Evaluation of the Accuracy of Modern Toric Intraocular Lens Formulas Based on Anterior Corneal Measurements

We aimed to evaluate the postoperative refractive astigmatism prediction error (RAPE) of three modern toric intraocular lens (IOL) formulas, which estimate posterior corneal astigmatism, using anterior corneal measurements with a swept-source optical coherence tomography biometer, ARGOS^® (ARGOS).

This retrospective study included 33 eyes of 33 patients who underwent uneventful cataract surgery implanted with Clareon toric IOLs (model CNW0Tx) using digital guidance. Back-calculation was performed using the KANE toric, Emmetropic Verifying Optical (EVO) toric v2.0, and Barrett toric IOL formulas employing preoperative measurements with the ARGOS biometer, and surgically induced astigmatism was set to 0 diopter (D). RAPE was defined as the difference between the predicted refractive astigmatism and actual subjective refractive astigmatism 1 month after surgery. The Wilcox-Holladay-Wang-Koch statistical method was used to compare the toric IOL formulas.

Mean and standard deviation (SD) of RAPE for KANE toric, EVO toric, and Barrett toric formulas were -0.02 ± 0.44 D (95% CI: -0.18 to 0.14), 0.03 ± 0.44 D (95% CI: -0.13 to 0.19), and 0.03 ± 0.43 D (95% CI: -0.12 to 0.18), respectively. The median absolute RAPE were 0.20 D for the KANE toric, 0.23 D for the EVO toric, and 0.27 D for the Barrett toric. The centroid and SD of RAPE for each toric formula were 0.09 ± 0.46 D at 3.25°, 0.04 ± 0.50 D at 172.34°, and 0.02 ± 0.45 D at 103.50°, respectively. No significant difference was noted between the Barrett toric and KANE toric in the x axis (p = 0.001).

This study provides a further evaluation of the accuracy of modern toric IOL formulas for astigmatism prediction, even though only anterior corneal measurements from ARGOS were used for toric IOL calculations.