Covariate-Adaptive Optimization in Online Clinical Trials
The decision of how to allocate subjects to treatment groups is of great importance in experimental clinical trials for novel investigational drugs, a multibillion-dollar industry. Statistical power, the ability of an experiment to detect a positive treatment effect when one exists, depends in part...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Institute for Operations Research and the Management Sciences (INFORMS),
2021-02-18T15:24:43Z.
|
| Subjects: | |
| Online Access: | Get fulltext |
| Summary: | The decision of how to allocate subjects to treatment groups is of great importance in experimental clinical trials for novel investigational drugs, a multibillion-dollar industry. Statistical power, the ability of an experiment to detect a positive treatment effect when one exists, depends in part on the similarity of the groups in terms of measurable covariates that affect the treatment response. We present a novel algorithm for online allocation that leverages robust mixed-integer optimization. In all tested scenarios, the proposed method yields statistical power at least as high as, and sometimes significantly higher than, state-of-the-art covariate-adaptive randomization approaches. We present a setting in which our algorithm achieves a desired level of power at a sample size 25%-. smaller than that required with randomization-based approaches. Correspondingly, we expect that our covariate-adaptive optimization approach could substantially reduce both the duration and operating costs of clinical trials in many commonly observed settings while maintaining computational efficiency and protection against experimental bias. United States. Office of Naval Research (Grant 021152-00001) |
|---|