PURPOSE: Research involving adolescent risk behaviors must balance data confidentiality with participant safety when risky behaviors are revealed. This report details a safety protocol and reports the experience of two contemporaneous studies that used it with variant safety thresholds.
METHODS: We developed a safety protocol for research with adolescent patients and used it in two concurrent studies of adolescent patients, aged 14-18 years. Study "PC" recruited participants from a primary care adolescent medical clinic (N = 490), and Study "SP" recruited participants from subspecialty pediatric clinics (N = 434); both studies involved a similar self-administered assessment of health behaviors. The protocol sets thresholds for clinical intervention (positive safety flags) for past 3-month heavy alcohol consumption (Study PC: 10 or more drinks and Study SP: "binge-"level drinking), illicit drug use other than marijuana and alcohol in combination with a substance other than marijuana, and sets a positive screen for depression. We examined the rates of positive safety flags in both protocol settings, used significance testing to describe demographic differences between participants with and without positive flags in both studies, and described clinician experiences with protocol implementation.
RESULTS: In studies PC/SP, .6%/8.8% of participants were flagged for heavy alcohol consumption, respectively; .2%/0% for illicit drug use, 2.2%/.7% for combination substance use, and 14.9%/4.8% for depression. Some clinicians found managing positive flags challenging, although both studies completed recruitment on time and without serious adverse events.
CONCLUSIONS: The protocol was feasible in clinical settings. The findings and experiences documented in this report could be useful for future protocols.