IOL Power Calculation Project: Accuracy of 36 Formulas.

Abstract

To ascertain the refractive accuracy of 36 intraocular lens (IOL) power calculation formulas in unoperated eyes. Retrospective accuracy and validity analysis. Six hundred fifty-five patients undergoing phacoemulsification and implantation of the Tecnis 1 ZCB00 IOL (Johnson & Johnson Vision, Jacksonville, FL, USA). Thirty-six formulas were evaluated including some that have never been tested, such as 3C 2.0, Eom, Hoffer H, Hoffer H-5, Fam adjusted methods, Norrby Regression Formula (Norrby RF), Norrby thin lens paraxial Ray-Tracing (Norrby RT), and VRF Cooke modified axial length (VRF CMAL). Optical biometry with the IOLMaster 700 (Carl Zeiss Meditec AG) was performed preoperatively. All descriptive statistics and the percentage of eyes within prediction error thresholds were evaluated with optimized lens constants. The Formula Performance Index (FPI) and Formula Performance Index for subgroup (FPI sub) were used as the primary formula outcomes. The highest FPI indexes were yielded by the VRF-G (0.590), Hoffer QST (0.575), VRF CMAL (0.574), Eom (0.572), EVO 2.0 (0.569), and Kane (0.568) formulas. The heteroscedastic test revealed statistically significant differences (P < .05) among formulas. The standard deviation (SD) of VRF-G (0.353 D), EVO 2.0 (0.362 D), Kane (0.366 D), Hoffer QST (0.371 D), and Eom (0.372 D) were lower than other methods (P < .05). The highest percentage of eyes with a PE within ±0.50 D was achieved by VRF-G (87.48%), Kane (86.41%), Hoffer QST (86.26%), and PEARL-DGS (86.26%). Contemporary IOL power calculation formulas (Eom, EVO 2.0, Hoffer QST, VRF CMAL, and VRF-G) improved accuracy in all axial length ranges compared to traditional and updated methods. The CMAL method raised the accuracy of the VRF formula.