Implementation of a peer-led school based smoking prevention programme: a mixed methods process evaluation

BACKGROUND: Smoking prevention programmes that reach adolescents before they experiment with tobacco may reduce the prevalence of tobacco use. ASSIST is a school-based, peer-led smoking prevention programme that encourages the diffusion of non-smoking norms among secondary school students (aged 12–13), and was shown in a randomised control trial (conducted 2001–2004) to reduce the prevalence of weekly smoking. This paper presents findings from a process evaluation of the implementation of ASSIST in Scotland in 2014–2017. It examines acceptability and fidelity of implementation and explores the context of message diffusion between peers. METHODS: Mixed method implementation study with students (n = 61), school staff (n = 41), trainers (n = 31) and policy and commissioning leads (n = 17), structured observations (n = 42) and student surveys (n = 2130). RESULTS: ASSIST was delivered with a high degree of fidelity to the licensed manual with all elements of the programme implemented. Student survey findings indicated that the frequency of conversations about smoking increased over the ASSIST delivery period (18% at baseline, 26% at follow-up), but student recollection of conversations about smoking with peer supporters was low (9%). The delivery context of ASSIST was important when considering perceptions of message diffusion. In the study schools, survey findings showed that 0.9% (n = 19) of participants were regular smokers (at least once a week), with nine out of ten (89.9%, n = 1880) saying they had never smoked. This very low prevalence may have affected when and with whom conversations took place. Study participants indicated that there were wider benefits of taking part in ASSIST for: peer supporters (i.e. personal and communication skills); schools (an externally delivered health promotion programme that required minimal resource from schools); and communities (via communication about the risks of smoking to wider social networks). CONCLUSIONS: ASSIST in Scotland was delivered with a high degree of fidelity to the licensed programme and was acceptable from the perspective of schools, students and trainers. Targeting ASSIST in deprived areas with higher youth smoking prevalence or in other countries where youth smoking rates are rising or higher than in Scotland may be particularly relevant for the future delivery.