Evaluation of Methods to Improve Accuracy of IOL Calculations in Long Eyes
Narrative Responses:
Purpose
In long axial length (AL) eyes, current unmodified intraocular lens (IOL) calculation formulas frequently result in post-operative hyperopia. Optimization of the AL has been proposed to minimize this calculation error. We evaluate the accuracy of 3 formulas after optimization and 2 formulas intended for use without optimization.
Methods
Consecutive cases with an AL of > 25mm from April 2010 - December 2012 were analyzed; 162 eyes of 115 patients met inclusion criteria. The refractive prediction error (RPE) for each eye was determined by comparing the actual postoperative refraction to the predicted postoperative refraction for the implanted IOL. The predicted refraction was back calculated using both the actual and the optimized AL with the Holladay 1, SRK/T, and Haigis formulas; the RPE for the Barrett Universal formula was also calculated. Similar calculations will also be performed with the PhacoOptics IOL program. P < 0.05 was considered statistically significant.
Results
AL optimization resulted in improved RPE with all 3 formulas evaluated, reducing the mean RPE for the Holladay 1 from 0.48 D to 0.04 D, for the SRK/T from 0.31 D to 0.16 D, and for the Haigis from 0.61 D to -0.01 D (all P < 0.001). The Barrett Universal formula had a mean RPE of 0.28D, which was better than the errors for the Holladay 1 and Haigis formulas prior to AL optimization, but with a larger error than all the formulas after AL optimization (all P<0.001). Results for the PhacoOptics IOL program will be presented.
Conclusion
AL optimization for all the tested formulas resulted in a more accurate refractive prediction in eyes with AL > 25 mm; the refractive prediction was more accurate with the optimized AL with all 3 formulas than with the Barrett Universal formula.