Refractive outcomes of 8 biometric formulas in combined phacovitrectomy with internal limiting membrane peeling for epiretinal membrane

Abstract

Purpose:To investigate the accuracy of 8 different biometric formulas in combined phacovitrectomy and the effect of constant optimization on refractive outcome.Setting:Ludwig-Maximilians-University, Munich, Germany.Design:Retrospective observational case series.Methods:In this single-center study, patients with cataract and epiretinal membrane who underwent combined phacovitrectomy with internal limiting membrane peeling (Group B) and axial length-matched patients who underwent phacoemulsification (Group A) were included. In Group C, optimized constants from Group A were applied in patients of Group B. One eye of each patient was included. Main outcome measures after constant optimization for each biometric formula were refractive prediction error (PE), mean absolute error (MAE), and percentages of eyes with a PE within ±0.25 diopters (D), ±0.5 D, and ±1.0 D.Results:The study comprised 128 patients. For all formulas in Group A and Group B, refractive PE was 0.000 (P =.964 and P =.967, respectively). For formulas Barrett, Haigis, Hill, Hoffer Q, Holladay 1, Holladay 2, Kane, and SRK-T, refractive PE was -0.147, -0.204, -0.180, -0.212, -0.180, -0.178, -0.153, and -0.159, respectively, in Group C (P =.569); MAE was 0.346, 0.375, 0.382, 0.379, 0.355, 0.377, 0.318, and 0.364, respectively, in Group A (P =.286); 0.402, 0.422, 0.417, 0.427, 0.417, 0.402, 0.370, and 0.401, respectively, in Group B (P =.364); and 0.401, 0.424, 0.419, 0.444, 0.424, 0.404, 0.391, and 0.422, respectively, in Group C (P =.767). Effect of constant optimization in phacovitrectomy was statistically significant for all formulas (P