The scenarios assessed were well described and justified. The details of the specific resources for the smoking cessation programme were available from a previous study (Olsen, et al. 2006, see 'Other Publications of Related Interest' below for bibliographic details).
The predicted health outcomes for the strategies were based on relevant national epidemiological studies and these reports should be consulted to assess their validity. The authors acknowledged a number of limitations to their analysis, including the potential overestimation of the health effects, in the context of declining prevalence; the omission of the quality of life; the omission of the cost impact of reduced passive smoking; and that the quit rates were self-reported, without biochemical validation. The lack of evidence for smoking bans was also highlighted as a limitation.
The costing methods were briefly described, but the unit costs and quantities for each resource input were not presented. These should be available in the previous publication (Olsen, et al. 2006), which would need to be consulted to fully understand the costing analysis. The productivity losses were presented separately to the health system cost impacts. All costs relevant to the perspective appear to have been considered.
Analysis and results:
The structure of the model and a full description, with a diagram, were presented. Scenario analyses were undertaken, but they did not fully address uncertainty. Analyses varying the key inputs, such as the quit rates, costs and relative risks of death, could have characterised more of the parameter uncertainty. The authors highlighted that their results should be interpreted cautiously due to the evidence for the effects of prevention activities being from an English study rather than a Danish one.
The quality of the analysis was sufficient, but the uncertainty in the inputs was not fully captured. The conclusions are likely to be valid, but the presence of uncertainty cannot be overlooked.