Combining Aggregate Data and Individual Patient Data in Model-Based Meta-Analysis: An Illustrative Case Study of Tofacitinib in Rheumatoid Arthritis Patients

This article examines how different types of clinical trial data influence the performance of model-based meta-analysis (MBMA), a quantitative approach that integrates results across multiple randomized controlled trials to compare treatment effects. Using tofacitinib in rheumatoid arthritis as a case study, the authors evaluate the relative value of incorporating individual patient data (IPD) versus relying on aggregate data (AD) alone when performing MBMA. Aggregate trial data were sourced from the CODEX RA clinical outcomes database, which compiles systematically extracted results from the published literature. Through a structured simulation-based framework, the study compares MBMA performance with and without IPD, both in the absence of predictive covariates and in scenarios where a clinically relevant predictive covariate (Asian race) is available and used for stratification. The findings demonstrate that access to IPD alone provides limited benefit for MBMA unless it enables covariate stratification, either through IPD-derived stratified results or published stratified AD. Overall, the work highlights the central role of stratified covariate data—rather than raw IPD per se—in improving MBMA performance, supporting more robust evidence generation for drug development and strategic decision-making.