To assess the impact of measurement error on cylinder treatment planning in corneal refractive surgery and to compare the reliability of two indices of cylinder change: the Correction Index, based on the surgically induced astigmatism, and the Flattening Index, based on the flattening effect.
Preoperative refractions and surgical outcomes were evaluated using a Monte Carlo simulation. Cylinder change was calculated using vector analysis. The initial distribution of cylinder magnitudes was matched to a population sample of 1,000 eyes prior to LASIK. Refraction measurement error was modeled both preoperatively and postoperatively, with a non-linear optimization solver optimizing treatment effect for each amount of preoperative cylinder. Model assumptions were subsequently tested against LASIK outcomes data.
The mean Flattening Index was approximately 1 for all levels of preoperative cylinder, whereas the Correction Index was significantly higher than 1 for cylinder treatments of 1.00 diopters (D) or less when preoperative and postoperative measurement errors were taken into consideration. Targeting undercorrection minimized postoperative cylinder compared with targeting full correction, with the optimal target dependent on the amount of random error in refraction measurement. It was optimal to partially treat 0.25 D of cylinder, even if the presumed level of measurement error was relatively high.
The Flattening Index outperforms the Correction Index when accounting for errors in preoperative and postoperative cylinder measurement, particularly when pre-operative cylinder is 1.00 D or less. Treatment nomograms should be adjusted to use the Flattening Index to assess cylinder change, and reporting guidelines should be updated to include the Flattening Index. [J Refract Surg. 2020;36(3):169-174.].

Copyright 2020, SLACK Incorporated.