Familywise error rate control for block response-adaptive randomization

TL;DR

This paper improves a method for controlling the familywise error rate in block response-adaptive randomization, simplifying the approach and enhancing power while ensuring error control in clinical trials.

Contribution

It introduces a simpler, more robust method for familywise error rate control in block response-adaptive designs with fixed control allocation, offering practical power benefits.

Findings

Guarantees no negative weights for data blocks

Provides a substantial power advantage in practice

Ensures familywise error rate control

Abstract

Response-adaptive randomization allows the probabilities of allocating patients to treatments in a clinical trial to change based on the previously observed response data, in order to achieve different experimental goals. One concern over the use of such designs in practice, particularly from a regulatory viewpoint, is controlling the type I error rate. To address this, Robertson and Wason (Biometrics, 2019) proposed methodology that guarantees familywise error rate control for a large class of response-adaptive designs. In this paper, we propose an improvement of their proposal that is conceptually simpler, in the specific context of block-randomised trials with a fixed allocation to the control arm. We show the modified method guarantees that there will never be negative weights for blocks of data, and can also provide a substantial power advantage in practice.