Scroll to top

On MOST, SMART and Factorial Designs: New Experimental Approaches for Optimizing Interventions

October 30, 2013

Inbal (Billie) Nahum-Shani, PhD
Research Assistant Professor
Institute for Social Research, University of Michigan

About the Presentation: Behavioral and health scientists have been increasingly interested in building and evaluating interventions. These interventions often include multiple components (i.e., multiple aspects of the intervention program itself as well as aspects of the program delivery or the implementation) that may be interdependent in their functions and effects. The traditional approach to intervention development involves constructing an intervention a priori and then evaluating it in a standard randomized controlled trial (RCT) with one treatment group and one control group. This approach provides vital information about the efficacy of the overall intervention as a package. However, it provides little information about (a) the efficacy of individual intervention components (e.g., which components of the intervention are contributing to the overall effect); and (b) effective ways to sequence and adapt the intervention components over time (e.g., how to best tailor the intervention components to the changing needs of the participant). These questions are critical for designing new treatment packages, or optimizing existing ones for effectiveness or cost-effectiveness. Factorial designs and SMART designs (Sequential Multiple Assignment Randomized Trial) are two experimental approaches useful for addressing these critical questions. Factorial designs can aid investigators to determine which of several possible components of a proposed intervention have effects of practical significance. The SMART experimental approach was specifically developed to enable investigators obtain data that informs the construction of high-quality adaptive interventions (in which intervention options are adapted over time in response to the changing needs and ongoing performance of the participant). I will present examples of completed or ongoing studies in which Factorial and SMART designs were used, and discuss their potential for building efficacious behavioral interventions.

About the Presenter: Inbal (Billie) Nahum-Shani is a Research Assistant Professor in the Survey Research Center of the Institute for Social Research, at the University of Michigan. Her research focuses on developing and employing behavioral theory and novel methodology to construct adaptive interventions, namely interventions that modify the type, timing, dose, or delivery mode of support in order to address the unique and changing needs of individuals. A more recent focus of her research is on the use of mobile technologies to facilitate the timely delivery and adaptation of interventions, known as Just-In-Time Adaptive Interventions (JITAIs). Dr. Nahum-Shani is providing leadership for two, five year, federally funded research projects; one seeks to develop new data analytic methods to inform the development of adaptive interventions (funded by NIH/NIDA); and the other seeks to optimize a stepped-care adaptive intervention that integrates mobile-health components in the treatment of obese adults (funded by NIH/NIDDK). As a member of MD2K, one of 11 national Big Data Centers of Excellence awarded by the NIH, Dr. Nahum-Shani works on investigating whether sensor-enabled measures of stress can be used to trigger and adapt the timely delivery of stress-regulation support via a mobile device. Dr. Nahum-Shani’s work also contributes to several projects aiming to advance the health of young adults. This includes being a site- PI of a four site longitudinal study (funded by NIH/NIAAA) seeking to inform the development of an adaptive intervention that targets at-risk drinking in young adults transitioning from college to work; as well as collaborating on the Mental Health for College Students (eBridge) project, and the Healthy Minds Network for Research on Adolescent and Young Adult Mental Health.