Predictive accuracy of patient satisfaction in aphakic correction surgery using the random forest algorithm

Abstract

This study evaluates the efficacy of a random forest model in predicting patient satisfaction following ophthalmic surgery for the correction of aphakia. Utilizing preoperative and postoperative clinical metrics, the model was trained to identify patterns associated with patient outcomes. In validation, it demonstrated robust performance on a designated test set, achieving an accuracy of 94%, sensitivity of 92%, specificity of 91%, and an area under the receiver operating characteristic curve of 0.95. Stability and reliability were further confirmed through 5-fold cross-validation, consistently showing an average accuracy of 93.5%. Feature importance analysis identified axial length and best corrected visual acuity as key predictors. These findings establish the model‘s potential as a reliable tool for healthcare providers to predict patient satisfaction in postoperative settings, enhancing clinical decision-making in the treatment of aphakia.