Interventions:
The selection of the comparators was appropriate. The amiodarone strategy was based on the actual regimen administered in the authors’ setting.
Effectiveness/benefits:
The clinical evidence came from a well-conducted and well-presented clinical trial, which was published in a companion paper (Zebis, et al. 2007, see ‘Other Publications of Related Interest’ below for bibliographic details). The study design should ensure a high internal validity and the characteristics of the eligible population were clearly reported. Details of the study design, patient selection, and randomisation procedure were extensively presented and the baseline comparability of the groups will have made the comparison more robust. No summary benefit measure was used and, in effect, a cost-consequences analysis was conducted.
Costs:
The economic analysis was consistent with the perspective. A breakdown of cost items and their sources were reported for most of the costs together with the price year. The economic data were treated deterministically, but alternative cost estimates were appropriately tested in the sensitivity analysis. The resource use reflected the actual consumption of health services in the study sample. The authors acknowledged that the assessment of resource use might have been inaccurate due to the retrospective analysis and the imprecision of the hospital accounting system.
Analysis and results:
The cost-consequences framework precluded a synthesis of the costs and benefits, but the better economic and clinical profile of amiodarone suggests its dominance over placebo. The potential uncertainty in the analysis was investigated using a partial approach that focused on cost variations. The study appears to have limited external validity due to the use of local data and the restricted use of sensitivity analysis. The authors acknowledged that a limitation of their study was that only clinical atrial fibrillation was considered and not subclinical atrial fibrillation.
Concluding remarks:
The analysis was based on a well-conducted clinical study, although some caution will be needed when interpreting the authors’ conclusions, given the limited external validity.