Abstract
PURPOSE: To investigate the influence of posterior corneal astigmatism on the prediction accuracy of toric multifocal intraocular lens (IOL) calculation. METHODS: The keratometric astigmatism measured by Lenstar LS 900 (KCA(L)), keratometric astigmatism (KCA(P)) and total corneal astigmatism (TCA) measured by Scheimpflug camera (Pentacam HR) were documented and analyzed accordingly. Three deduction models using different parameters were compared. Model 1: KCA(L) + keratometric corneal surgically induced astigmatism (KCSIA, 0.30 D @ 50°); Model 2: KCA(P) + KCSIA); Model 3: TCA + total CSIA (TCSIA, 0.23 D @ 50°). The prediction errors of each model as the difference vector between the actual and the intended residual astigmatism were compared. RESULTS: Seventy-six eyes implanted with toric multifocal IOLs were included in this study. The vector differences of the actual KCSIA and TCSIA were statistically significant in the total sample and against-the-rule (ATR) subgroup (both P < 0.05). Model 1 deduced the smallest mean values of prediction error, while that of Model 3 were smaller than that of Model 2, both in the total sample and the ATR subgroups (all P < 0.05). Meanwhile, in the total sample and ATR subgroups, the centroid vector magnitudes of Model 3 were smaller than that of Model 1 (0.31 ± 0.76 D and 0.39 ± 0.76 D). CONCLUSIONS: The calculation of toric multifocal IOL should be individualized especially in the ATR eyes for the impact of PCA on the estimation of the preoperative corneal astigmatism and the CSIA.