Proportion of treatment effect (PTE) explained by a surrogate marker

Area of Science:

• Biostatistics
• Clinical Trials
• Survival Analysis

Background:

• Estimating the proportion of treatment effect (PTE) explained by surrogate markers is crucial in clinical survival data analysis.
• Conventional methods involve fitting separate models, which complicates confidence interval construction and is computationally intensive.

Purpose of the Study:

• To propose a novel, computationally efficient procedure for estimating PTE and its confidence intervals in time-varying covariate analysis.
• To enhance the applicability of PTE estimation to multiple-covariate models and facilitate comparisons between surrogate markers.

Main Methods:

• A new procedure is introduced that simultaneously estimates treatment effects before and after covariate adjustment within a single statistical model.
• This approach contrasts with the conventional method of using two separate working models.

Main Results:

• The proposed procedure significantly simplifies computation compared to the conventional approach.
• It demonstrates enhanced flexibility, enabling effective application to multiple-covariate models for treatment effect decomposition and comparison of multiple surrogate markers.
• The method was successfully applied to data from the LIFE study.

Conclusions:

• The new simultaneous estimation procedure offers a more computationally efficient and flexible alternative for analyzing time-varying covariates in clinical survival data.
• This method overcomes limitations of conventional approaches, particularly in complex scenarios involving multiple surrogate markers and effect decomposition.