COMPARATIVE ACCURACY OF FIVE MODERN TORIC IOL FORMULAS

To compare the clinical accuracy of five modern toric intraocular lens (IOL) calculation formulas—Barrett, EVO, Naeser/Savini combined with Hoffer QST, Kane, and Abulafia-Koch combined with Hill RBF—in predicting postoperative refractive outcomes in cataract patients with astigmatism.

Eunpyeong St.Mary’s Hospital, Seoul, Republic of Korea

This retrospective, comparative study included 146 eyes from 146 cataract patients who underwent toric intraocular lens (IOL) implantation. Biometric measurements were obtained using the Lenstar LS900, with personalized surgically induced astigmatism values applied for each surgeon. Postoperative outcomes, including mean absolute prediction error (MAE), centroid error, and the proportion of eyes within a prediction error of ±0.50 D and ±0.75 D, were evaluated. Subgroup analyses were performed for with-the-rule (WTR), against-the-rule (ATR), and oblique astigmatism. Statistical significance was assessed using non-parametric and multivariate tests.

No significant differences in MAE were observed among the formulas in the overall cohort (p > 0.05). The proportion of eyes with a prediction error ≤ ±0.50 D was highest for EVO (71.92%), followed by Kane and Abulafia-Koch with Hill RBF (70.55%), Naeser/Savini with Hoffer QST (69.18%), and Barrett (67.12%). Among ATR cases, EVO, Abulafia-Koch with Hill RBF, and Naeser/Savini with Hoffer QST demonstrated statistically lower MAEs compared to Barrett (p < 0.05).

All five toric IOL calculation formulas demonstrated excellent overall accuracy, with slight variations observed in specific subgroups. The EVO formula showed superior precision in ATR astigmatism. These findings provide real-world insights to enhance refractive outcomes and optimize formula selection in toric IOL surgery.