Comparison of the SRK/T formula and other theoretical and regression formulas

We compared the predictive accuracy of the SRK/T formula to the SRK II, Binkhorst II, Hoffer, and Holladay formulas in seven series of cases totaling 1,050 eyes. In the combined group, the SRK/T and Holladay formulas performed only slightly better than the other formulas. In short eyes (less than 22 mm), all formulas performed well, with the SRK/T, SRK II, and Holladay formulas performing marginally better. In moderately long eyes (greater than 24.5 mm, less than or equal to 27 mm), the Hoffer and Binkhorst II formulas had a greater proportion of cases with greater than 2 diopters (D) of error and the SRK/T and Holladay were again marginally better. In the very long eyes (greater than 27 mm and less than or equal to 28.4 mm), there were only 11 cases and all formulas performed well since none had greater than 2 D of prediction error. In an extremely long eye data set (greater than 28.4 mm), the SRK II formula clearly gave the poorest result. Eyes of this length occurred in only 0.1% of cases in our unselected series. Results support the contention that the present second and third generation IOL power formulas give fairly equivalent accuracy. Other factors, such as availability, ease of use, and ability to tailor or individualize, become major considerations.