Modeling of the impact on health outcomes of the use of dabigatran in patients with atrial fibrillation.

Background: Atrial fibrillation (AF) is the main direct cause of stroke. Prevention by anticoagulation or antithrombotic treatment is required, vitamin K antagonists (VKAs) and aspirin being the main agents. Dabigatran etexilate is a novel oral direct thrombin inhibitor. The RE-LY study demonstrated that in patients with AF, the rates of stroke and systemic embolism were similar (at a dose of 110 mg) or lower (at a dose of 150 mg) than those observed in patients treated with warfarin, a VKA. The aim of the present study was to estimate, through modeling, the number of severe events avoided with dabigatran at dosages of 110 mg (D110) or 150 mg (D150) twice daily compared to warfarin, when prescribed in the French population for patients with AF who meet the inclusion criteria of the Randomized Evaluation of Long-Term Anticoagulant Therapy (RE-LY) study.

Methods: We used a decision tree type model to simulate the outcome at 5 years in a cohort of patients eligible for treatment with dabigatran. We compared 3 hypothetical cohorts: all AF patients are treated with D110, D150 and warfarin. Based on the probabilities of occurrence of the different outcomes observed in the RE-LY study and in the Dijon Stroke Registry, we simulated for each year during 5 years the evolution of the 3 cohorts. The model allows to simulate a prevalent cohort of 461,392 patients at year zero, with a follow-up of 5 years ('constant prevalence model'), and it can also take into account incident patients during 5 years ('dynamic prevalence model'). The different events taken into account were: major hemorrhages (excluding hemorrhagic strokes), myocardial infarctions, hemorrhagic strokes, ischemic strokes, recurrence of strokes (without differentiating the mechanism) or deaths.

Results: Considering the constant prevalence model, the use of D110 instead of warfarin for the whole target population in France would permit to avoid 10,012 events and to save 18,879 years of life in a period of 5 years. These figures are 13,484 and 27,736 for D150 instead of warfarin. Considering the dynamic prevalence model, the use of D110 for the whole target population in France would permit to avoid 13,620 events and to save 22,674 years of life in a period of 5 years. These figures are 18,186 and 33,091 for D150.

Conclusion: The use of dabigatran would lead to a significant reduction of strokes and deaths attributable to AF in France.