Objective: This article presents a meta-analysis of 48 teen cigarette smoking cessation studies, the 1st meta-analysis of its kind. The authors conducted searches of electronic databases and unpublished manuscripts from 1970 to 2003. Fifty contextual elements were coded from each article. A weighted least squares random effects method was used to pool results from individual study net effects estimates. Multilevel random coefficients modeling was applied to control for the intrastudy variation.
Main outcome measures: The primary outcome for the present analysis was quit rate. Data were entered as intent-to-treat (ITT) quit rates (not compliance sample rates). Overall treatment effect size and treatment effect sizes as a function of program content, modality, number of sessions, and length of follow-up were examined.
Results: Across studies, program conditions, compared with control conditions, appeared to give smokers a 2.90% (95% confidence interval = 1.47-4.35%) absolute advantage in quitting, increasing the probability of quitting by approximately 46% (9.14% vs. 6.24%). Relatively higher quit rates were found in programs that included a motivation enhancement component, cognitive-behavioral techniques, and social influence approaches. Also, relatively higher quit rates were found in school-based clinic and classroom modalities. Furthermore, relatively higher quit rates were found for programs consisting of at least 5 quit sessions. Data also indicated that the effects were maintained at short-term (1 year or less) and longer term (longer than 1 year) follow-ups.
Conclusion: Much more teen smoking cessation research is needed, but teen smoking cessation programming is effective, and the present study provides a framework to move forward.