Comparison of six methods for the intraocular lens power calculation in high myopic eyes

Abstract

Purpose: The aim of this study was to compare the accuracy of Barrett Universal II and Hill-Radial Basis Function with other four popular formulas for the calculation of intraocular lens power in high myopic eyes. Methods: A total of 56 eyes with an axial length of more than 26.0 mm were retrospectively reviewed. Six intraocular lens power calculation methods, including Barrett Universal II, Hill-Radial Basis Function, SRK/T, Haigis, Holladay 2 and Holladay 1, were evaluated. The difference between the postoperative actual refraction and the refraction predicted by the six methods was evaluated as the prediction error. The absolute prediction error was also calculated. Results: The mean numerical prediction error ± standard deviation of the six intraocular lens power calculation methods, in order of lowest to highest, was Barrett Universal II (0.37 ± 0.54 D), Hill-Radial Basis Function (0.40 ± 0.56 D), SRK/T (0.44 ± 0.56 D), Haigis (0.53 ± 0.54 D), Holladay 2 (0.88 ± 0.62 D) and Holladay 1 (1.00 ± 0.60 D). The median absolute errors predicted by the Barrett (0.46 D), Hill-Radial Basis Function (0.47 D), SRK/T (0.53 D) and Haigis (0.58 D) were significantly lower than those of the Holladay 1 (0.90 D) and Holladay 2(1.10 D; all p < 0.001). There was no significant difference among the median absolute errors of Barrett, Hill-Radial Basis Function, SRK/T and Haigis (all p > 0.05). Conclusion: The prediction errors differed for each method in the selection of intraocular lens power for the long eyes. In terms of overall accuracy, the Barrett Universal II formula provided the lowest prediction error. The Hill-Radial Basis Function method was comparable to the theoretical formulas, such as SRK/T and Haigis.