A ray tracing approach to calculate toric intraocular lenses.

PURPOSE: To quantify the precision of astigmatic correction in routine cataract surgery with toric intraocular lenses (IOLs) and to evaluate the predictability of keratometric and anterior/posterior topographic measurement for the improvement of the overall accuracy. METHODS: Seventy-eight eyes of 56 patients were implanted with toric IOLs. Data acquired by the Lenstar optical biometer (Haag-Streit, Bern, Switzerland) and TMS5 topography (Tomey, Nagoya, Japan) were processed with the ray tracing software Okulix (Tedics, Dortmund, Germany) to predict the residual refraction. Four different inputs were examined: keratometry only, anterior topography, anterior and posterior topography/tomography, and combination of keratometry only and anterior and posterior topography/tomography. Four weeks postoperatively, the spherical prediction error and the cylindrical prediction error (difference vector between predicted and achieved cylindrical refraction) were determined. RESULTS: Mean absolute error of spherical prediction error was 0.27 diopter (D). Cylindrical prediction errors were 0.57 D (keratometry only), 0.56 D (anterior topography), 0.56 D (anterior and posterior topography/tomography), and 0.50 D (combination of keratometry only and anterior and posterior topography/tomography). Differences between intraocular lens groups were statistically significant (Friedman test, P < .05). CONCLUSION: The combination of keratometry and anterior and posterior topography/tomography of anterior and posterior surface yielded the best results for toric IOL power calculations.