Selection processes, transportability, and failure time analysis in life history studies

Area of Science:

• Biostatistics
• Epidemiology
• Life History Analysis

Background:

• Cohort studies often select participants based on health events, potentially violating independence assumptions.
• Selection bias can compromise the internal and external validity of cohort study findings.
• Transportability of results to the general population is a critical concern in health research.

Purpose of the Study:

• To propose a joint model for cohort selection and failure processes in life history analysis.
• To address violations of independence assumptions common in cohort study enrollment.
• To investigate the conditions for transportability of cohort study results.

Main Methods:

• Developed a joint statistical model integrating cohort selection and failure time processes.
• Analyzed the impact of dependent selection on life history analysis.
• Utilized numerical studies to illustrate the framework and discuss auxiliary data requirements.

Main Results:

• Demonstrated that transportability of cohort study results is not guaranteed without auxiliary population data.
• Identified conditions leading to dependent selection and discussed relevant auxiliary data types.
• Applied the joint model to a real-world study on psoriatic arthritis risk in psoriasis patients.

Conclusions:

• The proposed joint modeling framework can mitigate bias arising from dependent cohort selection.
• Auxiliary population information is essential for ensuring the transportability of findings.
• The methodology provides a robust approach for analyzing life history data in health research.