Adaptive treatment strategies (also referred to as dynamic treatment regimes and sequential multiple assignment randomized trials (SMART)) are powerful designs that provide ways of operationalizing clinical regimes. These designs are effective because they mimic clinical practice, allow one to tailor treatments to individuals over time, and allow one to answer questions such as what is the best sequence of treatment. These designs consist of multiple stages of randomization. For example, a two-stage adaptive treatment strategy has first stage random treatment assignment, and then second stage randomization depending on observed individual response after the first stage of randomization. These trials can be designed with little or no additional burden to investigators (for example, consent can be obtained for the entire trial; randomization assignments can be determined at the beginning of the trial -- i.e., flip a coin 3 times at the start of the trial, once for initial treatment assignment, once for second stage randomization to be used if respond, once for second stage randomization to be used if do not respond; and sample sizes to obtain adequate power for testing primary hypotheses need not be any larger than those of a traditional trial).

Tatsuki could probably add some things here.


S.A. Murphy (2003), Optimal Dynamic Treatment Regimes, Journal of the Royal Statistical Society, Series B (with discussion) 65(2), 331-366

S.A. Murphy (2005), An Experimental Design for the Development of Adaptive Treatment Strategies., Statistics in Medicine. 24:1455-1481.
Topic revision: r1 - 12 Jan 2006, BryanShepherd

This site is powered by FoswikiCopyright © 2013-2017 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback