Session: Methodological Approaches in Implementation: Innovations in Research Design and Data Modeling (Society for Prevention Research 23rd Annual Meeting)

A driving force behind one of the priorities of the nascent field of implementation science is the gap between the development and establishment of evidenced-based prevention programs through scores of rigorous randomized control trials, and the successful implementation of these evidence-based prevention programs in communities. Addressing major roadblocks to closing this gap, we present:  1) a new federal effort to create a flexible infrastructure to share human subjects research data, and 2) two innovative new approaches for evaluating and monitoring the implementation of prevention programs as they are delivered into communities. We first present the National Institute of Mental Health data infrastructure initiative. Ultimately, this initiative will create a data repository of millions of deidentified human subjects records from NIMH funded trials along with the data dictionaries that accompany the records. We discuss the opportunities for modeling intervention effects utilizing this data repository. The second presentation uses an automated system for monitoring and providing feedback on the implementation process of a prevention program. This approach discusses how large amounts of data can be used in combination with calibration to provide high quality assessments for monitoring the implementation process of a given prevention program. Two examples are used to illustrate the potential of using this automated system in prevention research and practice. Blending methodology and frameworks from SMART trials, the third presentation examines how realistic intervention regimens can be tested and improved in ongoing studies, as programs are implemented in new settings. These SMART-AR (SMART with adaptive randomization) designs incorporate two features: 1) allowing for second stage interventions to be informed by a person’s or a system’s response to a first level intervention; and 2) adaptive designs that use existing case data to modify intervention assignment rates. SMART-AR designs show great promise in improving both individual care and system performance. This organized presentation explores the impact of how procedures from all three presentations can be used to conduct trials to inform prevention, effectiveness, and implementation practices and research.