Dabigatran Increases “Real-World” Bleeding

Payal Kohli, MD

The risk for major bleeding and GI hemorrhage in patients taking dabigatran was greater than in those taking warfarin; warfarin, however, was associated with a greater risk of intracranial hemorrhage.

The last few years have witnessed a surge of clinical trials and excitement over the novel oral anticoagulants (NOACs) and their clinical application for atrial fibrillation (AF). Their ease of use, lack of required monitoring, and limited drug-drug interactions have facilitated their uptake in the cardiovascular community. However, there is now emerging data from “real world” samples and post-marketing surveillance that the NOACs may be associated with more risks than previously demonstrated.

The risk for bleeding remains the greatest hazard with NOACs. A study recently published in JAMA-Internal Medicine analyzed 5% of random Medicare beneficiary pharmacy and claims data to determine if there were any differences in major or minor bleeding events among new users of dabigatran (n=1302) compared to new users of warfarin (n=8012), who initiated either treatment within 60 days of diagnosis. Four high-risk subgroups were also analyzed: age≥75 years; African Americans; those with chronic kidney disease (CKD); and those with >7 comorbidities. Whereas dabigatran was associated with a 1.3 times higher risk for any bleeding and a 1.85 times higher risk for GI bleeding relative to warfarin, warfarin was associated with a 3.1 times higher risk for intracranial hemorrhage. Dabigatran was consistently associated with risk of major bleeding and GI hemorrhage across all high risk subgroups.

The vasculature in the brain somehow appears pathophysiologically distinct from the vasculature in other anatomical sites. In this “real-world” analysis of dabigatran, there was a lower risk of intracranial hemorrhage compared to warfarin but there continued to be higher bleeding rates at other sites. Particularly interesting was a higher risk for bleeding observed in African Americans. One can imagine, then, that perhaps the least optimal patient for dabigatran therapy would be an elderly, African American with CKD and multiple comorbidities. This study also nicely illustrates the differences in results that are often seen between randomized controlled trials and post-marketing surveillance and observational data. It also underscores the importance of both types of studies for making the most highly informed clinical decision.







Hernandez I, Baik SH, Pinera A, Zhang Y. Risk of bleeding with dabigatran in atrial fibrillation. JAMA Intern Med. Published online November 03, 2014. doi: 10.1001/jamainternmed.2014.5398