Adaptation of the ASSIST peer-led smoking intervention to deliver information from the Talk to FRANK drug education website (ASSIST+FRANK): A pilot cluster-randomised controlled trial

Background Illicit drug use increases the risk of poor physical and mental health. Few effective school-based drug prevention interventions are available. We adapted an effective school-based peer-led smoking prevention intervention (ASSIST) to deliver information from the UK national drug education website, FRANK Methods We conducted a four-arm pilot cluster-randomised control trial of 1567 students aged 12-13 years in 12 schools in Wales. Three schools each (mean number students per year 131) were randomly allocated to: ASSIST+FRANK, FRANK friends, ASSIST, or usual practice which may or may not include drug eduation. ASSIST+FRANK is a peer-led smoking prevention intervention in UK Year 8 (ASSIST) followed by a drug prevention adjunct in Year 9 (+FRANK). FRANK friends is a stand-alone peer-led drug prevention intervention in Year 9. ASSIST involves 2 days off-site training, plus four follow-up sessions. +FRANK involves 1 day off-site training plus 3 followups. FRANK friends involves 2 days off-site training and four follow-ups. Influential students were trained to disseminate information on the risks of drug use and minimising harms. Primary outcome was the acceptability of the +FRANK and FRANK friends, which was assessed through interviews with students, staff, parents, and intervention staff (n=66), structured observations of delivery (195), exploratory effectiveness analysis (multilevel logistic models), and estimated costs. Secondary outcome was the lifetime use of any illicit drugs at 18 months. Intention-to-treat analyses were conducted unmasked to the intervention allocation. This trial was approved by Cardiff University School of Social Sciences Research Ethics Committee (SREC/1103; ISRCTN 14415936). Findings All 12 schools were recruited, randomised, and retained (ASSIST+FRANK 419 students, FRANK friends 440, ASSIST 347, usual practice 361). At the 18 month follow-up the student retention rate was 93% (1460/1567). +FRANK and FRANK friends were acceptable to students, teachers, and parents, and delivered as intended, but FRANK friends was preferred over +FRANK. Some staff felt pressured to meet all the learning objectives for +FRANK in 1 day training versus 2 days in FRANK friends. There could also have been a lack of novelty for +FRANK peer supporters since they had previously experienced the ASSIST intervention. The odds of lifetime drug use at 18 months was marginally lower in both the +FRANK arm and the FRANK friends arm than in the usual practice arm (38 [12·4%] of 307 vs 34 [13·4%] of 254 [odds ratio 0·96, 95% CI 0·58-1·59] and 30 [9·3%] vs 34 [13·4%] [0·70, 0·39-1·24], respectively). The estimated cost per school was £1942 for +FRANK and £3041 for FRANK friends. All progression criteria were met. Interpretation Although both interventions were acceptable to students, teachers, and parents, FRANK friends was preferred over +FRANK. There is, therefore, sufficient evidence to warrant a full-scale cluster-randomised controlled trial of FRANK friends.