Treatment Total Cases Months after Surgery
No. Patients
No. Eyes
1 2 3 6 12
344 216 293 328 329
686 431 578 647 646
Mild Halos
Moderate Halos
Mild Halos (%)
Moderate Halos (%)
One Eye
Both Eyes
One Eye
Both Eyes
Mild and Moderate Halos in Both Eyes
155 (22.5) 79 (18.3) 133 (23.0) 144 (22.2) 113 (17.4)
53 (7.8) 23 (5.3) 39 (6.7) 38 (5.9) 23 (3.6)
9 10 5 10 6
72 34 63 66 53
7 2 5 2 2
22 10 16 17 10
2 1 2 2 1
Odds ratio (probability) 95% Confidence interval
1:0.28 (⬍0.0001) 0.23–0.35
the maximum occurrence (% of eyes) by the risk to the fellow eye should indicate the maximum occurrence of bilateral events. The maximum loss of two lines of BSCVA or more occurred in 1.4% of eyes during the first postoperative year with odds ratio to the fellow eye of 0.18 (Table 5). Moderate halos had a higher bilateral than unilateral incidence, occurring in 7.8% of eyes. A total of 3.9% of patients had retreatments with an odds ratio to the fellow eye of 0.22 (Table 8); overcorrections more than 1 D occurred in 6.9% of eyes, and undercorrections occurred in 4.7% of eyes, with odds ratios below 0.25 for both events (Table 8).
Discussion The results of this bilateral study were not expected to be better than the results of sequential surgery. However, the
1:7.20 (⬍0.0001) 4.85–10.69
1:4.17 (⬍0.0001) 2.17–8.03
1:0.44 (0.05) 0.14–1.38
present study showed the amount of postoperative residual anisometropia and odds ratios of different postoperative events after bilateral surgery. Similar to recent studies, the refractive outcomes of this study, as analyzed per eyes, are approximately the same or higher at the different follow-up examinations.3– 6 However, the results of this study were dependent on the initial amount of myopia per eye. It was estimated, because these patients were consecutively selected, that the distribution of patient’s myopia of this study may represent the average population of a surgeon with a private practice in refractive surgery. The literature has numerous references to evaluate the results of PRK for low, moderate, and high myopia.3,4,5,17,21,22 A study comparing sequential and bilateral procedures
Table 8. Statistics for All 418 Patients during the First Postoperative Year Events (Maximum Occurrence)
No. Patients
Odds Ratio (Probability)
95% Confidence Interval
One eye retreated Both eyes retreated
23 5
0.22 (0.001)
(0.08–0.58)
One eye with undercorrection more than 1 D Both eyes with undercorrection more than 1 D
29 5
0.17 (⬍0.0001)
(0.07–0.45)
One eye with overcorrection more than 1 D Both eyes with overcorrection more than 1 D
40 9
0.23 (⬍0.0001)
(0.11–0.47)
One eye with loss of BSCVA of one Snellen line Both eyes with loss of BSCVA of one Snellen line
20 8
0.44 (0.002)
(0.22–0.92)
0.27 (⬍0.032)† 0.18 (0.013)†
(0.05–1.42)† (0.02–1.17)†
One eye with loss of BSCVA of one Snellen line and one eye with loss of BSCVA of two lines or more in the fellow eye* One eye with loss of BSCVA of two Snellen lines or more Both eyes with loss of BSCVA of two Snellen lines or more*
2 4 1
One eye with haze more than 1 One eye with haze more than 1 and fellow eye with haze 0.5/0.75 Both eyes with haze more than 1
0 1 0
One eye with moderate halos Both eyes with moderate halos
7 22
NA 4.17 (⬍0.0001)
NA (2.17–8.03)
BSCVA ⫽ best spectacle-corrected visual acuity; D ⫽ diopters; NA ⫽ could not be statistically evaluated. *All eyes that lost two lines of BSCVA at 1, 2, or 3 months after surgery regained at least one line of BSCVA at 6 and 12 months except for one eye. This eye lost two lines of BSCVA at 3, 6, and 12 months after surgery and ended with BSCVA of 20/40 with haze of 1. †Statistical test was performed with very low power.
477
Ophthalmology Volume 107, Number 3, March 2000 Table 9. Recapitulation on the Occurrence of Events for All Eyes and Risks to the Fellow Eye During the First Postoperative Year (Consult Tables 5– 8)
Events
Maximum Occurrence (% of All Eyes)
Risk to Fellow Eye (%)
Undercorrection ⬍1 D Overcorrection ⬎1 D Loss of ⱖ2 lines of BSCVA* Haze ⱖ1 Moderate halos Retreatment
4.7 6.9 1.4 0.2 7.8 3.9
⫻ 17 ⫻ 23 ⫻ 18 NA ⫻ 417 ⫻ 22
BSCVA ⫽ best spectacle-corrected visual acuity; NA ⫽ could not be statistically evaluated. *Maximum occurred 2 months after surgery. No patient lost two lines of BSCVA in both eyes after second postoperative month.
would not specifically determine bilateral patterns for some patients, such as bilateral loss of BSCVA, bilateral haze, and bilateral halos. However, in a bilateral study, the first eye to undergo surgery could be evaluated as an unilateral procedure. In this study, the first operated eye and the fellow eye were compared as part of internal consistency and found not to be statistically significantly different. Bilateral PRK resulted in predictable refractive outcomes for both eyes in approximately 95% of cases. Using the slit-scanning laser, the percent of patients who lost two lines of BSCVA in both eyes was below 1.4% during the first postoperative year. The results showed that the risk of losing lines of BSCVA in both eyes was lower than the risk for one eye. As shown in Tables 8 and 9, bilateral events occurred 2.5 to 4 times less often than unilateral events, except for the occurrence of halos and haze of 0.5 to 0.75, on the 0 to 4 scale. With a 0.2% occurrence of haze more than 1, the present study showed that haze was not problematic. One patient had haze of 0.75 in one eye and haze of 1 in the fellow eye, as previously discussed, and was treated accordingly. No patient had haze more than 1 for both eyes during the first postoperative year. Although a greater number of patients had bilateral haze of 0.5 to 0.75, this fact was of no consequence at the end of the first postoperative year and was expected because PRK commonly creates nonsignificant haze. The slit-scanning laser also contributed to low haze after surgery.22 Halos were four times as common bilaterally compared with unilaterally during the first postoperative year. The odds ratio for halos showed a distinctive pattern toward bilateral halos. Therefore, this could be interpreted as some patients being at higher risk to perceive halos in both eyes. Because reports of halos were subjective, it is likely that the perception of halos in an eye may be influenced by global bilateral vision or even by the global satisfaction of the vision. Further studies may be required to evaluate halos, because the scale used for this study should be correlated to objective measures. Satisfaction of the global vision should also be correlated with the presence of halos. Pupil size, depth of the ablation, optical zones types, and sizes are also possible variables in this equation.
478
In this study, the amount of residual anisometropia after a bilateral procedure led to interesting benefits for the patient, as for visual recovery, with low risks to the fellow eye. Patients experienced only one period of postoperative healing, pain, and visual recovery. Although other studies with greater numbers of patients and longer follow-ups may be needed, the present study showed that the possibility of an adverse event in the fellow eye is approximately 1 in 4 if a complication occurs in the first eye. No pattern could be established to predict postoperative events, except for the incidence of postoperative halos for the fellow eye when the first eye is affected. Considering the probabilities of obtaining bilateral events, the bilateral probabilities can be easily evaluated by multiplying the maximum event incidence (% of all eyes) by the risk to fellow eye (Table 9). The probability of obtaining two consecutive events is the square product of the probability of one event. For example, obtaining two consecutive “ones” in two rows with a dice would be 1/6 ⫻ 1/6 ⫽ 1/36; P ⫽ 1/36 being the bilateral probability of obtaining “ones.” Using the present study results (Table 9), the bilateral probabilities were not the square product of unilateral probabilities, meaning that bilateral risks were not equal to the same unilateral risks applied to both eyes. Bilateral probabilities of event incidence were generally higher than the square product of unilateral probabilities, but were different, and statistically lower, than unilateral probabilities. The fact that the risk to the fellow eye was higher after a problem had occurred in the first eye can lead to the hypothesis that bilateral events are not independent. These rare events may be related to unidentified healing properties of the eye, or unknown predisposing factors that may take months to observe and that influence both eyes. However, the comparison of the unilateral and bilateral probabilities confirms that event incidence cannot be predicted using the results of a patient’s first operated eye. To reveal the true difference between unilateral and bilateral probabilities, a study with greater occurrence of events may be needed to increase statistical power. Bilateral surgery is still controversial. The present study does not state that bilateral simultaneous surgery is safer than sequential surgery, nor can the opposite be also stated. Bilateral surgery is a procedure with refractive benefits, and low risks, that should not be rejected by the back of the hand because of general usage and custom, and it has become a valid option for patients. None of the complications faced in the present study could have led a surgeon not to operate in the fellow eye. One of the risks not assessed in this study was the infection rate. Although no infection occurred during this study, risks of infection should be thoroughly explained to every patient before bilateral surgery. The present study does not offer a conclusion on the risk of bilateral infection. However, one may consider these risks to be extremely rare. By cumulating the study results of 9375 eyes, Maloney (see Discussion)23 showed that the probability of a catastrophic complication, including infections, after PRK or LASIK could be statistically estimated to be less than 1 in 3129. Using a ratio of 1:3 for bilateral events, he estimated the bilateral probability of catastrophic complication to be less
Pop and Payette 䡠 Results of Bilateral PRK than 1 in 10,000, and concluded that as long as the patient did not have an unusual risk factor, bilateral simultaneous photorefractive surgery was sufficiently safe, and that this type of surgery may be compared with other types of bilateral surgeries.
References 1. Goggin MJ, Kenna PF, Lavery FL. Photoastigmatic refractive keratectomy for compound myopic astigmatism with a Nidek laser. J Refract Surg 1997;13:162– 6. 2. Stephenson CG, Gartry DS, O’Brart DPS, et al. Photorefractive keratectomy. A 6-year follow-up study. Ophthalmology 1998;105:273– 81. 3. Tuunanen TH, Tervo TT. Results of photorefractive keratectomy for low, moderate, and high myopia. J Refract Surg 1998;14:437– 46. 4. Kim JH, Kim MS, Hahn TW, et al. Five year results of photorefractive keratectomy for myopia. J Cataract Refract Surg 1997;23:731–5. 5. Zadok D, Haviv D, Vishnevskia-Dai V, et al. Excimer laser photoastigmatic refractive keratectomy. Eighteen-month follow-up. Ophthalmology 1998;105:620 –3. 6. Carr JD, Stulting RD, Sano Y, et al. Prospective comparison of single-zone and multizone laser in situ keratomileusis for the correction of low myopia. Ophthalmology 1998;105:1504 – 11. 7. Grosvenor T. Primary Care Optometry: Anomalies of Refraction and Binocular Vision, 3rd ed. Boston: Butterworth-Heinemann, 1996. 8. Kohl M. To wait or not to wait: opinion varies on simultaneous bilateral surgery. J Refract Surg 1997;13:611–2. 9. Chynn EM. Merits of bilateral refractive surgery are debated. Ophthalmology Times, April 1st, 1999: p. 41. 10. Zaldivar R, Oscherow S, Ricur G, Piezzi V. Bilateral simultaneous laser in situ keratomileusis. J Refract Surg 1999;15(2 Suppl):S202– 8.
11. Gimbel HV, van Westenbrugge JA, Anderson Penno EE, et al. Simultaneous bilateral laser in situ keratomileusis: safety and efficacy. Ophthalmology 1999;106:1461– 68. 12. El-Maghraby A, Salah T, Polit F, et al. Efficacy and safety of excimer laser photorefractive keratectomy and radial keratotomy for bilateral myopia. J Cataract Refract Surg 1996;22: 51– 8. 13. Loewenstein A, Lipshitz I, Ben-Sirah A, et al. Symmetry of outcome after photorefractive keratectomy for myopia. J Refract Surg 1995;11(Suppl):S268 –9. 14. Gimbel HV, Van Westenbrugge JA, Johnson WH, et al. Visual, refractive, and patient satisfaction results following bilateral photorefractive keratectomy for myopia. Refract Corneal Surg 1993;9(Suppl):S5–S10. 15. Hashim A. Eye dominance in excimer laser photorefractive keratectomy. J Refract Surg 1997;13(Suppl):S461–2. 16. Pop M, Aras M. Multizone/multipass photorefractive keratectomy: six month results. J Cataract Refract Surg 1995;21:633– 43. 17. Pop M, Aras M. Photorefractive keratectomy retreatments for regression. One-year follow-up. Ophthalmology 1996;103:1979 – 84. 18. Fantes FE, Hanna KD, Waring GO III, et al. Wound healing after excimer laser keratomileusis (photorefractive keratectomy) in monkeys. Arch Ophthalmol 1990;108:665–75. 19. Pop M. Prompt re-treatment after photorefractive keratectomy. J Cataract Refract Surg 1998;24:320 – 6. 20. Dawson-Saunders B, Trapp RG. Basic and Clinical Biostatistics, 2nd ed. Norwalk, CT: Appleton & Lange, 1994. 21. Piovella M, Camesaca FI, Fattori C. Excimer laser photorefractive keratectomy for high myopia. Four-year experience with a multiple zone technique. Ophthalmology 1997;104: 1554 – 65. 22. Pop M, Payette Y. Multipass versus single pass photorefractive keratectomy for high myopia using a scanning laser. J Refract Surg 1999;15:444 –50. 23. Maloney RK. Discussion. Ophthalmology 1999;106:1467– 68.
479