A0473
Title: Calibrated mixtures of g-priors for assessing treatment effect moderation in Bayesian meta-analysis
Authors: Hwanhee Hong - Duke University (United States)
Qiao Wang - East Carolina University (United States) [presenting]
Abstract: Assessing treatment effect moderation is critical in biomedical research and many other fields, as it guides personalized intervention strategies to improve individual outcomes. Individual participant-level data meta-analysis (IPD-MA) offers a robust framework for such assessments by leveraging data from multiple studies. However, its performance is often hindered by challenges such as high between-study variability. Traditional Bayesian shrinkage methods are less suitable in MA, as their priors do not discern heterogeneous studies. The calibrated mixtures of g-priors are proposed in IPD-MA to enhance efficiency and reduce risks in estimating moderation effects, providing a novel series of priors tailored for multiple studies by incorporating a study-level calibration parameter and a moderator-level shrinkage. This design offers a flexible range of shrinkage levels, allowing practitioners to evaluate moderator importance from both conservative and optimistic perspectives. Compared with existing Bayesian shrinkage methods, the extensive simulation studies demonstrate that the calibrated mixtures of g-priors exhibit superior performances in terms of efficiency and risk metrics, particularly under high between-study variability, high model sparsity, weak moderation effects, and correlated design matrices. Their applications are illustrated in assessing the effect moderator of two active treatments for major depressive disorder, using IPD from randomized controlled trials.
