Authors: Soumya Sharma, Harrish Nithianandan, Eric S. Tam, Hannah Chiu, Raj Maini, Sohel Somani

Purpose: The purpose of this study was to compare the performance and accuracy of two intraocular lens (IOL) power calculation formulae, Haigis and Barrett Universal II, on a Tecnis IOL platform in patients who have undergone either Manual Cataract Surgery (MCS) or Refractive Laser-assisted Cataract surgery (ReLACS). 

Study Design: Retrospective EMR chart review. 

Methods: Included patients who had pre-operative biometric testing done with IOL Master 700 using both Haigis and Barrett formulas and implantation of a Tecnis IOL. Predicted spherical equivalence (SE) from each formula was compared to the post-operative SE taken at one month. The Mean Absolute Error (MAE) and Median Absolute Error (MedAE) were compared for both formulae. 

Results: Of the total 158 eyes studied, 64 eyes underwent MCS and 94 eyes underwent ReLACS. The overall MAE was statistically significantly less using the Barrett (0.29±0.25D) versus Haigis (0.36±0.30D) (p<0.002). The MAE in the MCS group showed a trend towards less error with the Barrett (0.27±0.21D) versus the Haigis (0.35±0.26D) (P=0.06). The MAE in the ReLACS group was statistically significantly less using the Barrett (0.29±0.27D) versus Haigis (0.37±0.33D) (P<0.016). The MedAE in the overall, MCS and ReLACS groups using the Barrett formula was 0.24D, 0.25D, and 0.22D respectively; and for the Haigis formula was 0.27D, 0.26D and 0.28D respectively. The expected differences (ED) between Barrett and the fourth generation Haigis formula were significantly correlated with axial length (r=0.169, p= 0.017), corneal curvature (r = 0.227, p = 0.002), and lens thickness (r = 0.354, p = 0.000). 

Conclusions: The Barrett formula was found to be statistically more accurate than the Haigis formula in all eyes, and specifically in eyes undergoing ReLACS. This trend was also observed in patients undergoing MCS, albeit not statistically significant.