Theoretical vs empirical prediction of aphakic refraction

The aphakic refraction was predicted from preoperative and postoperative keratometry and axial length determinations in 43 patients subjected to cataract extraction without intraocular lens implantation using a theoretical and an empirical (Sanders, Retzlaff, Kraff [SRK]) method of calculation. The error (mean +/- SD; observed value minus expected value) of the theoretical vs the empirical method was 0.02 +/- 0.76 diopters (D) vs 0.66 +/- 0.72 D, respectively, when the preoperative data were used in the calculations, and 0.21 +/- 0.80 D vs 0.84 +/- 0.79 D, respectively, when the postoperative data were used in the calculations. The mean error was significantly different from zero (predicted values were lower than observed values) for the SRK method but not for the theoretical method. The mean error of the SRK method could be eliminated by adjusting the offset constant in the regression formula.