Comparison of Intraocular Lens Power Prediction Accuracy Between 2 Swept-Source Optical Coherence Tomography Biometry Devices

To compare intraocular lens (IOL) power prediction accuracy of the Eyestar 900 (EyeS900) and the IOLMaster 700 (IOLM700) based on estimated and measured posterior corneal power. Retrospective, interinstrument reliability study. Setting: Institutional. Participants: Two hundred twenty-five eyes of 225 cataract surgery patients. Measurements: Patients underwent measurements by both devices preoperatively. Main Outcome Measures: Spherical Equivalent Prediction Error (SEQ-PE), spread of the SEQ-PE (precision) and the absolute SEQ-PE (accuracy) of each device using Barrett Universal II (BUII) formula with either estimated posterior keratometry (E-PK) or measured posterior keratometry (M-PK). Trimmed mean SEQ-PEs of EyeS900 E-PK, EyeS900 M-PK, IOLM700 E-PK, and IOLM700 M-PK were 0.03, 0.08, 0.02, and 0.09 D, respectively with no significant differences between EyeS900 E-PK and IOLM700 E-PK ( P = 0.31) as well as between EyeS900 M-PK and IOLM700 M-PK ( P = 0.31). Statistically significant SEQ-PE differences were found when E-PK and M-PK were compared, regardless of the device used, showing hyperopic SEQ-PE in M-PK calculations. Excellent correlation and agreement in SEQ-PE were found between the devices for both E-PK ( P < 0.001, r = 0.848, mean bias: +0.01 D, 95% LOA of −0.32 to +0.34 D) and M-PK ( P < 0.001, r = 0.776, mean bias: −0.01 D, 95% LOA of −0.42 to +0.39 D). No significant differences were found comparing absolute SEQ-PE and precision of the devices. The Eyestar 900 and the IOLMaster 700 show comparable IOL power prediction accuracy by the BUII formula using either estimated or measured posterior keratometry. An adjusted lens factor may be required for BUII when utilizing measured posterior keratometry in both devices.