Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Consistency Assessment of Regional Treatment Effect for Multi-Regional Clinical Trials in the Presence of Covariate Shift

Kunhai Qing, Xinru Ren, Jin Xu, Menggang Yu

TL;DR

The paper tackles covariate shift in multi-regional clinical trials by moving beyond region-wise marginal treatment effects to conditional average treatment effects $Δ_r({\bm X})$. It introduces a two-step consistency assessment: a traditional one-step check using region-specific ATEs, followed by a salvage step that tests CATE similarity via ITE estimation and, when appropriate, a covariate-shift adjusted ATE $δ_r^*$. Through extensive simulations and an BELIEVE trial-inspired application, the authors show that the two-step method improves true positive consistency detection while maintaining control over false positives, particularly when effect modifiers differ in distribution across regions. The approach integrates density-ratio adjustments, LOOP-based ITE estimation, and RMST-based endpoints to handle censoring, offering a practical framework for regulatory-ready consistency assessment in MRCTs under covariate shift.

Abstract

Multi-Regional Clinical Trials (MRCTs) play a central role in the development of new therapies by enabling the simultaneous evaluation of drug efficacy and safety across diverse global populations. Assessing the consistency of treatment effects across regions is a fundamental aspect of MRCTs. Existing methods typically focus on region-specific marginal treatment effects. However, when treatment effect heterogeneity arises due to effect-modifying baseline covariates, distributional differences in these covariates can lead to erroneous conclusions. In this paper, we explicitly account for this phenomenon in the consistency assessment by considering the conditional average treatment effect. We propose a two-step assessment strategy that complements existing methods and mitigates the impact of treatment effect heterogeneity. Results from numerical studies demonstrate the effectiveness of the proposed approach.

Consistency Assessment of Regional Treatment Effect for Multi-Regional Clinical Trials in the Presence of Covariate Shift

TL;DR

The paper tackles covariate shift in multi-regional clinical trials by moving beyond region-wise marginal treatment effects to conditional average treatment effects . It introduces a two-step consistency assessment: a traditional one-step check using region-specific ATEs, followed by a salvage step that tests CATE similarity via ITE estimation and, when appropriate, a covariate-shift adjusted ATE . Through extensive simulations and an BELIEVE trial-inspired application, the authors show that the two-step method improves true positive consistency detection while maintaining control over false positives, particularly when effect modifiers differ in distribution across regions. The approach integrates density-ratio adjustments, LOOP-based ITE estimation, and RMST-based endpoints to handle censoring, offering a practical framework for regulatory-ready consistency assessment in MRCTs under covariate shift.

Abstract

Multi-Regional Clinical Trials (MRCTs) play a central role in the development of new therapies by enabling the simultaneous evaluation of drug efficacy and safety across diverse global populations. Assessing the consistency of treatment effects across regions is a fundamental aspect of MRCTs. Existing methods typically focus on region-specific marginal treatment effects. However, when treatment effect heterogeneity arises due to effect-modifying baseline covariates, distributional differences in these covariates can lead to erroneous conclusions. In this paper, we explicitly account for this phenomenon in the consistency assessment by considering the conditional average treatment effect. We propose a two-step assessment strategy that complements existing methods and mitigates the impact of treatment effect heterogeneity. Results from numerical studies demonstrate the effectiveness of the proposed approach.
Paper Structure (15 sections, 27 equations, 3 figures, 7 tables, 1 algorithm)

This paper contains 15 sections, 27 equations, 3 figures, 7 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. One-step consistency assessment
  3. Two-step consistency assessment
  4. Evaluating similarity between $\Delta_{r}({\bm X})$ and $\Delta_{-r}({\bm X})$
  5. Covariate-shift adjusted ATE
  6. Algorithm for two-step assessment and practical consideration
  7. Simulation
  8. Setup
  9. Results in the absence of covariate distributional shift
  10. Results in presence of covariate distributional shift
  11. Application to the BELIEVE trial
  12. Discussion
  13. Computation of pseudo-observations of RMST
  14. CATE functions for binary and survival endpoints
  15. Details of the setup of scenarios for Sections \ref{['sec:results-without-distribution-shift']} and \ref{['sec:results-with-distribution-shift']}

Figures (3)

  • Figure 1: Empirical CPs of region $r$ with respect to its complementary regions obtained by the one-step method and the proposed two-step method for various types of endpoints under different levels of discrepancy in regional CATEs (decreasing in $\kappa_r / \kappa_{-r}$) in the absence of covariate distributional shift
  • Figure 2: Empirical CPs of region $r$ with respect to its complementary regions obtained by the one-step method and the proposed two-step method under scenario (i), where a covariate shift takes place in $X_1$ alone, for various types of endpoints under different levels of discrepancy in regional CATEs (decreasing in $\kappa_r / \kappa_{-r}$)
  • Figure 3: Empirical CPs of region $r$ with respect to its complementary regions obtained by the one-step method and the proposed two-step method under scenario (ii), where a covariate shift takes place in both $X_1$ and $X_2$, for various types of endpoints under different levels of discrepancy in regional CATEs (decreasing in $\kappa_r / \kappa_{-r}$)