Inference for restricted mean survival time as a function of restriction time under length-biased sampling

Area of Science:

• Biostatistics
• Survival Analysis
• Epidemiology

Background:

• Restricted mean survival time (RMST) is crucial for censored survival data in clinical studies.
• Length-biased sampling, common in observational studies, complicates RMST estimation for multiple time points.
• Existing methods struggle to estimate RMST across various restriction times within a single model under length-biased sampling.

Purpose of the Study:

• To develop a novel statistical model for estimating RMST as a continuous function of restriction time.
• To address the challenges posed by length-biased sampling in survival data analysis.
• To provide a unified approach for estimating time-varying covariate effects on RMST.

Main Methods:

• Proposed two novel methods based on estimating equations to model RMST under length-biased sampling.
• Developed estimators for time-varying covariate effects.
• Established asymptotic properties for the proposed statistical estimators.

Main Results:

• The developed methods effectively model RMST as a continuous function of restriction time.
• Simulation studies confirmed the good finite sample performance of the proposed estimators.
• The procedures were successfully applied to analyze two real-world datasets.

Conclusions:

• The proposed approach offers a robust method for estimating RMST in the presence of length-biased sampling.
• This work provides valuable tools for survival data analysis in observational studies.
• The findings contribute to a more accurate understanding of survival outcomes in complex sampling settings.