Influence of the invariant refraction assumption in studies of formulas for monofocal and multifocal intraocular lens power calculation.

Abstract

To assess the influence in paired design studies of formulae comparison for intraocular lens (IOL) power calculation of using a single formula for deciding the implanted power with monofocal (mIOL) and multifocal (MIOL) lenses. Retrospective observational. Ninety-six right eyes were retrospectively analyzed. Eyes were assigned in two independent groups, SG and HG, depending on the formula used for deciding the implanted power, SRK-T (n = 54) and Haigis (n = 42), respectively. Median absolute prediction error (MedAE) was evaluated between independent samples (SRK-T in SG vs Haigis in HG) and between paired samples (SRK-T vs Haigis in both SG and HG). Percentages of eyes within a specific range of prediction error (PE) were also calculated for both, the standard steps and the clinically relevant steps. MedAE difference was lower than 0.09 D between both formulas for the comparison of independent samples in the mIOL (p = 0.62) and MIOL (p = 0.83) groups. However, paired samples resulted in better MedAE for SRK-T in the SG (0.14 D lower, p = 0.003) and for Haigis in the HG (0.07 D lower, p = 0.015), but only in the mIOL group. These small differences were also manifested, but not reaching statistical significance (p > 0.05), in the percentage of eyes achieving a specific range of PE, especially in the mIOL group. A small superiority for the formula used for selecting the final implanted IOL power can appear in studies following current standards. These studies should clearly specify which formula was used for selecting the implanted power.