A methodology for analysing a repeated measures and survival outcome simultaneously

Area of Science:

• Biostatistics
• Biomedical Data Analysis
• Longitudinal Data Analysis

Background:

• Extensive literature exists for univariate repeated measures data analysis.
• Biomedical research frequently involves multiple outcome measures, often exceeding univariate analysis.
• Prior work addressed bivariate repeated measures with discrete or continuous outcomes using GEE and seemingly unrelated regression.

Purpose of the Study:

• To extend existing statistical frameworks for analyzing paired outcome data.
• To accommodate scenarios with one repeated measures outcome and one survival endpoint.
• To provide a robust methodology for complex biomedical research questions.

Main Methods:

• Utilized generalized estimating equations (GEE) to model the repeated measures outcome.
• Employed the seemingly unrelated regression (SUR) model to integrate GEE models.
• Extended the SUR-GEE framework to incorporate a survival endpoint alongside a repeated measures outcome.

Main Results:

• Developed a statistical methodology for analyzing mixed repeated measures and survival data.
• Addressed key estimation and hypothesis testing challenges for this combined data type.
• Demonstrated the practical application of the methodology through a relevant example.

Conclusions:

• The proposed method offers a unified approach for analyzing complex longitudinal and survival data in biomedical research.
• This extension enhances the analytical capabilities for studies with multiple, distinct endpoint types.
• The methodology provides a valuable tool for researchers dealing with mixed-effects data structures.