Minimizing selection bias in randomized trials: A Nash equilibrium approach to optimal randomization |
| |
| Affiliation: | 2. Department of Public Health, Grand Valley State University, Allendale, Michigan;3. Department of Epidemiology, University of Michigan, Ann Arbor, Michigan;1. Service d''Anesthésie Réanimation, CHU de Dijon, Dijon, France, BP 77908, 21709 Dijon Cedex, France;2. Service d''Epidémiologie et d''Hygiène Hospitalières, CHU de Dijon, Dijon, France, BP 77908, 21709 Dijon Cedex, France;3. Service d''imagerie diagnostique et thérapeutique Neuroradiologie et Urgences CHU de Dijon, Dijon, France, BP 77908, 21709 Dijon Cedex, France;2. Department of Neurology, Botucatu Medical School-UNESP, São Paulo State University, São Paulo, Brazil;3. Department of Tropical Diseases and Diagnostic Imaging, Botucatu Medical School-UNESP, São Paulo State University, São Paulo, Brazil;4. Department of Neuroscience and Behavior, Ribeirão Preto Medical School-USP, São Paulo State University, São Paulo, Brazil;2. Monash University, Commercial Road, Melbourne, Victoria, Australia;3. Ochsner Medical Center, New Orleans, LA;4. Missoula Anesthesiology and The International Heart Institute of Montana, Missoula, MT;5. Cleveland Clinic, Cleveland, OH;11. University Hospital Münster, Münster, Germany;12. University of Alberta, Edmonton, AB, Canada |
| |
| Abstract: | Randomized trials can be compromised by selection bias, particularly when enrollment is sequential and previous assignments are unmasked. In such contexts, an appropriate randomization procedure minimizes selection bias while satisfying the need for treatment balance. This paper presents optimal randomization mechanisms based on non-cooperative game theory and the statistics of selection bias. For several different clinical trial examples, we examine subgame-perfect Nash equilibrium, which dictates a probability distribution on suitable assignment sequences. We find that optimal procedures do not involve discrete uniform distributions, because minimizing predictability is not equivalent to minimizing selection bias. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
