Mon, April 10, 2017, noon:
Daniel Almirall (PSC, U of M)
02/13/2017, 12:00:00, 6050 ISR - Thompson
The effective treatment and management of a wide variety of health disorders often requires individualized, sequential decision-making. To do this, each patient's treatment is dynamically tailored over time based on the patient's history and changing disease state. Adaptive interventions (or, dynamic treatment regimens) operationalize such individualized decision making using a sequence of decision rules that specify which intervention option to offer, for whom, and when. Intervention options in this case correspond to varying doses, types or delivery modes of pharmacological, behavioral and/or psychosocial treatments. There has been a surge of clinical and methodological interest in developing and evaluating adaptive interventions via clinical trials. Specifically, there is great interest in the use of sequential multiple assignment randomized trials (SMART), a type of multi-stage randomized trial design, to build high-quality adaptive interventions. The aim of this talk is to provide a brief, conceptual introduction to adaptive interventions and SMART designs. To illustrate ideas, we draw on three example SMART studies: the first is a study to develop a clinic-based adaptive intervention to improve social communication skills in minimally verbal children with autism; the second is a study to develop a school-based adaptive intervention to improve academic outcomes in school-children with autism; the third is a cluster-randomized study to develop an adaptive implementation intervention to improve the uptake/adoption of an evidence-based intervention for mood disorders in community mental health settings. We end with a discussion of potential applications in health policy or population health.
[The talk is based on joint work with Inbal Nahum-Shani, Susan A. Murphy, Connie Kasari (UCLA), and Amy Kilbourne.]