Performance metrics for models designed to predict individualized treatment effect

Objective Measuring the performance of models designed to predict individualized treatment effect is challenging, because the outcomes of two alternative treatments are inherently unobservable in one patient. The C-for-benefit was proposed to measure discriminative ability. We aimed to propose metrics of calibration and overall performance for models predicting treatment effect. Study Design and Setting Similar to the previously proposed C-for-benefit, we defined the observed treatment effect as the difference between outcomes in pairs of matched patients. Thus, we redefined the E-statistics, the logistic loss and the Brier score into metrics for measuring a model's ability to predict treatment effect. In a simulation study, the metric values of deliberately perturbed models were compared to those of the data generating model. To illustrate the performance metrics, different models predicting treatment effect were applied to the data of the Diabetes Prevention Program. Results As desired, performance metric values of perturbed models were consistently worse than those of the optimal model (Eavg-for-benefit[≥]0.070 versus 0.001, E90-for-benefit[≥]0.115 versus 0.002, log-loss-for-benefit[≥]0.757 versus 0.733, Brier-for-benefit[≥]0.215 versus 0.212). Calibration, discriminative ability, and overall performance of three different models were similar in the case study. Conclusion The proposed metrics are useful to assess the calibration and overall performance of models predicting individualized treatment effect, accessible via (https://github.com/CHMMaas/HTEPredictionMetrics).