Accuracy of intraocular lens power calculation using three optical biometry measurement devices: the OA-2000, Lenstar-LS900 and IOLMaster-500.

Purpose: To compare ocular measurements of three optical biometry devices and their application in intraocular lens (IOL) power calculations.

Methods: One hundred and forty eyes which had undergone cataract extraction surgery with preoperative biometry with OA--2000, IOLMaster-500, and Lenstar-LS900 were enrolled. Biometry measurements of the three devices were compared. The deviation of the postoperative refraction from the preoperative refractive target was calculated with different formulas (Barrett Universal II, Hoffer Q, Holladay I, and SRK/T). Errors in the predicted astigmatism using the Barrett toric calculator were calculated for the toric IOLs. Additional 6465 eyes in which the IOLMaster-500 failed to measure axial length (AL) were reviewed. The percentage of successful measurements using the OA-2000 in those eyes was calculated.

Results: High agreement was found between the three devices for AL, anterior chamber depth, and average keratometry measurements (interclass correlation confidents: 1.000, 0.970, and 0.998, respectively, P < 0.001). The mean absolute prediction errors were similar using all formulas, ranging from 0.25 to 0.29 D, with no statistical significant difference between the three devices per each formula. The OA-2000 yielded a lower against-the-rule (ATR) centroid error in the predicted astigmatism than the IOLMaster-500 and Lenstar-LS900 (0.06 D ± 0.59 at 13.4° vs. 0.20 D ± 0.61 at 14.8° and 0.16 D ± 0.55 at 21.4°, respectively, P < 0.001, X-axis). Among 301 cases with unsuccessful AL readings using the IOLMaster-500, the OA-2000 had 284 (94.35%) successful measurements.

Conclusions: The OA-2000 measurements showed good agreement with those of the IOLMaster-500 and Lenstar-LS900. Our results may suggest a potential advantage of the OA-2000 device in toric IOLs calculations and AL measurement success rate.