Background:
ASSIST (A Stop Smoking in Schools Trial) is a peer-led, smoking prevention, social network intervention developed in the UK. ASSIST trains students (aged 11-13) to work as peer supporters to disseminate non-smoking norms in their year. To date, evaluation of ASSIST has focused on delivery within schools only. This study sought to address this gap by using qualitative social network research to map the social networks of peer supporters to explore the extent, nature, content and perceived impact of message diffusion beyond the school year.

Methods:
Qualitative, face to face interviews conducted in school with 16 young people aged 12-13 incorporating the creation of 'egocentric sociograms' to collect network data. Sociograms were then used to create a qualitative narrative to explore the extent, and perceived impact, of message diffusion in the wider social networks of peer supporters. Computer packages UCINET and NVivo were used to conduct analysis.

Results:
Peer supporter networks were fairly large with a total of 155 people. Peer supporters recalled having conversations about smoking with 102 network members, two thirds (n=63) of which were with people out with their school year (e.g. parents, siblings, other family members, out of school friends, neighbours). Thematic analysis of conversation content revealed three types of conversation: protecting non-smokers from starting to smoke; encouraging smokers to stop; and protecting people in wider social networks. Perceived impact was noted for 37 people in peer supporter networks, with examples of positive and negative impact, focusing on the dynamics of a child speaking to an adult.

Conclusions:
Smoking prevention message diffusion was not limited to school year, reaching in to the wider networks of peer supporters. This creates learning opportunities for the future delivery of ASSIST or other similar peer-led interventions for smoking prevention, and the role of social network interventions in prevention.