Number of Repetitions in Re-randomization Tests

TL;DR

This paper examines the optimal number of repetitions in re-randomization tests used in covariate- or response-adaptive randomization, proposing adaptive methods to reduce computational burden in clinical trial analyses.

Contribution

It introduces an adaptive procedure to determine the number of re-randomization repetitions, improving computational efficiency in statistical inference for adaptive randomization designs.

Findings

Proposed an adaptive method to reduce repetitions in re-randomization tests.

Demonstrated through simulations that the method maintains validity with fewer repetitions.

Provided practical strategies for implementing the approach in clinical trial settings.

Abstract

In covariate-adaptive or response-adaptive randomization, the treatment assignment and outcome can be correlated. Under this situation, re-randomization tests are a straightforward and attractive method to provide valid statistical inference. In this paper, we investigate the number of repetitions in the re-randomization tests. This is motivated by the group sequential design in clinical trials, where the nominal significance bound can be very small at an interim analysis. Accordingly, re-randomization tests lead to a very large number of required repetitions, which may be computationally intractable. To reduce the number of repetitions, we propose an adaptive procedure and compare it with multiple approaches under pre-defined criteria. Monte Carlo simulations are conducted to show the performance of different approaches in a limited sample size. We also suggest strategies to reduce total computation time and provide practical guidance in preparing, executing and reporting before and after data are unblinded at an interim analysis, so one can complete the computation within a reasonable time frame.