Interventions:
The rationale for the comparators was clear. The proposed intervention had demonstrated its efficacy in various studies, while the brief intervention was what students might have received in a typical school setting. A clear description of the two strategies was given.
Effectiveness/benefits:
Most of the evidence was derived from the analysis of a cohort of students, who were in a nationwide programme. The study had a large sample size and was a prospective analysis. The students in the two groups were matched in their characteristics. The authors pointed out that a conservative approach was taken, especially in terms of the ITT analysis, which might exclude the additional efficacy of the Not On Tobacco programme over the brief intervention. A limitation of the clinical study was the short follow-up, which might not have captured the true impact of the two programmes. This also appears to have been biased against the Not On Tobacco intervention which was more effective in smoking cessation. The benefit measure was appropriate for capturing the most relevant dimension of health for smoking adolescents, namely survival. LYs also allow comparisons to be drawn with the benefits of other health care interventions. Appropriate discounting was applied.
Costs:
The authors justified their selection of the perspective, on the grounds that schools incurred all the costs and responsibility for programme implementation. The authors stated that medical cost savings were not taken into account. In general, the cost study appears to have been valid, but the economic inputs were not fully described. The cost estimates and the data on resource consumption were only partially reported. All the economic inputs were treated deterministically, but some cost categories were omitted or only included in the sensitivity analysis.
Analysis and results:
The costs and benefits were appropriately combined using an incremental approach. The issue of uncertainty was investigated using a partial approach, but the worst- and best-case scenarios gave the range of possible cost-effectiveness estimates. The study results were clearly presented and discussed. The authors discussed some potential limitations of their analysis and these mainly related to the exclusion of some cost categories.
Concluding remarks:
The study was well conducted, but limited in its perspective and in the follow-up of clinical effectiveness. In general, the authors’ conclusions appear to be valid.
