A decision model was used to combine published evidence for the costs and outcomes of the interventions. A hypothetical cohort of 70-year old patients, with atrial fibrillation, who had a moderate risk of stroke and no contraindication to anticoagulant therapy was used. The time horizon was 20 years. The authors reported that the perspective was that of an insurance company or Medicare.
The effectiveness data were mainly from clinical trials, using an intent-to-treat analysis. The effectiveness of dabigatran, compared with warfarin, was from the Randomized Evaluation of Long-term Anticoagulant Therapy (RE-LY) trial. The effectiveness of the other interventions, in reducing stroke incidence, was from other published trials. The main clinical effectiveness estimate was the risk of ischaemic stroke.
Monetary benefit and utility valuations:
The utility estimates were from several published studies of the quality of life of patients with atrial fibrillation, incident stroke, recurrent stroke, and bleeds.
Measure of benefit:
The measure of benefit was quality-adjusted life years (QALYs) gained, and these were discounted at an annual rate of 3%.
The direct cost categories were adverse events and drugs. The costs of adverse events, including ischaemic neurological events, intracranial haemorrhage, bleeding, and myocardial infarction, were from published studies, the Healthcare Cost and Utilization Project, and Medicare. The costs of drugs, which included monitoring, were from prescriptions data and Medicare reimbursement rates. All costs were reported in 2010 US dollars ($) and were discounted at an annual rate of 3%.
Analysis of uncertainty:
One-way sensitivity analysis was undertaken, and the model parameters were varied over plausible ranges. The results of this analysis were presented in a bar graph. Two-way and three-way sensitivity analyses were undertaken.