Abstract / Summary
Background/Objectives: To compare treatment-zone decentration after small-incision lenticule extraction (SMILE) performed with VISUMAX 500 and VISUMAX 800 using corneal thickness difference maps derived from anterior segment optical coherence tomography. Methods: This retrospective comparative study included 44 eyes of 22 participants who underwent SMILE with VISUMAX 500 or VISUMAX 800 by a single surgeon and had preoperative and 3-month postoperative CASIA2 examinations. Treatment-zone centers were primarily identified manually relative to the corneal vertex. Between-platform differences were analyzed using generalized estimating equations (GEE) accounting for inter-eye correlation, with additional adjustment for preoperative cylinder. Participant-level analyses and semi-automated radial color-symmetry localization were performed as sensitivity analyses. Results: Twenty-two eyes of 11 participants were included in each group. Mean radial decentration was 0.219 ± 0.134 mm with VISUMAX 500 and 0.225 ± 0.171 mm with VISUMAX 800. The mean between-group difference was 0.006 mm (95% confidence interval [CI], −0.074 to 0.085; p = 0.890), and remained similar after adjustment for preoperative cylinder (0.006 mm; 95% CI, −0.082 to 0.094; p = 0.888). Participant-level analysis likewise showed no evidence of a between-group difference (mean difference, 0.0056 mm; 95% CI, −0.083 to 0.094; exact permutation p = 0.907). Reanalysis using semi-automated localization yielded a similarly small between-platform difference (0.0114 mm; 95% CI, −0.063 to 0.086; p = 0.763). Conclusions: In this small single-surgeon cohort, treatment-zone decentration did not differ significantly between VISUMAX 500 and VISUMAX 800. The findings were consistent across participant-level and semi-automated sensitivity analyses. Because the study was not designed as an equivalence or noninferiority trial, these results should be interpreted as an absence of evidence for a between-platform difference rather than evidence of equivalent centration performance.