Restricted mean survival time regression model with time-dependent covariates

Area of Science:

• Biostatistics
• Clinical Epidemiology
• Survival Analysis

Background:

• Restricted mean survival time (RMST) is an interpretable survival analysis metric, advantageous over traditional hazard rate methods.
• Existing RMST regression models primarily handle baseline covariates, limiting their application with time-dependent factors.

Purpose of the Study:

• To develop a novel regression model for RMST that accommodates time-dependent covariates.
• To evaluate the performance and predictive accuracy of the new model.

Main Methods:

• Developed a time-dependent RMST regression model utilizing the inverse probability of censoring weighting (IPCW) method.
• Validated the model's parameter estimation through Monte Carlo simulations.
• Compared the model's predictive performance against the time-dependent Cox model and a fixed-covariate RMST model.

Main Results:

• The proposed time-dependent RMST regression model accurately estimates regression parameters.
• The model demonstrated superior prediction ability compared to the time-dependent Cox model and the fixed-covariate RMST model.
• The model's utility was confirmed using a real-world heart transplantation dataset.

Conclusions:

• The developed time-dependent RMST regression model offers a robust approach for analyzing survival data with time-varying covariates.
• This method enhances prediction accuracy in clinical and epidemiological follow-up studies.
• The model provides a valuable tool for researchers dealing with complex longitudinal data.