To compare the accuracy of the Barrett Universal II, Emmetropia Verifying Optical (EVO), Haigis, Kane, and SRK/T formulas for intraocular lens power calculation in patients with high axial myopia. In this retrospective study, 175 eyes (175 patients) that underwent uneventful cataract surgery were enrolled. According to the axial length (AL), the eyes were divided into long AL (26 ⩽ AL < 28 mm), super long AL (28 ⩽ AL < 30 mm), and extremely long AL (⩾ 30 mm). The mean absolute prediction errors (MAE) 3 months postoperatively and the percentage of eyes within different prediction error were compared, followed by subgroup analysis. The MAE and percentage of eyes within ±0.50 diopters (D) of the five formulas were as follows: Barrett Universal II (0.342, 74.9%), EVO 2.0 (0.314, 82.3%), Haigis (0.336, 74.9%), Kane (0.318, 78.9%), and SRK/T (0.398, 69.7%) (P = .552 and .071, respectively). Although no significant difference was found among the five formulas in the super and extremely long AL groups (P = .792 and .227, respectively), the EVO 2.0 formula achieved the highest accuracy (88.9%, 72 of 81) in the long AL group (P = .049). Moreover, the accuracy of the EVO 2.0 and Haigis formulas was stable, regardless of AL. The SRK/T formula showed a negative trend in the long and super long AL groups, whereas the Barrett Universal II, Kane, and SRK/T formulas showed positive trends in the extremely long AL group. Overall, the EVO 2.0 and Kane formulas achieved better results in patients with high axial myopia, whereas the other three formulas showed slightly poor outcomes. [J Refract Surg. 2021;37(11):754-758.].
Read full abstract