Robust and Data-Adaptive Integration of Nonconcurrent Data in Platform Trials via Gaussian Processes

2026-05-13

AI Summary

Platform trials often exclude non-concurrent randomized data due to concerns about time trends, resulting in efficiency loss. This work proposes an adaptive integration framework based on Gaussian processes that jointly models concurrent and non-concurrent data by leveraging temporal smoothness, thereby improving estimation efficiency while controlling bias. The method accommodates discrete outcomes and covariate adjustment. Theoretically, it establishes—for the first time—that incorporating non-concurrent controls reduces the posterior variance of treatment effects and provides a non-increasing upper bound on potential bias. A frequentist interpretation is offered via connections to kernel ridge regression. The approach is validated through simulations mimicking the SURMOUNT-1 platform trial, and an open-source R package, RobinCID, is provided to facilitate practical implementation.

Abstract

A platform trial is an innovative clinical trial design that enables simultaneous and continuous evaluation of multiple treatments within a single master protocol. Existing robust methods restrict analyses to concurrently randomized participants due to concerns that including nonconcurrent data may introduce bias from temporal trends. However, this exclusion represents a missed opportunity to improve efficiency. We propose a Gaussian process framework for incorporating nonconcurrent data that exploits temporal smoothness, a key feature of platform trials. The framework includes single-task and multi-task formulations and provides data-adaptive integration of nonconcurrent data with uncertainty quantification. The connection to kernel ridge regression yields a transparent frequentist interpretation of how nonconcurrent data are integrated. We establish two theoretical guarantees: incorporating nonconcurrent controls reduces the posterior variance of the treatment effect, and the resulting bias is controlled by a non-increasing bound. We extend the framework to discrete outcomes and to covariate adjustment, illustrate it on a hypothetical platform trial constructed from SURMOUNT-1, and provide an implementation in the R package RobinCID.

Problem

• platform trials
• nonconcurrent data
• temporal trends
• bias
• efficiency

Innovation

• Gaussian processes
• platform trials
• nonconcurrent controls
• data-adaptive integration
• temporal smoothness