Please take our brief survey

Blueprints Programs = Positive Youth Development

Return to Search Results

Promising Program Seal

A Stop Smoking in Schools Trial (ASSIST)

Blueprints Program Rating: Promising

A peer support program to reduce the uptake of smoking among young adolescents.

  • Suzanne Audrey
  • University of Bristol
  • Dept. of Social Medicine
  • Canynge Hall, Whiteladies Road
  • Bristol BS2 8PS
  • UK
  • suzanne.audrey@bristol.ac.uk
  • Tobacco

    Program Type

    • Peer Counseling and Mediation
    • School - Individual Strategies

    Program Setting

    • School

    Continuum of Intervention

    • Universal Prevention (Entire Population)

    A peer support program to reduce the uptake of smoking among young adolescents.

      Population Demographics

      ASSIST targets young adolescents in school.

      Age

      • Early Adolescence (12-14) - Middle School

      Gender

      • Male and Female

      Gender Specific Findings

      • Male
      • Female

      Race/Ethnicity

      • All Race/Ethnicity

      Race/Ethnicity/Gender Details

      Program effects did not differ for boys and girls, while differences across ethnic groups were not examined.

      • Individual
      Risk Factors
      • Neighborhood/Community: Laws and norms favorable to drug use/crime

      See also: A Stop Smoking in Schools Trial (ASSIST) Logic Model (PDF)

      A Stop Smoking in Schools Trial (ASSIST) is a smoking prevention intervention based on an informal, educational, peer-led approach. Influential students nominated by their peers are trained by health promotion trainers for two days on the risks of smoking, the benefits of remaining smoke-free, and skills for promoting non-smoking among their peers. The trained students then use informal contacts with peers over a 10-week period to promote non-smoking and keep a diary record of these conversations.

      A Stop Smoking in Schools Trial (ASSIST) is a smoking prevention intervention based on an informal, educational, peer-led approach. Selected influential students are trained for two days to use informal contacts with peers to encourage them not to smoke. They are asked to intervene informally in everyday situations over a 10-week period to promote non-smoking and keep a diary record of these conversations. Students in participating schools are given a brief questionnaire to nominate influential students. The nominated students receive a two-day training delivered by health promotion trainers that aims to: increase knowledge about the health, economic, social, and environmental risks of smoking; emphasize the benefits of remaining smoke-free; and encourage the development of skills to enable the selected "peer supporter" students to promote non-smoking among their peers.

      ASSIST is based on the "diffusion of innovation" model, whereby the diffusion of new behavioral norms through social networks is effected by influential local opinion-formers.

      • Normative Education

      The evaluation used a clustered randomized controlled design with 59 selected secondary schools randomized to either the ASSIST intervention or a control group that used regular smoking education curricula. The 29 schools assigned to the control group contained 5,562 students and the 30 schools assigned to the intervention group contained 5,481 students. Self-reported smoking at baseline, posttest, 1-year follow-up, and 2-year follow-up served as the outcome, while saliva tests for a smoking by-product validated the smoking measure.

      For all students, the program had a marginally significant effect (p = .058) at posttest, a significant effect at 1-year follow-up (p = .043), and a marginally significant effect at 2-year follow-up (p = .067). Across all three assessments, ASSIST achieved a significant reduction of 22% in the odds of being a regular smoker.

      Compared to students in control schools, students in ASSIST schools reported significantly lower

      • regular smoking at the 1-year follow-up

      Not examined.

      The odds ratio of .78 indicates a small effect size.

      The evaluation was limited to year 8 (ages 12-13) secondary school students in west England and southeast Wales from urban and rural districts.

      • Did not follow all students or impute data for intent-to-treat analysis
      • No formal tests for differential attrition, but data were imputed using a growth model that accounted for missing data

      • Blueprints: Promising

      Campbell, R., Starkey, F., Holliday, J., Audrey, S., Bloor, M., Parry-Langdon, N., ... Moore, L. (2008). An informal school-based peer-led intervention for smoking prevention in adolescence (ASSIST): A cluster randomised trial. The Lancet, 371, 1595-1602.

      Starkey, F., Audrey, S., Holliday, J., Moore, L., & Campbell, R. (2009). Identifying influential young people to undertake effective peer-led health promotion: The example of A Stop Smoking In Schools Trial (ASSIST). Health Education Research, 24(6), 977-988.

      Starkey, F., Moore, L., Campbell, R., Sidaway, M., & Bloor, M. (2005). Rationale, design and conduct of a comprehensive evaluation of a school-based peer-led anti-smoking intervention in the UK: The ASSIST cluster randomised trial. BMC Public Health, 5(43), 1-10.

      Sally Good
      CEO (Acting)
      DECIPHer IMPACT Limited
      2 Farleigh Court
      Old Weston Road
      Flax Bourton
      Bristol
      BS48 1UR
      Tel: +44 (0)1275 464779
      Mob: +44 (0)7791 692815
      www.decipher-impact.com

      Study 1

      Campbell, R., Starkey, F., Holliday, J., Audrey, S., Bloor, M., Parry-Langdon, N., ... Moore, L. (2008). An informal school-based peer-led intervention for smoking prevention in adolescence (ASSIST): A cluster randomised trial. The Lancet, 371, 1595-1602.

      Starkey, F., Audrey, S., Holliday, J., Moore, L., & Campbell, R. (2009). Identifying influential young people to undertake effective peer-led health promotion: The example of A Stop Smoking In Schools Trial (ASSIST). Health Education Research, 24(6), 977-988.

      Starkey, F., Moore, L., Campbell, R., Sidaway, M., & Bloor, M. (2005). Rationale, design and conduct of a comprehensive evaluation of a school-based peer-led anti-smoking intervention in the UK: The ASSIST cluster randomised trial. BMC Public Health, 5(43), 1-10.

      Campbell, R., Starkey, F., Holliday, J., Audrey, S., Bloor, M., Parry-Langdon, N., ... Moore, L. (2008). An informal school-based peer-led intervention for smoking prevention in adolescence (ASSIST): A cluster randomised trial. The Lancet, 371, 1595-1602.

      Starkey, F., Audrey, S., Holliday, J., Moore, L., & Campbell, R. (2009). Identifying influential young people to undertake effective peer-led health promotion: The example of A Stop Smoking In Schools Trial (ASSIST). Health Education Research, 24 (6), 977-988.

      Starkey, F., Moore, L., Campbell, R., Sidaway, M., & Bloor, M. (2005). Rationale, design and conduct of a comprehensive evaluation of a school-based peer-led anti-smoking intervention in the UK: The ASSIST cluster randomised trial. BMC Public Health, 5 (43), 1-10.

      Campbell et al. (2008) presented the program impact results on smoking that are reported below. Starkey et al. (2005) described the design but did not present results, and Starkey et al. (2009) described the peer nomination process and teacher and student satisfaction with the program but only summarized previous results on smoking impact.

      Evaluation Methodology

      Design:

      Recruitment: Of 223 secondary schools in west England and southeast Wales invited to participate in the study, 127 expressed initial interest, 113 agreed to participate, 66 were randomly selected, and 59 agreed to accept the randomization.

      Assignment: The evaluation used stratified block randomization to assign schools to treatment and control conditions. The 29 schools randomly assigned to the control group, which continued the usual smoking education and policies, contained 5,562 students, and the 30 schools assigned to the intervention group contained 5,481 students. Two schools withdrew after randomization, one from the control group and one from the intervention group. These schools were replaced by two schools from the same strata in the list of 113 interested schools, which were then “randomly allocated to treatment group as a block of size two.” The cluster design added new students entering the intervention and control schools and excluded students leaving the schools.

      Attrition: Students were assessed at baseline, posttest (about 4 months later), 1-year follow-up, and 2-year follow-up. According to the CONSORT diagram in Figure 1, the percentage of eligible students analyzed was 92-95% at baseline, 89-95% at posttest, 92-95% at the 1-year follow-up, and 89-94% at the 2-year follow-up. However, the count of eligible students excludes those leaving the schools and adds new students entering the schools.

      Loss of subjects came from students lacking parental consent, leaving the study, or moving from an intervention or control school. Two schools closed, but most of the subjects (95%) in the schools moved to a participating intervention school or control school, and remained in the study.

      Sample: The sample appears diverse. About 44% of the sample schools had 20% or more students eligible for free school lunches. The family affluence scale was distributed normally, with a mix of low (25%), medium (55%), and high (20%) scores. A majority of the families (53%) had two or more cars. About 6% of the students reported being weekly smokers at baseline.

      Measures: The primary outcome measure was self-reported smoking prevalence over the past week. The study distinguished regular smokers from occasional, experimental, or ex-smokers, who were considered at high risk for future regular smoking. The outcome measure was validated by saliva samples measuring cotinine, a metabolite of nicotine and a biomarker of smoke exposure in the previous two to three days. The saliva samples were obtained for all students at baseline and 1-year follow-up and for 24 schools (39% of the sample) at 2-year follow-up.

      Analysis: The separate analysis of outcomes at each wave calculated odds ratios and confidence intervals with adjustment for design effects and clustering. The combined analysis of all waves used random effects logistic regression models with school as a random effect and covariates for baseline smoking and the five school-level stratifying variables. The three-level model treated schools at level 3, students at level 2, and repeated follow-up measurements at level 1.

      Intent-to-Treat: The study used all available data but dropped randomized subjects who left the schools. The repeated-measures multilevel design allowed the inclusion of individuals with missing data for some of the assessments.

      Outcomes

      Implementation Fidelity: A survey of students identified influential peers, and the 17.5% of students in the intervention schools with the most nominations were invited to participate as peer supporters. Of the 5,358 nominated students, 835 (16%) completed the training and agreed to work as peer supporters, achieving the prespecified target. Of the students trained, 99% continued to work as peer supporters, and 84% handed in a completed diary at the end of the intervention period. Of possible relevance, the program worked no better for the peer supporters than regular students.

      Baseline Equivalence: Figures for schools and students (Table 1) showed no large differences but did not list significance tests. The text notes that more students in control schools (7%) reported smoking every week than did those in intervention schools (5%).

      Differential Attrition: The study provided no tests.

      Posttest and Long-Term: For all students, the adjusted odds ratios (Table 3) indicate a marginally significant effect (p = .058) at posttest, a significant effect at 1-year follow-up (p = .043), and a marginally significant effect at 2-year follow-up (p = .067). The overall odds ratio across all waves of .78 indicates a small effect size.

      For the subsample of students who were occasional, experimental, or ex-smokers at baseline, the adjusted odds ratios were insignificant at posttest, significant at 1-year follow-up, and marginally significant at 2-year follow-up.

      Subgroup analyses provided no evidence of the intervention having a differential effect by sex, peer supporter status, or free school meal entitlement. However, the intervention had a “more pronounced effect in schools located in south Wales valleys.”

      The concentration of salivary cotinine measured at the 1-year and 2-year follow-ups indicated that only 1-3% of students had high levels but reported not smoking in the survey. There was almost no difference in incorrect reporting between intervention and control schools.