Addiction can occur rapidly in young smokers with consequent health problems in later life. Valuable resources are wasted when school interventions are ineffective, underscoring the need for evidence-based interventions. Approaches are often school-based owing to consistent access to students, with peer-led programmes proposed as a way forward.

This cluster randomised controlled trial targeted children aged 12–13 years, aiming to spread non-smoking behaviour in schools with peer help but in a hitherto uncharted informal out-of-classroom setting, since a systematic review of peer-led studies that used a conventional classroom-based approach showed variable effectiveness, perhaps contributed to in part by the inevitable loss of credibility when peers adopt formal teaching roles.

A total of 10 730 students in 59 schools were randomised to usual smoking education or trial intervention with follow-up being immediately and at 1 and 2 years after randomisation. The primary outcome was smoking in the past week. On intention-to-treat analysis, the odds of smoking at all three follow-up periods were lower in intervention schools, with a significant risk reduction at 1 year that diminished to a non-significant reduction at 2 years. Significance, however, was retained after the planned multilevel modelling to compensate for loss to follow-up.

This robust study with its well planned statistical modelling and encouraging retention rates demonstrates the success of an informal peer-led approach with an effect that seems to last. The importance of the social aspect to smoking shown by this study suggests that other potential vistas may include exploration of the role played by siblings and targeting smoking promotion campaigns.

▸ Campbell R, Starkey F, Holliday J, et al. An informal school-based peer-led intervention for smoking prevention in adolescence (ASIST): a cluster randomised trial. Lancet 2008;371:1595–602.