Agreement and variability of subjective refraction, autorefraction, and wavefront aberrometry in pseudophakic patients
Manuel Ruiß, Oliver Findl, Petra Draschl, Annette Harrer-Seely, Nino Hirnschall
Abstract
PURPOSE: To evaluate the comparability and variability of subjective refraction (SR), autorefraction (AR), and wavefront aberrometry (WA) in pseudophakic patients. SETTING: Hanusch Hospital, Vienna, Austria. DESIGN: Prospective study. METHODS: Subjective refraction was performed by 2 independent examiners at 2 study visits. Furthermore, 5 measurements of AR and WA were performed in each patient at both visits. Agreement between the 3 refraction methods for spherical equivalent (M) and cylindrical vectors (J0, J45) was analyzed using Friedman multiple comparison and Bland-Altman plots. Predictability of spherical equivalent determined by SR from AR and WA measurements was tested with partial least squares regression and random forest regression. RESULTS: Ninety eyes of 90 pseudophakic patients were measured. The mean absolute error and arithmetic mean difference of SR measurements of M, J0, and J45 were comparable between both examiners. A small mean difference was found for SR between both visits. Spherical equivalent was more negative when measured by AR [-0.87 diopters (D)] and WA (-0.90 D) compared with SR (-0.60 D), whereas astigmatic vectors agreed well. Good test-retest reliability was found between all 3 refraction methods for M, J0, and J45. Partial least squares regression and random forest regression showed moderate predictive power for M measured by objective refraction and SR. CONCLUSIONS: Reproducibility and reliability of SR measurements in pseudophakic patients showed good agreement. AR and WA measured the spherical equivalent more myopic than SR, whereas astigmatic vectors were comparable between the 3 methods after uneventful cataract surgery.