R and Python codes for estimating heterogeneous survival treatment effect via Machine/Deep Learning methods in observational studies.

We considered 7 machine/deep learning methods designed for right-censored survival outcomes, and contributed a series of simulations to evaluate the operating characteristics of these methods for estimating the treatment effect heterogeneity. A case study examining the heterogeneous treatment effects of two radiotherapy approaches for treating high-risk prostate cancer demonstrates the methods.

Accompanying article is: Hu L, Ji J, Li F (2021). Estimating heterogeneous survival treatment effect in observational data using machine learning. Stat Med 40 (21), 4691-4713.