Participant Attrition Associated With Longitudinal Use of Safety Flag Protocols in Adolescent Substance Use Research.

PURPOSE Safety flag (SF) protocols are increasingly used in adolescent substance use research to protect of minor participants. This report examines the relationship between different thresholds for reporting participant substance use and study attrition. METHODS Data were analyzed from 2 concurrent adolescent studies that used an identical SF protocol with the exception of the threshold for heavy episodic drinking (HED). The first study recruited participants from a primary care adolescent medical clinic and the threshold for clinical intervention for past 3-month heavy episodic drinking was 10+ drinks on a single occasion. The second study recruited youth with chronic medical conditions from subspecialty pediatric clinics and used a lower threshold of 3+ to 5+ drinks on a single occasion. Generalized estimating equations were used to assess associations between SF type and subsequent attrition. RESULTS The baseline analytic sample size was 921 (10+ threshold = 487, 3+ to 5+ threshold = 434). No significant relationship between SF type and attrition was observed in either cohort. CONCLUSION Positive SFs were not associated with significant differences in participant attrition. These results suggest that researchers can use different SF thresholds to accommodate adolescent populations with varying levels of risk without compromising study goals.