Response-Adaptive Designs for Clinical Trials: Simultaneous Learning from Multiple Patients
University of Chicago
Date: January 20, 2014
Time: 02:30 pm - 04:00 pm
Location: Room 245
Clinical trials have traditionally followed a fixed design, in which patient allocation to treatments is fixed throughout the trial and specified in the protocol. The primary goal of this static design is to learn about the efficacy of treatments. Response-adaptive designs, on the other hand, allow clinicians to use the learning about treatment effectiveness to dynamically adjust patient allocation to treatments as the trial progresses. An ideal adaptive design is one where patients are treated as effectively as possible without sacrificing potential learning or compromising the integrity of the trial. We propose such a design, one that uses forward-looking algorithms to fully exploit learning from multiple patients simultaneously. Compared to the best existing implementable adaptive design (e.g. using Berry, 1978), we show that our proposed design improves patient outcomes by up to 8.6% under a set of considered scenarios. Further, we demonstrate our design's effectiveness using data from a recently conducted stent trial. The design is general and applicable to any Markov decision process setting where learning takes place from multiple simultaneous individual experiments as, for example, in customized consumer offers. This paper also adds to the general understanding of such models by showing the value and nature of improvements over heuristic solutions for problems with small delays in observing patient outcomes. We do this by showing the relative performance of these schemes for maximum expected health and maximum expected learning objectives, and by demonstrating the value of a truncated-horizon approximation in a practical example.