Reporting and Design Considerations for SMART Behavioral Science Research

Sequential Multiple Assignment Randomized Trials (SMARTs) are increasing in popularity in behavior science research with over 130 federally funded studies currently active. SMARTs use multiple randomizations to experimentally evaluate the impact of different decisions during an ongoing adaptive intervention. While many aspects of SMARTs are similar to more traditional single-randomization trials, the additional complexity and variability introduced during subsequent randomization requires specificity in reporting. Due to the increase in the application and reporting of SMARTs, this paper serves as a guide to support transparent reporting of adaptive intervention research. Further, these guidelines provide detail to ensure that future meta-analyses examining SMARTs can be robust. We provide recommendations for reporting and how reporting in SMARTs warrant additional detail above and beyond a traditional randomized trial.