Instrumental variables estimation with competing risk data.

Area of Science:

• Biostatistics
• Epidemiology
• Health Services Research

Background:

• Time-to-event analyses frequently encounter challenges with unmeasured confounding and competing risks.
• Instrumental variables (IVs) are increasingly utilized for robust effect estimation in observational studies.

Purpose of the Study:

• To develop and validate a novel methodology for incorporating instrumental variables into competing risk analyses.
• To estimate the effect of arbitrary exposures on cause-specific hazard functions under a semi-parametric model.

Main Methods:

• The proposed approach employs a semi-parametric model with minimal data distribution restrictions.
• It utilizes flexible, closed-form estimators that can be computed recursively.
• The method accommodates various exposure and IV types and allows for covariate adjustment.

Main Results:

• Simulation studies demonstrated the favorable performance of the proposed method.
• The approach effectively addresses both confounding and competing risks in time-to-event data.
• The methodology proved effective in a real-world application concerning mammography screening and breast cancer mortality.

Conclusions:

• The developed instrumental variable approach offers a powerful tool for analyzing time-to-event data with competing risks and confounding.
• This flexible methodology enhances the reliability of causal inference in complex epidemiological and clinical settings.
• It provides a valuable framework for evaluating interventions and exposures in health research.