Purpose
To compare the incidence and timing of postoperative complications between laser-assisted in situ keratomileusis (LASIK) and photorefractive keratectomy (PRK) within a national clinical registry.
Design
Retrospective registry-based clinical cohort study.
Participants
A total of 48,892 eyes of 27,372 patients who underwent refractive surgery in the IRIS Registry (Intelligent Research in Sight).
Methods
LASIK and PRK procedures were identified using Healthcare Common Procedure Coding System codes. Postoperative complications, including dry eye disease (DED), visual disturbances, and ocular pain, were identified using International Classification of Diseases codes and analyzed at the eye level. Incidence was evaluated across three postoperative intervals (1 day to 1 month, 1 to 3 months, and 3 to 12 months). Interval-specific incidence rates and incidence rate ratios (IRRs) comparing LASIK and PRK were estimated using multivariable-adjusted Poisson generalized estimating equations.
Main Outcome Measures
Interval-specific incidence of postoperative complications and comparative IRRs between LASIK and PRK.
Results
DED was the most frequently documented postoperative complication, occurring in 16.1% of LASIK eyes and 15.6% of PRK eyes over the follow-up period. During the 1 to 3 months interval, LASIK was associated with a lower incidence of DED compared with PRK (IRR, 0.8 [95% confidence interval (CI), 0.7-0.9]). In the first postoperative month, visual disturbances occurred more frequently after PRK (IRR, 0.3 [95% CI, 0.3-0.4]), but differences between procedures were not significant at later intervals. Ocular pain was rare after both procedures, with no consistent differences in early intervals, but showed a lower incidence after LASIK in the 3 to 12 months interval (IRR, 0.5 [95% CI, 0.3-0.9]).
Conclusions
Postoperative complication profiles following refractive surgery varied by procedure and postoperative interval. PRK was associated with a higher risk of visual disturbances in the early postoperative period and a higher incidence of DED during the 1 to 3 months interval compared with LASIK. Ocular pain was rare overall but showed a lower incidence after LASIK in the 3 to 12 months interval, whereas most other complication rates were comparable between procedures. Recognition of these temporal patterns may help inform patient counseling, optimize postoperative follow-up strategies, and guide postoperative management during different phases of corneal recovery after refractive surgery.
INTRODUCTION
R efractive surgery, including laser-assisted in situ keratomileusis (LASIK) and photorefractive keratectomy (PRK), is among the most frequently performed elective procedures in ophthalmology. These procedures provide highly effective correction of refractive error and are associated with high patient expectations for long-term safety, visual quality, and satisfaction. ,,,,
Although LASIK and PRK achieve comparable visual outcomes, their differing surgical approaches result in distinct corneal wound-healing responses and complication profiles. Advances in technology have substantially reduced the overall risk of adverse events, yet postoperative complications continue to influence visual quality, patient comfort, and long-term satisfaction. According to the Academy Refractive Surgery Preferred Practice Pattern (PPP), three of the most commonly observed complications after both LASIK and PRK are dry eye disease (DED), visual disturbances, and ocular pain. ,,,,,
Despite extensive literature on refractive surgery outcomes, the timing of these complications in routine clinical practice remains poorly characterized. In particular, data describing how complication risk evolves across discrete postoperative intervals are limited. Understanding these temporal patterns is essential for optimizing follow-up schedules, guiding prophylactic interventions, and tailoring postoperative management to the expected stage of corneal recovery.
The purpose of this retrospective clinical cohort study was to use the IRIS Registry (Intelligent Research in Sight) to cf incidence and timing of key postoperative complications following LASIK and PRK in a large, real-world population.
METHODS
The IRIS Registry is a centralized data repository and reporting tool that can be used for research purposes. This does not constitute human subject research because data in the IRIS Registry are deidentified, and the investigator does not have access to study identifiers. Therefore, institutional review board and informed consent are not required. This study adheres to the Declaration of Helsinki.
STUDY POPULATION AND PROCEDURE IDENTIFICATION
Adults aged 18 years or older who underwent laser vision correction between January 2, 2013, and March 29, 2024, were eligible for inclusion. Refractive procedures were identified using Healthcare Common Procedure Coding System (HCPCS) codes S0810 for LASIK and S0800 for PRK. To ensure correct identification of excimer laser procedures, each surgical event was required to have documentation of a perioperative topical medication, such as an ophthalmic antibiotic, corticosteroid, or nonsteroidal anti-inflammatory drugs, within the 3 days preceding or 7 days following the procedure date. Laterality was obtained from procedure coding; when laterality was not explicitly recorded, procedures occurring less than 3 months apart were assumed to represent treatment of different eyes and not enhancement of the first eye, in alignment with the Academy Refractive Surgery PPP.
EXCLUSION CRITERIA
Eyes were excluded if they had prior refractive surgery or had International Classification of Diseases 9th Revision (ICD-9) and 10th Revision (ICD-10) codes for keratoconus or corneal ectasia (Supplemental Table 1), an established contraindication for refractive surgery. Eyes were also excluded if follow-up was less than 6 months after surgery. Because demographic completeness varies across IRIS Registry practices, eyes from patients with missing data on age, sex, race, ethnicity, or income category were excluded to avoid misclassification. Practice region was the only demographic variable permitted to be missing and was categorized as “unknown.”
DEMOGRAPHIC AND POSTOPERATIVE COMPLICATION VARIABLES
Demographic characteristics were assessed at the patient level and included age at surgery, sex (male or female), race, ethnicity, income category, and surgeon practice region. Postoperative complications were identified using diagnostic codes or structured test values and were classified as incident if first recorded after the index refractive procedure. Preexisting conditions such as DED, visual disturbances, and ocular pain were not used as exclusion criteria, as they are not established contraindications to refractive surgery.
Postoperative DED was defined only when all criteria were met in combination, including abnormal objective testing (Schirmer’s test ≤10 mm or tear break-up time <10 seconds), relevant diagnostic coding, and postoperative initiation or documentation of topical medications indicated for DED management (Supplemental Table 1). Visual disturbances and ocular pain are patient-reported, subjective symptoms; therefore, documentation on at least two separate postoperative encounters with relevant ICD-9 and ICD-10 codes was required to improve specificity and reduce miscoding of transient or nonspecific symptoms (Supplemental Table 1).
FOLLOW-UP INTERVALS
To characterize postoperative risk over clinically relevant intervals, incidence was evaluated within three time periods after surgery selected to correspond with distinct clinical phases of refractive recovery: acute epithelial and flap healing (1 day to 1 month), refractive stabilization and tapering of topical medications (1 to 3 months), and late remodeling (3 to 12 months). These intervals reflect expected stages of postoperative healing and follow recommendations from the Academy Refractive Surgery PPP. Demographic characteristics were summarized at the patient level, whereas postoperative complications were analyzed at the eye level.
STATISTICAL ANALYSIS
For each postoperative complication, incidence was evaluated across three predefined postoperative intervals measured from the surgery date (1 to 30 days, 31 to 90 days, and 91 to 365 days). The day of refractive surgery was defined as day zero. Eyes began contributing eye-time (total follow-up time per eye) on postoperative day 1 and entered subsequent intervals only if they remained complication-free at the start of each interval. Within each interval, eyes were followed until the earliest of: (1) onset of the complication, (2) the end of the interval, or (3) the last documented follow-up. Interval-specific cumulative incidence was calculated as the number of eyes developing a new complication during the interval divided by the number of eyes at risk at the start of that interval. Incidence rates were computed as events per eye-year, with 95% confidence intervals (CIs) derived under a Poisson assumption accounting for within-patient correlation. To cf LASIK and PRK, incidence rate ratios (IRRs) with 95% CI were estimated separately for each complication and interval using multivariable-adjusted Poisson generalized estimating equations models accounting for patient-level correlation and baseline covariates (age at surgery, sex [male or female], race, ethnicity, income category, and surgeon practice region, medical history [ie, DED, visual disturbances, and ocular pain], spherical equivalent, surgical year). A pooled model including procedure type, complication type, follow-up interval, and their interaction terms was also used to estimate overall covariate effects across outcomes and time intervals. All analyses were performed using R Statistical Software (Version 4.5.1; Foundation for Statistical Computing).
RESULTS
STUDY POPULATION
A total of 48,892 eyes of 27,372 patients who underwent refractive surgery in the IRIS Registry between 2013 and 2024 were included in the study cohort; 22,263 patients (81.3%) who had LASIK, and 5109 patients (18.7%) had PRK ( Table 1 , Figure 1 ). Patient-level demographic analysis showed that the median age at surgery was similar between groups (39 years [IQR, 31-52 years] for LASIK and 40 years [IQR, 30-57 years] for PRK; P <.001). Women comprised 57.4% of the LASIK cohort and 54.8% of the PRK cohort. Most patients in both groups were White, with slightly higher proportions of Asian and Black or African American patients in the PRK group. Median income category was predominantly in the $35,000 to $74,999 and $75,000 to $149,999 ranges for both LASIK and PRK, with relatively few patients in the lowest and highest income strata. Geographic distribution differed between procedures: LASIK was more frequently performed in the Midwest (43.5% of LASIK patients), whereas PRK had a larger proportion of patients from the South (37.7%) and West (22.4%). Bilateral surgery was more common among LASIK patients (81.9%) compared with PRK patients (64.3%).
TABLE 1
Demographic Characteristics of Patients With LASIK and PRK, 2013 to 2024 in the IRIS Registry.
| Variable | Total | LASIK | PRK | P Value |
|---|---|---|---|---|
| N = 27,372 | N = 22,263 | N = 5109 | ||
| Age, median years (Q1, Q3) | 39.00 (31.00, 53.00) | 39.00 (31.00, 52.00) | 40.00 (30.00, 57.00) | <.001 |
| Sex, n (%) | <.001 | |||
| Female | 15,583 (56.9%) | 12,782 (57.4%) | 2801 (54.8%) | |
| Male | 11,789 (43.1%) | 9481 (42.6%) | 2308 (45.2%) | |
| Race, n (%) | <.001 | |||
| White | 25,700 (93.9%) | 20,963 (94.2%) | 4737 (92.7%) | |
| Asian | 1078 (3.9%) | 831 (3.7%) | 247 (4.8%) | |
| Black or African American | 594 (2.2%) | 469 (2.1%) | 125 (2.4%) | |
| Median income category, n (%) | <.001 | |||
| <$34,999 | 171 (0.6%) | 124 (0.6%) | 47 (0.9%) | |
| $35,000-$74,999 | 12,245 (44.7%) | 10,023 (45.0%) | 2222 (43.5%) | |
| $75,000-$149,999 | 14,395 (52.6%) | 11,701 (52.6%) | 2694 (52.7%) | |
| >$150,000 | 561 (2.0%) | 415 (1.9%) | 146 (2.9%) | |
| Surgeon practice region, n (%) | <.001 | |||
| Midwest | 10,612 (38.8%) | 9684 (43.5%) | 928 (18.2%) | |
| South | 5830 (21.3%) | 3906 (17.5%) | 1924 (37.7%) | |
| West | 4778 (17.5%) | 3633 (16.3%) | 1145 (22.4%) | |
| Unknown | 3709 (13.6%) | 3027 (13.6%) | 682 (13.3%) | |
| Northeast | 2443 (8.9%) | 2013 (9.0%) | 430 (8.4%) | |
| Laterality, n (%) | <.001 | |||
| Bilateral | 21,520 (78.6%) | 18,233 (81.9%) | 3287 (64.3%) | |
| Left | 2433 (8.9%) | 1739 (7.8%) | 694 (13.6%) | |
| Right | 3419 (12.5%) | 2291 (10.3%) | 1128 (22.1%) |
Demographic analysis is patient-level analysis.
Bolded P values indicate statistical significance ( P <.05)
Flowchart of patient selection and study cohort definition.
SUMMARY OF HISTORY AND POSTOPERATIVE COMPLICATIONS
Among surgical eyes, preoperative spherical equivalent was −1.25 D (IQR, −3.88 to 0.00 D) for LASIK eyes and −0.75 D (IQR, −3.50 to 0.00 D) for PRK eyes. The history of DED was most common, present in 5553 of 40,496 LASIK eyes (13.7%) and 1477 of 8396 PRK eyes (17.6%). A history of visual disturbances was documented in 821 LASIK eyes (2.0%) and 269 PRK eyes (3.2%), whereas a history of ocular pain was recorded in 154 LASIK eyes (0.4%) and 57 PRK eyes (0.7%) ( Table 2 ).
TABLE 2
Summary of History and Postoperative Complications Per Eye After LASIK and PRK, 2013 to 2024 in the IRIS Registry.
| Complication | Total | LASIK | PRK |
|---|---|---|---|
| N = 48,892 Eyes | N = 40,496 Eyes | N = 8396 Eyes | |
| History of dry eye disease, n (%) | 7030 (14.4%) | 5553 (13.7%) | 1477 (17.6%) |
| History of visual disturbances, n (%) | 1090 (2.2%) | 821 (2.0%) | 269 (3.2%) |
| History of ocular pain, n (%) | 211 (0.4%) | 154 (0.4%) | 57 (0.7%) |
| Postoperative dry eye disease, n (%) | 7826 (16.0%) | 6513 (16.1%) | 1313 (15.6%) |
| Postoperative visual disturbances, n (%) | 320 (0.7%) | 227 (0.6%) | 93 (1.1%) |
| Postoperative ocular pain, n (%) | 36 (0.1%) | 29 (0.1%) | 7 (0.1%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
