Comparison of the accuracy of six intraocular lens power calculation methods in post-myopic-LASIK eyes

To compare the accuracy of six intraocular lens (IOL) power calculation methods in post-myopic-LASIK eyes.

Post-myopic-LASIK patients scheduled for cataract surgery were enrolled. Mean error (ME), mean absolute error (MAE), and median absolute error (MedAE) of ASCRS (ASCRS-Max, ASCRS-Average, ASCRS-Min), EVO 2.0, Pearl-DGS, Barrett True-K no-history, Shammas-PL, and Haigis-L were compared. The ASCRS method employed four formulas, including Shammas-PL, Haigis-L, Potvin-Hill Pentacam, and Barrett True-K no-history. Trueness, precision, and accuracy indices were evaluated by comparing trimmed-mean values with heteroscedasticity adjustment. Subgroup analyses were performed based on K value, axial length (AL), and corneal decentered ablation, respectively. Factors influencing prediction errors were analyzed.

Totally, 87 eyes were analyzed. ASCRS-Min had the lowest MAE (0.79 D) and MedAE (0.62 D), followed by EVO 2.0, Pearl-DGS, and Barrett True-K no-history. It also had the highest percentage of absolute errors within 1.00 D. All methods outperformed ASCRS-Max in accuracy indices, and ASCRS-Min and EVO 2.0 showed superior accuracy indices compared to Shammas-PL and Haigis-L. In the subgroups of lower K value, longer AL, and larger decentration of ablation zone, ASCRS-Min, EVO 2.0, Pearl-DGS, and Barrett True-K no-history generated statistically lower MedAEs than ASCRS-Max. The accuracy of Shammas-PL was associated with AL; the accuracy of Haigis-L was associated with K value and AL; no other significant associations were found.

Generally, ASCRS-Min, EVO 2.0, Pearl-DGS, and Barrett True-K no-history achieve relatively better accuracy than the other methods, which might be considered as first choices for IOL power calculation in post-myopic-LASIK eyes.