Optimizing IOL calculation in triple-DMEK: data from a real-life cohort

Abstract

Aim To enhance the accuracy of intraocular lens (IOL) power calculation in patients undergoing simultaneous cataract surgery and Descemet membrane endothelial keratoplasty (triple-DMEK) by predicting corneal power changes.Methods Observational ambispective monocentric cohort study. Real life refractive outcomes and pre/post-operative corneal data from the IOL Master 700 (Zeiss) were merged in a linear corneal change model (LCCM), including traditional and new-generation formulae. The model was compared with the outcomes of optimizing IOL constants specifically for Triple-DMEK.Results 97 eyes of 69 patients with a follow-up of more than 3 months were analysed. Preoperative keratometry was unavailable in 9 eyes, in which IOL calculation was performed by manually with autorefractometer data and were analysed separately. Mean absolute error (MAE) for the former group was higher than for the measured one, 1.35 ± 1.16 D and 0.75 ± 0.70 D respectively. The median change in simulated keratometry (SimK) was − 0.21 ± 0.68 D and in total keratometry (TK) was − 0.62 ± 1.09 D, both statistically significant (p < 0.001). The lowest MAE was achieved by SRK/T with constant optimization (0.60 ± 0.58 D). The LCCM applied to any formula, showed similar MAE results to the constant optimization method (p > 0.05). However, the maximum absolute error for the optimization method was 2.08 ± 0.16 D, while for the LCCM method was 1.87 ± 0.13 D.Conclusions Keratometric readings in corneas with Fuchs´ dystrophy should be carefully measured. SimK and TK show a significant change during Triple-DMEK surgery. Optimizing IOL constants for this surgery improves the MAE, while using the LCCM could lead to a reduction of refractive surprises.