Purpose: To examine the risk of serious visual loss in contemporary LASIK.
Design: Retrospective, noncomparative, consecutive case series combined with a structured literature review and comparison with historical controls.
Participants: One thousand consecutive cases of LASIK performed from January 2000 to January 2004 by 1 surgeon. One thousand nine hundred eighty-two cases with 6-month review data in postmillennial Food and Drug Administration (FDA) trials of LASIK for myopia and astigmatism and 5203 patients with 6 month review data in similar premillennial FDA trials.
Methods: LASIK was performed with the Hansatome microkeratome and a VISX Star (S2, S3, or S4) laser. There were 899 myopic treatments (spherical equivalent < or =-10.5 diopters [D]; average -4.2 D; standard deviations [SD] D = 1.9 D; astigmatism < or =4.75 D), 87 hyperopic treatments (spherical equivalent < or =+4.25 D; average +2.4 D; SD = 0.9 D; astigmatism < or =3.5 D), and 14 treatments for mixed astigmatism (< or =4.5 D). Outcome data were obtained from a review of case notes and compared with historical data from FDA trials.
Main outcome measures: Best-spectacle corrected visual acuity (BSCVA) preoperatively and at final review was compared in our case series with reference to the United Kingdom driving standard (BSCVA> or =20/30). Subsidiary outcome measures included duration of follow-up, intraoperative, and postoperative complication rates. Six-month postoperative data on standard safety criteria (numbers of patients losing > or =2 lines BSCVA, BSCVA <20/40, and final BSCVA <20/25 for patients with preoperative BSCVA > or =20/20) were examined in FDA trials.
Results: In our clinical series, no case with a preoperative BSCVA of 20/20 or better had a final BSCVA of <20/30 at final review (97.5% confidence interval [CI], 0%-0.37%); 4 eyes had a BSCVA <20/25 (0.41%; 95%CI, 0.11%-1.02%). The median interval between surgery and final review was 2 months (range, 1-45 months). In FDA studies recruiting pre-2000, 1.4% of patients lost > or =2 lines of BSCVA versus 0.6% in postmillennial studies (P = 0.005); 0.45% of patients in premillennial studies lost BSCVA to <20/40 compared with no patients in postmillennial studies (P = 0.001); and 1.68% of patients in premillennial studies with a preoperative BSCVA > or =20/20 had a postoperative BSCVA <20/25 compared with 0.16% in postmillennial studies (P< or =0.001).
Conclusions: Compared with premillennial results, the risk of visual loss is significantly reduced in contemporary LASIK.