Assessing the Accuracy of Refractive Prediction of Different IOL Formulas in Medium Long Eyes

Authors: Tasneem Arsiwalla A* , Cenanning Li and Nishant Gupta

DOI: 10.23880/jcstm-16000125

Volume 7, Issue 1

Abstract

Purpose: To evaluate and compare the prediction errors (PE) between seventeen different IOL power calculation formulas in medium long eyes (>25mm). Design: Retrospective study. Setting: Public hospital in Victoria, Australia. Methods: Medical records of uncomplicated cataract surgery patients from 2008-2019 with axial length of >25mm were collected. The post-operative refraction for each patient was noted and the PE was calculated for each of the 17 different formulas. Mean, median, and absolute prediction errors (MPE, MedPE, APE respectively) and standard deviation (SD) was calculated. Results: 106 patients (106 eyes) were included in the study. Okulix showed the lowest MedAPE (0.224), and SRK-T had the highest (0.400). The SD for Okulix and Olsen PO formulas was the lowest. After reducing the mean to zero, the MedAPE of all formulas ranged from 0.155 to 0.379 and Okulix and Olsen PO had the lowest APE and SD (0.155±0.164 and 0.169±0.168 respectively). Furthermore, Okulix formula has the most patients with APE of within ± 0.25D (42%) and Olsen PO and Okulix have the least number of patients with APE of > ±1.0D (2%). Conclusion: Okulix and Olsen PO formulas have the MedPE closest to zero with smaller standard of deviation then the rest of the formulas. The Barrett U2, EVO, Haigis, Hill RBF 3.0, Holladay 2 NLR, Kane, K6, Okulix, Olsen Lenstar, Olsen PO, Pearl DGS, T2 formulas tend to have more myopic MPE.

Keywords: Prediction Errors; IOL Formulas; Myopic Eyes; Refractive Prediction

View PDF

View Appendix