Prediction of spectacle-corrected visual acuity using videokeratography

PURPOSE: Our aim was to improve prediction of spectacle-corrected visual ac uity (SCVA)using indices derived from the EyeSys System 2000 data (version 3.1). METHODS: We studied corneal topography in 182 eyes from 8 groups of patient s. Holladay Diagnostic Summary indices were recorded. Nine statistical indi ces calculated with the first 8-ring data and refractive power symmetry ind ex were also studied. Correlation with SCVA (LogMAR units) was studied by m eans of Pearson's regression. Multiple linear regression was used to obtain linear equations combining several indices. RESULTS: At a univariate level, total astigmatism cylinder showed the stron gest correlation with SCVA (r = .63, P = .0001). At a multivariate level, t he predicted visual acuity obtained by linear equation combining the aspher icity coefficient, the predicted corneal acuity, the mean of the means, and the total astigmatism cylinder was closely associated with SCVA (r = .72, P = .0001). It was identical to SCVA in 58.2% of the cases, within one line in 75.8%, and within two lines in 91.2%. CONCLUSION: Multiple linear regression resulted in the best prediction of s pectacle-corrected visual acuity, giving notable improvement in prediction of spectacle-corrected visual acuity as compared to the predicted corneal a cuity available in the EyeSys System 2000.