Home-monitoring of oral anticoagulation vs. dabigatran. An indirect comparison

Oral anticoagulation with vitamin k antagonists (VKAs) requires regular testing and dose adjustment. Home-monitoring (self-testing or self-management) is more effective than usual management. Dabigatran, does not require dose-adjustment and appears to be more effective at reducing the risk of stroke with similar risks of bleeding in patients with atrial fibrillation (AF). Dabigatran, however, has not been compared to the home-monitoring. It was the objective to evaluate the efficacy of dabigatran compared with home-monitoring of oral anticoagulation with VKAs. Randomised controlled trials (RCTs) comparing usual management of oral anticoagulation with home-monitoring, dabigatran with usual management, and RCTs comparing dabigatran with home-monitoring and including patient-important outcomes (thromboembolic events, death and major bleeding) were eligible. For our direct comparison we calculated pooled relative risks (RRs) using the Mantzel-Haenzel random effect model. For the indirect comparison we estimated lnRRs and back transformed to RR. We evaluated the quality of the evidence with the GRADE system. Dabigatran, compared with warfarin, was associated with lower rates of stroke or thromboembolism and systemic embolism but similar rates of major haemorrhage and death. Dabigatran 150 mg also increased non-significantly the rate of myocardial infarction. The quality of the evidence was high. Our indirect comparison of home-monitoring of oral anticoagulation versus dabigatran showed no convincing differences in the risk of thromboembolism, death or major bleeding. The estimates for self-management vs. dabigatran showed stronger but still non-significant trends. The quality of the evidence was low. In conclusion, the indirect comparison of home monitoring of oral anticoagulation with dabigatran suggests that the treatments have similar impact on thrombosis, bleeding and death. However, the confidence in the estimate of effect is low to very low. Our analyses contrast with the available comparison of dabigatran with conventional warfarin monitoring.