Descemet stripping automated endothelial keratoplasty (DSAEK) is a predominant surgical method of endothelial keratoplasty for treating corneal endothelial dysfunction. 1 Prediction, variable selection, and determining factors associated with graft failure may lead to improved clinical decisionmaking guidelines and surgical outcomes of DSAEK. Survival data allow for time-to-event analysis, which can be defined as death, onset of disease, or the success or failure of procedures. The traditional statistical methods used to analyze survival data include Kaplan-Meier analysis for estimating the survival function, the log-rank test for 2-group comparisons, and the Cox proportional hazards regression model to determine associations between risk factors and the probability of survival. 2 More recently, machine learning techniques have been used to expand our ability in including high-dimensional data with multiple features for predictions and variable selection. In this issue of JAMA Ophthalmology, O'Brien et al 3 use data from the Cornea Preservation Time Study to apply a random survival forest (RSF), an ensemble tree method for analysis of right-censored survival data, as a method to select important variables that predict graft failure after DSAEK.For censored survival data, Cox proportional hazards regression models are known for their ease of use and interpretability in determining associations between factors, such as sociodemographic, clinical, and environmental factors, and time to event. However, the proportional hazards assumption can restrict the application of this model. Cox proportional hazards regression and logistic regression were used in a previous study to determine factors associated with DSAEK graft success using the same data from the Cornea Preservation Time Study. 4 Donor diabetes status and intraoperative complications were associated with graft success in the early and entire postoperative period, whereas preoperative diagnosis (Fuchs dystrophy compared with pseudophakic or aphakic corneal edema) was associated with long-term postoperative graft success. What differentiates this study 3 is the use of a machine learning model with survival statistics, an RSF, to incorporate 67 predictive donor, recipient, eye bank, and intraoperative variables and rank their importance in predicting graft failure. Survival trees and RSF models are approaches for prediction and feature selection when the proportional hazards assumption is violated and when there is a need to include multiple variables for exploratory analysis. The notable finding in the study by O'Brien et al 3 is that, with the RSF, intraoperative complications were ranked as one of the most important predictors of graft failure after DSAEK. Specifically, the authors were further able to identify that donor lenticule flipping on insertion and difficulty unfolding, as well as positioning donor tissue without use of the positioning hook, were the most informative variables for predicting graft failure.