ACCURACY USING DIFFERENT KERATOMETRIC PARAMETERS IN CATARACT SURGERY: A COMPARATIVE STUDY

Objective: To assess whether standard keratometry (K), total keratometry (TK), simulated keratometry (SimK), total corneal refractive power (TCRP), and 65% Holladay equivalent keratometry reading (EKR65) differ in postoperative refractive prediction accuracy.Materials and Methods: This retrospective methodological study evaluated 110 eyes of 110 patients after uncomplicated cataract surgery. Keratometry and TK were obtained from the IOLMaster 700, and SimK, TCRP, and EKR65 were obtained from the Pentacam AXL. Using the actual implanted intraocular lens power, postoperative refraction was recalculated with the Barrett Universal II formula for each keratometric parameter, while all other biometric inputs remained unchanged.Results: Flat, steep, and mean keratometry values differed significantly among the methods (all p<0.001), whereas corneal astigmatism did not (p=0.248). Total corneal refractive power yielded the lowest mean corneal power and SimK the highest. Prediction errors differed significantly (p<0.001): K 0.29±0.51 diopter (D), TK 0.35±0.51 D, SimK 0.39±0.53 D, and EKR65 0.25±0.56 D showed mean hyperopic errors. TCRP −0.18±0.53 D showed a mean myopic error. Absolute prediction errors also differed overall (p=0.008), with TCRP showing the lowest mean value; however, no pairwise comparison of absolute prediction error remained significant after Bonferroni correction. The proportions of eyes within ±0.25 D, ±0.50 D, and ±1.00 D of the predicted refraction were similar among the methods (all p>0.05).Conclusion: Keratometric parameters influenced the direction of refractive prediction error, but no method showed a clear advantage in overall accuracy. Although TCRP yielded the lowest numerical absolute prediction error, it showed no significant advantage in pairwise comparisons or clinical accuracy thresholds.