Refractive predictability using two optical biometers and refraction types for intraocular lens power calculation in cataract surgery

To evaluate the accuracy of intraocular lens (IOL) power calculation in relation to optical biometry devices and refraction types. Patients undergoing cataract phacoemulsification and insertion of the MX60 IOL were enrolled. Optical biometric measurements were performed with both IOLMaster 700 and Lenstar 900. Biometry measurements were compared between devices. A subsample of 133 eyes (81.1%) had examination for both autorefraction and subjective refraction postoperatively. The differences between the postoperative refraction and the refraction predicted by eight formulas (Kane, Hill-RBF 2.0, Barrett Universal II, Olsen, Haigis, SRK/T, Holladay 1 and Hoffer Q) were calculated. Overall, this study comprised 164 eyes of 164 patients. High agreement between the two biometers for axial length, average keratometry readings, anterior chamber depth, lens thickness and central corneal thickness was found (interclass correlation confidents: 0.999, 0.988, 0.965, 0.865 and 0.972, respectively, all P < 0.001). The absolute prediction error calculated with IOLMaster 700 measurements was significantly lower than that calculated with Lenstar 900 measurements for Olsen (P = 0.003), Haigis (P < 0.001) and Hoffer Q (P = 0.028). OPD-Scan III gave slightly more negative readings than subjective refraction (mean difference − 0.107 ± 0.553, P = 0.003 for spherical equivalent). However, no significant difference in absolute prediction error was found between the two refraction types per each formula. IOLMaster 700 and Lenstar 900 showed good agreement in biometric measurements with a trend toward better refractive outcome using IOLMaster 700. The accuracy of IOL calculation assessed with OPD autorefraction was equivalent to that assessed with subjective refraction.