Accuracy of IOL Power Calculation in Eyes Needing a Myopic Target

Purpose: To compare the accuracy of formulas for intraocular lens (IOL) power calculation in eyes needing myopic targets between −2.50 and −3.00 diopters (D). Methods: Patients who underwent phacoemulsification needing a myopic target were enrolled in a single-center retrospective study. IOL power was calculated with newer formulas (Barrett Universal II [BUII], Cooke K6, EVO 2.0, Hoffer QST, and Kane) and older ones (Haigis, Hoffer Q, Holladay 1 and 2, and SRK/T). Trueness, precision, and accuracy of the prediction error (PE) and the percentages of eyes within certain PE thresholds were compared. Patients were grouped based on the implanted IOL (iSert255 [Hoya] or SN60WF [Alcon Laboratories, Inc]). Results: One hundred thirty eyes were enrolled in the iSert 255 group and 105 eyes in the SN60WF group. No statistically significant difference was observed in trueness among formulas, due to constant optimization. In the iSert 255 group, the accuracy of the Holladay 1 formula was lower ( P < .05) compared to the Cooke K6, EVO 2.0, and Kane formulas. In the SN60WF group, the accuracy of the Holladay 1 formula was lower ( P < .005) compared to the BUII, EVO 2.0, Hoffer QST, and Kane formulas. The percentage of eyes with an absolute PE less than 0.50 D was highest with newer formulas (77% to 81% in the iSert 255 group and 83% to 85% in the SN60WF group) and lowest with the Holladay 1 formula (58 and 69%, respectively). Conclusions: Newer formulas have good accuracy in eyes with intentional myopic targets, whereas the use of older formulas, especially the Holladay 1, is not recommended.