Interventions:
The selection of the comparators was appropriate as the proposed interventions were compared with the usual care in the authors’ setting.
Effectiveness/benefits:
The treatment effect was assumed by the authors based on the potential impact of different reductions in risk factors on future cardiovascular events. This assumed relative risk was appropriately used to produce the annual risk of a first cardiovascular event, based on the person’s age and gender, using standard Framingham equations. The treatment effect was varied in the sensitivity analysis. QALYs were an appropriate benefit measure, as they capture the impact of CVD on a patients’ health. The utility values were appropriately from a NICE report and UK population norms for quality of life. Life-years saved were reported.
Costs:
The economic analysis was consistent with the perspective of the public payer, but only considered the costs of CVD; the intervention costs were not calculated. The estimated CVD cost savings indicated the maximum cost at which each programme remained cost-effective. A list of cost items was not given and the unit costs and quantities of resources were not presented separately. The price year and discount rate were reported. The cost estimates appear to have been treated deterministically and they were not varied in the sensitivity analysis.
Analysis and results:
The results were extensively presented in an appendix. An incremental approach was used to synthesise the costs and benefits of each programme compared with no intervention. The uncertainty was partly investigated, using a deterministic approach that only varied the reduction in relative risk. The authors stated that a probabilistic analysis would have been difficult since most of the clinical data were only available as point estimates. The methods used to estimate the QALYs gained and costs saved, in the mathematical model, were extensively presented in the web appendix, and an example of a 65-year-old individual was given. The analysis was specific to the UK and the authors did not discuss the transferability of their results.
Concluding remarks:
A conventional cost-effectiveness framework was used, but most of the evidence was from a published model and the details were not reported. The authors’ conclusions appear to be robust.