WOW: WAIC-Optimized Gating of Mixture Priors for External Data Borrowing

2025-10-06

AI Summary

In clinical trials, borrowing external data without prior compatibility assessment often induces estimation bias. To address this, we propose a WAIC-based gating strategy that constructs a prior-agnostic compatibility test using KL divergence and the widely applicable information criterion (WAIC), enabling *a priori* dynamic gating for mixed prior methods and preventing excessive borrowing from incompatible external data. The strategy integrates seamlessly into both fixed- and adaptive-weight frameworks without altering the underlying model architecture. Simulation studies demonstrate that, compared to rMAP and SAM, our approach significantly improves parameter estimation accuracy—reducing average MSE by 32%—enhances inferential robustness, and safeguards clinical decision reliability. Our key innovation lies in the first application of WAIC for *a priori* compatibility discrimination, thereby achieving *verifiable borrowing*.

Technology Category

• Machine Learning: Bayesian Learning
• Reasoning under Uncertainty: Relational Probabilistic Models
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

The integration of external data using Bayesian mixture priors has become a powerful approach in clinical trials, offering significant potential to improve trial efficiency. Despite their strengths in analytical tractability and practical flexibility, existing methods such as the robust meta-analytic-predictive (rMAP) and self-adapting mixture (SAM) often presume borrowing without rigorously assessing whether, how, or when integration is appropriate. When external and concurrent data are discordant, excessive borrowing can bias estimates and lead to misleading conclusions. To address this, we introduce WOW, a Kullback-Leibler-based gating strategy guided by the widely applicable information criterion (WAIC). WOW conducts a preliminary compatibility assessment between external and concurrent trial data and gates the level of borrowing accordingly. The approach is prior-agnostic and can be seamlessly integrated with any mixture prior method, whether using fixed or adaptive weighting schemes, after the WOW step. Simulation studies demonstrate that incorporating the WOW strategy before Bayesian mixture prior borrowing methods effectively mitigates excessive borrowing and improves estimation accuracy. By providing robust and reliable inference, WOW strengthens the performance of mixture-prior methods and supports better decision-making in clinical trials.

Problem

• Assesses compatibility between external and concurrent trial data
• Mitigates excessive borrowing bias in Bayesian mixture priors
• Improves estimation accuracy in clinical trial data integration

Innovation

• WOW uses WAIC for data compatibility assessment
• It gates borrowing levels to prevent excessive integration
• The method works with any Bayesian mixture prior