Comparison of the Cox Proportional Hazards Model and Random Survival Forest Algorithm for Predicting Patient‐Specific Survival Probabilities in Clinical Trial Data
- 1Institute of Mathematics University of Augsburg Augsburg Germany
- 2Department of Mathematics and Statistics University of Reading Reading UK
Abstract
The Cox proportional hazards model is often used to analyze data from randomized controlled trials (RCTs) with time‐to‐event outcomes. Random survival forest (RSF) is a machine‐learning algorithm known for its high predictive performance. We conduct a comprehensive neutral comparison study to compare the performance of Cox regression and RSF in various simulation scenarios based on two reference datasets from RCTs. The motivation is to identify settings in which one method is preferable over the other when comparing different aspects of performance using measures according to the TRIPOD (Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis) recommendations. Our results show that conclusions solely based on the index, a performance measure that has been predominantly used in previous studies comparing predictive accuracy of the Cox‐PH and RSF model based on real‐world observational time‐to‐event data and that has been criticized by methodologists, may not be generalizable to other aspects of predictive performance. We found that measures of overall performance may generally give more reasonable results, and that the standard log‐rank splitting rule used for the RSF may be outperformed by alternative splitting rules, in particular in nonproportional hazards settings. In our simulations, the performance of the RSF suffers less in data with treatment–covariate interactions compared to data where these are absent. The performance of the Cox‐PH model is affected by the violation of the proportional hazards assumption.
Related
- Interval-Censored Survival Analysis: A Comparative Monte Carlo Study of Weibull AFT and Cox Proportional Hazards Models
- Critical branching random walks on ℤ2: local survival probabilities and Yaglom limit theorems
- Sharp Partial Identification for Survival Model Comparison Without Target Outcomes
- Traveling Salesman Problem on Graphs with Proportional Independent Set
- Fluid and Diffusion Approximation of a Multiclass Proportional Processor Sharing Queue with Reneging and General Distributions
BibTeX
@article{graf2026comparison,
title = {{Comparison of the Cox Proportional Hazards Model and Random Survival Forest Algorithm for Predicting Patient‐Specific Survival Probabilities in Clinical Trial Data}},
author = {Ricarda Graf and Susan Todd and M. Fazil Baksh},
journal = {Biometrical Journal},
year = {2026},
month = sep,
doi = {10.1002/bimj.70171},
url = {https://doi.org/10.1002/bimj.70171},
note = {CC BY},
}
RIS
TY - JOUR
TI - Comparison of the Cox Proportional Hazards Model and Random Survival Forest Algorithm for Predicting Patient‐Specific Survival Probabilities in Clinical Trial Data
AU - Ricarda Graf
AU - Susan Todd
AU - M. Fazil Baksh
PY - 2026
DA - 2026/09/25
JO - Biometrical Journal
DO - 10.1002/bimj.70171
UR - https://doi.org/10.1002/bimj.70171
AB - The Cox proportional hazards model is often used to analyze data from randomized controlled trials (RCTs) with time‐to‐event outcomes. Random survival forest (RSF) is a machine‐learning algorithm known for its high predictive performance. We conduct a comprehensive neutral comparison study to compare the performance of Cox regression and RSF in various simulation scenarios based on two reference datasets from RCTs. The motivation is to identify settings in which one method is preferable over the other when comparing different aspects of performance using measures according to the TRIPOD (Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis) recommendations. Our results show that conclusions solely based on the index, a performance measure that has been predominantly used in previous studies comparing predictive accuracy of the Cox‐PH and RSF model based on real‐world observational time‐to‐event data and that has been criticized by methodologists, may not be generalizable to other aspects of predictive performance. We found that measures of overall performance may generally give more reasonable results, and that the standard log‐rank splitting rule used for the RSF may be outperformed by alternative splitting rules, in particular in nonproportional hazards settings. In our simulations, the performance of the RSF suffers less in data with treatment–covariate interactions compared to data where these are absent. The performance of the Cox‐PH model is affected by the violation of the proportional hazards assumption.
SN - 0323-3847
SN - 1521-4036
ER -