Interventions:
The selection of the interventions was appropriate because the new programme was compared with no intervention, which represented the current pattern of care in the authors’ setting. This comparator was also likely to be relevant in other health care systems.
Effectiveness/benefits:
The data on treatment effect were taken from a randomised controlled trial, which should ensure high internal validity. Other clinical data were mainly derived from the programme implementation in a rural community. The participating individuals were interviewed to provide evidence on the risk of falling and other endpoints of interest. It was not clear whether a rigorous methodology was adopted to ensure the validity of these data. Furthermore, the authors supplemented these data with evidence from published studies, which may have referred to different populations or interventions. Thus, caution is required when interpreting the clinical findings.
Costs:
The economic viewpoint of the study was not explicitly stated. However the costs were mainly attributable to the third-party payer or health insurer. The steps of the cost calculations were clearly reported. Nevertheless, no clear information on the sources of data was given. Such information might be useful in defining the study perspective. The price year was not reported, which limits the possibility of replicating the cost analysis for other time periods. Finally, statistical analyses of the costs were not carried out and the validity of the cost estimates was not tested in a sensitivity analysis.
Analysis and results:
A synthesis of the costs and benefits was not carried out given the cost-consequences design of the analysis. In effect, the efficacy data were used only to calculate the costs. The issue of uncertainty was not extensively addressed because the sensitivity analysis focused on one uncertain parameter (the risk factors). The study findings were clearly presented. The authors discussed the additional non-financial benefits of the programme and some potential limitations were pointed out. Firstly, the evidence came from multiple sources which are difficult to pool because of possible background differences. Secondly, the analysis did not consider a drop-out rate from the programme as people moved out of their homes or died.
Concluding remarks:
As the authors acknowledged, the study was subject to some limitations, thus the conclusion should be treated with a degree of caution.