Abstract

Purpose: To compare different IOL calculation formulas(Barrett II, Haigis, Hoffer Q, SRK/T & AI based Hill RBF, Kane, Ladas super) for eyes with mean K>46D with AL>26.00 mm; and for short eyes<22mm
Method: Retrospectively 46 eyes with mean K> 46 D were included. Average absolute value of differences between actual and predicted spherical equivalent,SE of postop ref error (mean absolute error: MAE) was calculated for each formula.
Similarly, for short eyes,29 eyes with AL<22mm were included who undergone similar process.
Results: For long eyes, MAE was smallest in Ladas super(0.28±0.32) followed by Barrett(0.39 D ± 0.35), Holladay II(0.74 D ± 0.32), Haigis (0.85 D ± 0.35), Hoffer Q (0.81 D ± 0.38)and SRK/T(0.93± 0.42).
Similarly, for short eyes, Ladas super again predicted more eyes within SE ± 0.5 D 86% & within 1D 97% compared to other formulas
Conclusion: In eyes with extremes of AL, Ladas super formula predicts more accurate results than others including Barrett universal II

Full Text