Optimizing intraocular lens power calculations in eyes with axial lengths above 25.0 mm

PURPOSE: To evaluate the accuracy of refractive prediction of 4 intraocular lens (IOL) power calculation formulas in eyes with axial length (AL) greater than 25.0 mm and to propose a method of optimizing AL to improve the accuracy. SETTING: Cullen Eye Institute, Baylor College of Medicine, Houston, Texas, USA, and Department of Ophthalmology, Goethe University, Frankfurt am Main, Germany. DESIGN: Case series. METHODS: Refractive prediction errors with the Holladay 1, Haigis, SRK/T, and Hoffer Q formulas were evaluated in consecutive cases. Eyes were randomized to a group used to develop the method of optimizing AL by back-calculation or a group used for validation. Further validation was performed in 2 additional data sets. RESULTS: The optimized AL values were highly correlated with the IOLMaster AL (R(2) from 0.960 to 0.976). In the validating group, the method of optimizing AL significantly reduced the mean numerical errors for IOLs greater than 5.00 diopters (D) from +0.27 to +0.68 D to -0.10 to -0.02 D and for IOLs of 5.00 D or less from +1.13 to +1.87 D to -0.21 to +0.01 D, respectively (all P<.05). In 2 additional validation data sets, this method significantly reduced the percentage of eyes that would be left hyperopic. CONCLUSIONS: The proposed method of optimizing AL significantly reduced the percentage of long eyes with a hyperopic outcome. Updated optimizing AL formulas by combining all eyes from the 2 study centers are proposed.