Target Aggregate Data Adjustment Method for Transportability Analysis Utilizing Summary-Level Data From the Target...

Area of Science:

• Causal inference
• Biostatistics
• Epidemiology

Background:

• Transportability analysis assesses external validity by applying study findings to new populations.
• Existing methods often require individual patient-level data (IPD), limiting use with aggregate data (AgD).
• Transportability methods for survival data with informative censoring using AgD are underexplored.

Purpose of the Study:

• Propose a novel framework, Target Aggregate Data Adjustment (TADA), for transportability analysis using AgD.
• Address challenges of informative censoring and distributional imbalances in effect modifiers.
• Enhance the applicability of transportability analysis in data-limited settings.

Main Methods:

• Developed a two-stage weighting framework (TADA).
• Incorporated time-varying inverse probability of censoring weights.
• Utilized method of moments for participation weights to adjust for effect modifier distributions.

Main Results:

• TADA effectively controls bias from moderate censoring.
• The method enhances transportability analysis with AgD.
• Demonstrated performance via extensive simulations and a real-world lung cancer trial case study.

Conclusions:

• TADA provides a viable approach for transportability analysis with AgD, even with informative censoring.
• The framework improves the clinical interpretability of findings in data-scarce scenarios.
• Expands the utility of causal inference methods in real-world applications.