Does Foveal Morphology Influence Postoperative Refractive Prediction Error Following Cataract Surgery?

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Background: In this study, we evaluated whether preoperative foveal morphology assessed using optical coherence tomography is associated with postoperative absolute refractive prediction error after uncomplicated cataract surgery. Methods: This retrospective observational study screened 435 patients and included 91 eyes of 91 patients after exclusions. All included patients underwent phacoemulsification with monofocal intraocular lens implantation. Preoperative spectral-domain optical coherence tomography measurements were used to calculate a foveal steepness index (FSI) as the ratio of the mean inner Early Treatment Diabetic Retinopathy Study (ETDRS) ring thickness to minimum foveal thickness, and foveal depth as the difference between these measurements. The primary outcome was SRK/T-derived absolute prediction error (APE) at one month. A supplementary analysis used Barrett Universal II. Associations were examined using Spearman correlation and univariable and multivariable linear regression analyses. A sensitivity analysis was performed after zeroing the cohort mean signed prediction error. Results: The mean age was 68.2 ± 7.8 years. The mean absolute prediction error was 0.48 ± 0.37 D, with a median of 0.50 D (IQR, 0.25–0.75 D). Overall, 70.3% and 95.6% of eyes had errors within 0.50 and 1.00 D, respectively. FSI and foveal depth were significantly associated with APE in univariable analyses (p = 0.028 and p = 0.013, respectively). The adjusted foveal-depth association was not statistically significant in the primary analysis (p = 0.079) but reached significance after zeroing the mean prediction error (p = 0.032). The supplementary Barrett Universal II analysis did not confirm an adjusted association before or after zeroing the mean signed prediction error. Adding foveal depth did not meaningfully improve leave-one-out cross-validated performance. Conclusions: These exploratory findings suggest a possible association between foveal morphology and SRK/T-derived APE in this study population. Reproducibility across IOL calculation formulas and clinical applicability remain uncertain; validation in a larger prospective cohort is needed.

This publication has 20 references indexed in Scilit: