LAUSR

BACKGROUND This study examined approaches to delivering brief interventions (BI) for risky substance use and sexual behaviors in school-based health centers (SBHCs). METHODS 300 Adolescents (ages 14-18; 54 % female) with risky marijuana and/or alcohol use identified via CRAFFT screening (scores >1) were recruited from two SBHCs and randomized to computer-delivered BI (CBI) or nurse practitioner-delivered BI (NBI). Both BIs included motivational and didactic content targeting marijuana, alcohol, and risky sexual behaviors. Assessments at baseline, 3-month, and 6-month follow-up included past 30-day frequency of marijuana use, alcohol use, binge drinking, unprotected sex, and sex while intoxicated; marijuana and alcohol problems; and health-related quality-of-life (HRQoL). A focused cost-effectiveness analysis was conducted. An historical 'assessment-only' cohort (N=50) formed a supplementary quasi-experimental comparison group. RESULTS There were no significant differences between NBI and CBI on any outcomes considered (e.g., days of marijuana use; p=.26). From a cost-effectiveness perspective, CBI was 'dominant' for HRQoL and marijuana use. Participants' satisfaction with BI was significantly higher for NBI than CBI. Compared to the assessment-only cohort, participants who received a BI had lower frequency of marijuana (3-months: Incidence Rate Ratio [IRR] = .74 [.57, .97], p=.03), alcohol (3-months: IRR = .43 [.29, .64], p<.001; 6-months: IRR = .58 [.34, .98], p = .04), alcohol-specific problems (3-months: IRR = .63 [.45, .89], p=.008; 6-months: IRR = .63 [.41, .97], p = .04), and sex while intoxicated (6-months: IRR = .42 [.21, .83], p = .013). CONCLUSIONS CBI and NBI did not yield different risk behavior outcomes in this randomized trial. Supplementary quasi-experimental comparisons suggested potential superiority over assessment-only. Both NBI and CBI could be useful in SBHCs.