Self-management of Vitamin K Antagonists Is More Cost- effective Than Dabigatran for Stroke Prevention in Non- valvular Atrial Fibrillation in Spain

We read with great interest the article by González-Juanatey et al. on the cost-effectiveness of the new oral anticoagulant dabigatran compared with conventional oral anticoagulant therapy (OAT) in patients with nonvalvular atrial fibrillation. This study consisted of a simulation using modern computer models applied to the Spanish setting. It concluded that dabigatran would be an effective strategy for stroke prevention when compared with two different scenarios: 100% of patients on OAT, or the usual prescription pattern (60% on OAT, 30% on aspirin, and 10% untreated). The authors concluded that dabigatran is efficient because the increase in costs does not exceed 30 000 euros/quality-adjusted life year gained, an acceptability threshold proposed in 2002. Increases in the cost-effectiveness ratio would be 17 581 euros/quality-adjusted life year for the scenario in which 100% of patients received OAT and would be 14 118 euros/quality-adjusted life year when the usual prescription pattern was considered. These results rely exclusively on data from the RE-LY (Randomized Evaluation of Long Term Anticoagulant Therapy) clinical trial, which compared dabigatran with OAT managed exclusively in specialist units on a monthly basis. However, while González-Juanatey et al.’s simulation takes into account the costs of different OAT management scenarios and includes weekly self-management, it does not consider the very different rate of serious complications observed depending on the model of OAT management used. For example, risk is clearly reduced by self-management compared with conventional OAT. Given the clinical and economic importance of these issues, clarification of some issues would seem opportune. The design, procedure, and interpretation of the results of the RE-LY trial have attracted considerable criticism; one of the major criticisms focused on the total lack of an advantage for dabigatran compared with conventional OAT when levels of international normalized ratio control, measured as the proportion of time in therapeutic range, are over 65% to 70%. It is unsurprising, therefore, that a recent publication which indirectly compared dabigatran and OAT self-management showed no significant difference in serious complications. On the contrary, there was a clear trend favoring self-management: the relative risk and its confidence intervals were 0.73 (0.48-1.10) for thromboembolism, 0.64 (0.40-1.01) for mortality, and 1.15 (0.83-1.60) for severe bleeding. Although direct comparative studies are lacking, these data should prompt health officials to rethink strategies for stroke prevention. Such a reevaluation is especially pertinent, given that 50% of patients receiving OAT in Spain can perform self-management with clinical guarantees and that a rigorous metaanalysis has confirmed the superiority of the self-management model over conventional OAT in specialized centers. Consequently, the last