An analysis of FDA adverse event reports concluded that apixaban had the safest drug profile among anticoagulants and prasugrel had the highest safety risk among antiplatelet drugs.

AdverseEvents released a report in February that analyzed data from the FDA’s Adverse Event Reporting System (FAERS) with a focus on blood thinners. The medications were categorized as either anticoagulants—warfarin and the newer drugs dabigatran (Pradaxa, Boehringer Ingelheim), rivaroxaban (Xarelto, Janssen Pharmaceuticals/Bayer HealthCare) and apixaban (Eliquis, Bristol-Myers Squibb/Pfizer)—or antiplatelets—clopidogrel (Plavix, Bristol Myers-Squibb/Sanofi Aventis), ticagrelor (Brilinta, AstraZeneca) and prasugrel (Effient, Eli Lilly/Daiichi Sankyo).

“Given that blood-thinning drugs represent some of the most widely used medicines in the world and they are disproportionately used in both the elderly and other high-risk patient populations, we used multiple analytic platforms to analyze post-marketing safety signals for drugs in this class,” the authors wrote.

Analysts relied on the company’s RxFilter, which addresses some of the challenges with FAERS data: The filter identifies and corrects misspelled drug names, aggregates pharmaceuticals that have a variety of brand names into one name, removes duplicate reports and identifies common adverse events and conditions. They studied data from each drug’s FDA approval date through December 2012, with a few exceptions: the data for warfarin started at November 1997, decades after its approval; and for apixaban and ticagrelor, they obtained data up to December 2013.

They applied a proprietary algorithm based on six categories to calculate a risk-adjusted score, with 100 indicating the highest potential adverse event risk. In the anticoagulant group, they determined that warfarin had the highest risk score, at 67.56, followed closely by dabigatran (67.15) and rivaroxaban (67.08). Apixaban received the lowest score, at 39.45.

Under antiplatelets, prasugrel placed highest, with a score of 81.13, followed by ticagrelor (68.41) and clopidogrel (63.82).  

They acknowledged several limitations with using the FDA’s post-marketing data, including the potential for underreporting, overreporting and confounding. In a response, Boehringer Ingelheim also stressed the limitations of FAERS, citing the FDA’s cautions that reports can be incomplete and do not require a causal relationship, making the data inappropropriate for calculating incidence of adverse events.