Improving IOL Power Selection in Second Eye Using First-Eye Results
Narrative Responses:
Purpose
To improve results in IOL power selection of the second eye with the results of the first eye.
Methods
Retrospective analysis of refraction results from over 1000 paired postoperative catarct surgery eyes from 5 different surgeons yielded two methods of prediction to improve results in the second eye surgery. The first is a linear regression for adjusting the IOL to be implanted: -0.07582 + PredErrorFirstEye * 0.71561; The second was a Monte Carlo Markov Chain Prediction Process using the data set for analysis. A prospective study of 50 second eyes was performed. The Second Eye IOL power was modified by the linear regression prediction or the MCMC prediction, the method randomly assigned.
Results
Both methods, Linear Regression and MCMC prediction, improved results in the second eye. The MCMC method variance, 0.21, was smaller than the Linear Regression of 0.34. This difference was significant (p<0.05)
Conclusion
A simple linear regression (AdjustIOLPower = -0.07582 + PredError * 0.71561) correction of the second eye based on the the Prediction Error found in first eye at one week postop and an MCMC prediction both improved results in the second eye. The MCMC method performed slightly better.