Impact of Biometric Parameters on the Predictive Accuracy of Intraocular Lens Power Calculation Formulas Included in the ESCRS IOL Power Calculator

Purpose: To analyze the impact of keratometry, axial length (AL), anterior chamber depth (ACD), lens thickness (LT), central corneal thickness (CCT), and white-to-white distance (WTW) on the predictive accuracy of formulas included in the European Society of Cataract and Refractive Surgery (ESCRS) IOL power calculator. Methods: This retrospective case series included 748 eyes from 748 patients. Postoperative refraction data were used to calculate refractive prediction errors (PEs) for the Barrett, Cooke K6, EVO, Hill-RBF, Hoffer QST, Kane, and Pearl-DGS formulas. Low, medium, and high subgroups were created based on the value of selected anatomical parameters. Statistical analysis included standard deviation, root mean square absolute errors (RMSAE), and stepwise multiple regression. Results: Significant differences in formula performance were observed across subgroups. The Hoffer QST had a higher standard deviation than the Cooke K6, Kane, and Pearl-DGS in the subgroup with steep keratometry and a higher standard deviation than the Pearl-DGS in the subgroup with thin LT. The Kane had a lower standard deviation than the Cooke K6 in the subgroup with shallow ACD and a lower RMSAE than the EVO in the subgroup with deep ACD. The Cooke K6 had a lower standard deviation than the Hoffer QST in the subgroup with deep ACD. Anatomical parameters that significantly contributed to the PEs were: keratometry for the Pearl-DGS; AL for all formulas except the Hill-RBF; ACD for the Hill-RBF and Hoffer QST; LT for the Hill-RBF, Hoffer QST, and Pearl-DGS; CCT for the Pearl-DGS; and WTW for the Barrett and Kane. Conclusions: Anatomical parameters significantly influence the accuracy of modern IOL power calculation formulas. These findings suggest that tailoring formula selection based on specific anatomical characteristics can improve refractive outcomes in cataract surgery.