The approach used to select the two comparators was appropriate, as the routine care in the authors’ setting was compared against a physical activity promotion. The description of the two strategies was clear.
The clinical trial was well reported and appears to have been generally well conducted. The participants and practitioners appear to have been blinded to intervention allocation, power calculations were performed, and baseline characteristics were shown to be comparable between groups. These features strengthened the internal validity of the study. The authors stated that an intention-to-treat approach was not adopted because participants who dropped out refused to be tested after the intervention period. This use of a per-protocol analysis might have had an impact on the results, but it is not clear that an intention-to-treat analysis without the full data set would produce a less biased result. Overall the clinical trial was well reported and the uncertainty in the clinical results was appropriately investigated through bootstrapping.
The authors provided a full justification for the perspective and they appear to have considered those costs relevant to this perspective. The programme cost consisted of salary costs alone. This did not include other programme costs to the National Health System, such as medications and consultations, as these were found to be the same in the two groups. The exclusion of those costs occurring equally in both groups seems to have been valid. There were no other additional costs for the programme. Other details of the analysis, such as the price year and the sources of costs, were reported.
Analysis and results:
An appropriate incremental analysis was performed. The construction of a cost-effectiveness acceptability curve, using non-parametric bootstrapping, was an appropriate way to address sampling variation. The authors also presented a number of scenarios, in which the key assumptions or parameters were varied. This thorough exploration of uncertainty improves the quality of the analysis. The results of all scenarios and the baseline analysis were presented in detail.
Overall, the economic evaluation was well conducted and valid sources of data were used. These facts make the authors’ conclusions robust.