Repository logo
 

Multi-armed Bandit Models for the Optimal Design of Clinical Trials: Benefits and Challenges.

Accepted version
Peer-reviewed

Type

Article

Change log

Authors

Villar, Sofía S 
Bowden, Jack 

Abstract

Multi-armed bandit problems (MABPs) are a special type of optimal control problem well suited to model resource allocation under uncertainty in a wide variety of contexts. Since the first publication of the optimal solution of the classic MABP by a dynamic index rule, the bandit literature quickly diversified and emerged as an active research topic. Across this literature, the use of bandit models to optimally design clinical trials became a typical motivating application, yet little of the resulting theory has ever been used in the actual design and analysis of clinical trials. To this end, we review two MABP decision-theoretic approaches to the optimal allocation of treatments in a clinical trial: the infinite-horizon Bayesian Bernoulli MABP and the finite-horizon variant. These models possess distinct theoretical properties and lead to separate allocation rules in a clinical trial design context. We evaluate their performance compared to other allocation rules, including fixed randomization. Our results indicate that bandit approaches offer significant advantages, in terms of assigning more patients to better treatments, and severe limitations, in terms of their resulting statistical power. We propose a novel bandit-based patient allocation rule that overcomes the issue of low power, thus removing a potential barrier for their use in practice.

Description

Keywords

Gittins index, Multi-armed bandit, Whittle index, patient allocation, response adaptive procedures

Journal Title

Stat Sci

Conference Name

Journal ISSN

0883-4237
2168-8745

Volume Title

30

Publisher

Institute of Mathematical Statistics

Rights

All rights reserved
Sponsorship
Biometrika Trust (unknown)
Medical Research Council (MR/J004979/1)