# Comparison of Refractive Prediction Errors of Toric Intraocular Lens Formulae

## TL;DR

This study compares two formulas for predicting vision correction after toric intraocular lens implantation and finds both are highly accurate, with one slightly better for astigmatism.

## Contribution

The study provides a direct comparison of two toric IOL formulae's accuracy in predicting refractive outcomes after cataract surgery.

## Key findings

- Both BTT and HT formulae had similar median absolute errors in spherical equivalence prediction.
- HT formula showed lower mean absolute and centroid errors for astigmatism prediction compared to BTT.
- The difference in astigmatism prediction was statistically significant but not clinically meaningful.

## Abstract

Introduction

Toric intraocular lenses (IOLs) are effective in correcting astigmatism; however, with multiple IOL toric formulae available, identifying the most accurate formula is important. The IOLMaster 700 (Carl Zeiss Meditec AG, Jena, Germany) provides two in-built toric IOL formulae, the Barrett TK Toric (BTT) and Haigis Toric (HT), which sometimes differ in astigmatism predictions for a particular IOL power. This study investigates the accuracy of these formulae in predicting postoperative refraction.

Methods

This retrospective study included 52 eyes from 52 patients undergoing uncomplicated cataract surgery with AcrySof IQ Toric IOL implantation (Alcon Laboratories Inc., Fort Worth, Texas, United States). Biometric data, including axial length, anterior chamber depth, lens thickness, and total keratometry, were measured using the IOLMaster 700. Predicted residual spherical equivalence (SE) and astigmatism were calculated using BTT and HT formulae. Postoperative refractive outcomes were assessed at least four weeks post-surgery. Differences in prediction errors in SE and astigmatism were assessed.

Results

The median absolute error in SE prediction was comparable between BTT and HT formulae (0.195 D vs. 0.185 D; p=1.000), with no significant difference in the proportion of eyes achieving absolute prediction errors ≤0.25 D, ≤0.50 D, ≤0.75 D, or ≤1.00 D. For astigmatism, the HT formula demonstrated lower mean absolute errors (MAE: 0.41±0.33 D) and centroid errors (0.18 D @ 70°) compared to BTT (MAE: 0.47±0.36 D; centroid error: 0.32 D @ 76°; p<0.001).

Conclusion

Both BTT and HT formulae exhibit similarly high accuracy in predicting postoperative SE. HT performed statistically better for astigmatism prediction; however, the difference was negligible and not clinically significant.