Real-world bleeding risks of new anticoagulants similar to warfarin (or lower)
New-generation oral anticoagulants (NOACs) like dabigatran, rivaroxaban, and apixaban all had similar efficacy and bleeding risk compared to warfarin in the randomized trials that led to their FDA approval. That didn’t stop trial lawyers from filing bundles of lawsuits on behalf of patients experiencing hemorrhages after taking NOACs. (The large brand-name pharmaceutical manufacturers advertising new billion-dollar drugs are much better targets than the generic warfarin makers who eschew marketing and sell an old drug with well-known risks.)
One drugmaker paid out $650 million in out-of-court settlements, to plaintiffs asserting the drug was marketed as safer than it is. Another manufacturer had 18,000 lawsuits pending in mid-2017, and a third was ordered to pay a $28 million judgment to a single plaintiff who suffered GI bleeding.
All that bad publicity, and the frequent disparity between adverse outcomes in clinical trials and their real world use, have allowed questions to persist as to whether the bleeding risks of NOACs are really as low as advertised.
Fortunately, it sounds like they are.
Researchers in Canada analyzed retrospective data from almost 60,000 patients diagnosed with deep venous thrombosis or pulmonary embolism, in six different health care systems in Canada and the U.S.
About 12,000 had taken a NOAC (almost all were taking Xarelto); the rest had taken warfarin.
The rates of significant bleeding were the same or lower for patients taking NOACs (rivaroxaban) compared to warfarin: about 3% at 30 days, and about 4% at 60 days. Statistically, the rates of bleeding were equivalent, with a strong trend suggesting lower bleeding rates with NOACs. Bleeding was gastrointestinal in the vast majority of cases, for both warfarin and NOACs.
In a 2016 analysis of 45,000 patients insured by Medicare and commercial insurance with atrial fibrillation, those taking newer anticoagulants over a one-year period had a significantly lower or equivalent bleeding risk overall than those taking warfarin. A separate 2016 analysis covering 5 years of insurance claims data in tens of thousands of patients came to the same statistical conclusion.