Ground Report

BRISTOL, England, May 9 — Much as physicians take clinical clues from so-called thought leaders, adolescents may be ready to accept the no-smoking message from the most popular and influential kids in school. Action Points ——————————————————————————– Explain to interested patients that this study found that a smoking prevention program delivered by one eighth-grader to another was effective after two years. For adolescents, the message of smoking prevention may be best delivered by someone their own age outside the classroom, researchers here suggested. In a randomized trial involving 59 schools in England and Wales, eighth-grade students exposed to a smoking prevention program led by their peers outside a formal setting were 22% less likely to be smokers two years later than students at control schools (HR 0.78, 95% CI 0.64 to 0.96). So reported Rona Campbell, Ph.D., of the University of Bristol, and colleagues in the May 10 issue of The Lancet. "Although systematic reviews have shown mixed evidence that school-based approaches for smoking prevention are effective," they said, "the results presented here … suggest that abandoning interventions for smoking prevention in schools would be premature." Because most peer-led approaches to smoking prevention have been classroom-based and lacking in rigorous evaluation, the researchers set out to assess the effectiveness of an approach delivered by eighth-graders to their peers in more informal settings called A Stop Smoking In Schools Trial (ASSIST). The program consisted of identifying influential students by surveying their peers and training them to talk to their classmates about not smoking over a 10-week intervention. A total of 59 schools containing 10,730 students ages 12 and 13 were randomized to a peer-led intervention plus their regular anti-smoking initiatives (30 schools, 5,358 students) or to regular initiatives only (29 schools, 5,372 students). Smoking prevalence in the schools was assessed by self-reports and confirmed by measuring cotinine concentrations in saliva samples. More than 90% of students provided information at every data collection point — immediately following the intervention and at one and two years. In the intervention schools, 835 students (16%) were selected and trained as peer supporters to deliver the intervention. Most (84%) turned in a completed diary chronicling the conversations they had as part of the program at the end of the intervention. At baseline, more students in the control schools reported smoking regularly (7% versus 5%). Overall — including both intervention and control schools — the smoking rate increased from 5.7% when the students were 12 to 13 at baseline to 13.8% at the one-year follow-up to 20.3% at the two-year follow-up. Smoking prevalence was lower in the intervention schools than in the controls at each data collection point during follow-up but the difference was only significant at one year (12.49% versus 15.13%, P=0.043). That equated to a 23% reduced chance of being a smoker for students in the intervention schools (HR 0.77, 95% CI 0.59 to 0.99). The intervention did not have a greater effect on students who were deemed to be at high risk for smoking regularly — those who occasionally smoked or experimented or ex-smokers — as the researchers had expected. They said that performing the intervention more than once would probably extend the effect. "If the program was repeated every year with successive year groups, it would probably have an effect on the cultural norms surrounding smoking behavior in the whole school, magnifying the effect of the intervention." Robin Mermelstein, Ph.D., of the University of Illinois at Chicago, wrote in an accompanying editorial that these results provide "promising evidence for a way to bolster the effectiveness of school-based programs." She listed several strengths to the study but noted that a previous publication about the trial mentioned that the peer supporters had several conversations immediately following training but then engaged in fewer and fewer. Also, many of the conversations were with friends and non-smokers and not with regular smokers, she said. "Thus, the degree of penetration of the peer supporters into the many social networks might not have been deep or broad enough to show lasting effects," she said. "This tenet argues perhaps for a reinstatement of the peer-supporter nomination and training process on a yearly basis." Consideration should also be given to interventions involving other social influences, including family members and advertising, she said. "As encouraging as ASSIST’s findings are," she concluded, "an important message is the need to go beyond the classroom setting and into the many domains of social influence that adolescents encounter."