Variation in Intervention Response As a Function of Variation in Participation: Applying What We Know to Translate Efficacious Preventive Interventions to Real-World Delivery Settings

High levels of participation (i.e., dosage) are rarely feasible, even under the best circumstances in prevention efficacy trials, and are likely to be compromised further when interventions are implemented under real-world conditions.  Because prevention research unequivocally links variation in participation to variation in program proximal and distal outcomes, understanding what predicts levels of participation is critical in determining “what works for whom.” Identifying participant, program, and delivery-context characteristics that predict participation can inform translation of efficacious interventions to real-world settings. In this roundtable, discussants and the audience will dialogue about how prevention scientists can apply what we know about predictors of participation and the effects of dosage on outcomes to move effacious interventions to real-world delivery settings.  Discussant expertise will be diverse and include perspectives from program developers with experience adapting programs to maximize response, a representative from the NIDA, a methodologist with extensive experience evaluating preventive interventions, and a senior federal researcher involved in examining evidence-based prevention services in states. Sample questions we will address include: 

1. How do prevention researchers reengineer and trim interventions and delivery mechanisms to ensure participants receive the active ingredients and dosage needed to achieve effects? (e.g., Do video and other web-based technologies increase dosage or, alternatively, have detrimental effects on dosage?) 
2. Research shows that participation levels may be lowest for high-risk families for whom the program potentially has the most benefits and who also require the highest dosage. What are the implications of this research finding for understanding what interventions work for whom?    
3. Efficacious programs that achieve lower than expected participation levels when delivered in real-world settings are vulnerable to adoption failure. Given what prevention scientists know about the estimated effects of participant, program, and delivery-context characteristics on participation are there extant methodologies that can adjust dosage effects in efficacy trials to reflect real-world delivery expectations?
4. Strong attendance in the early sessions of an intervention is often followed by a decline in attendance. What are the implications of this phenomenon for understanding program change mechanisms and determining active ingredients? 
5. How can we harmonize the results of dose-response analyses across several studies, given that different programs vary in length, may use different methods to assess dose,  and target qualitatively different samples.