High Rates of Unnecessary Aspirin Use in A-fib/VTE Patients on Warfarin

Anticoagulation clinics offer a key opportunity to stop aspirin use if it’s harming more than it helps, the senior author says.

High Rates of Unnecessary Aspirin Use in A-fib/VTE Patients on Warfarin

More than one-third of patients taking warfarin for atrial fibrillation or venous thromboembolism who were being followed at anticoagulation clinics in Michigan over a 7-year period were also taking daily aspirin, despite any clear therapeutic indication for doing so, a new study shows.

In a propensity-matched analysis, rates of bleeding and bleeding-related hospital visits were higher among patients taking both the antiplatelet and the anticoagulant as compared with patients taking warfarin alone. Rates of thrombosis, however, were no different between the two groups.

Of note, no falloff in rates of concomitant aspirin prescription was seen following the release of major studies suggesting that combined use of antiplatelets and anticoagulants was not beneficial and guidelines, including 2012 guidance from the American College of Chest Physicians, recommending warfarin monotherapy for patients with atrial fibrillation and stable CAD. Likewise, the 2016 European Society of Cardiology atrial fibrillation guidelines also advise against combined antiplatelet and anticoagulant use.

That aspirin is still so frequently used by patients also taking warfarin is “surprising,” senior author Geoffrey D. Barnes, MD (University of Michigan, Ann Arbor), told TCTMD, adding that information in this study is “pretty reliable” since it’s abstracted from charts from the anticoagulation clinic, where staff rigorously question patients about the medications they’re taking. “I was hopeful that [the proportion taking aspirin] was going to be lower,” he admitted. “This number was higher than what we've seen in some other places and that did surprise me a little bit, how many people have no real reason to be on aspirin.”

The analysis, with lead author Jordan K. Schaefer, MD (University of Michigan), was published online March 4, 2019, ahead of print in JAMA Internal Medicine.

The study cohort included 6,539 adults (51% men) being managed for atrial fibrillation or venous thromboembolism (without recent MI) at one of six anticoagulation clinics in Michigan between 2010 and 2017. For the entire group, 37.5% were taking aspirin with no clear therapeutic indication.

Barnes and colleagues then analyzed two propensity-matched cohorts to compare 1-year outcomes among patients taking warfarin plus aspirin versus warfarin alone. Compared with patients taking warfarin alone, rates of bleeding, major bleeding, emergency department (ED) visits for bleeding, and hospitalizations for bleeding were all significantly higher among patients taking unnecessary aspirin on top of warfarin.

One-Year Outcomes


Warfarin + Aspirin


P Value

Overall Bleeding



< 0.001

Major Bleeding



< 0.001

ED Visits for Bleeding




Hospitalizations for Bleeding








“There's been a lot of studies showing that even when we have evidence-based practices, it can take up to 17 years for those practices to become routine in everyday care, so I'm not surprised that we didn’t see a decline during our study period,” even in the wake of different trials and guidelines, Barnes said. “I think it's going to take some really thoughtful implementation studies to figure out how we can reduce the risk.”

Empowering Allied Health Professionals

Barnes stressed that these were all patients taking aspirin at the time they enrolled at the clinics—the study didn’t look at patients started on aspirin for another reason after entering the clinic’s surveillance. And there are many factors that play into a patient taking aspirin, despite guidelines recommending against antiplatelet therapy in this setting.

“We have lots of different prescribers who may recommend an aspirin, or a patient might take it because they've heard, ‘Oh, I’m 50, I'm supposed to start taking it,’ or, ‘My mother took an aspirin every day so I should too.’ And there's this sense that it's safe,” Barnes said. “A second major factor is that it's not reliably measured in our electronic medical records because it's not a pill we prescribe—people get it over the counter—so doctors, especially the doctors managing anticoagulants, may not know that their patients are taking aspirin. They may not have that reliable information.”

Barnes said he believes pharmacists and nurses working at anticoagulation clinics like the ones studied in the current research are well positioned to play a bigger role in catching the patients who are taking aspirin and shouldn’t be.

“When different healthcare providers realize that a patient is on aspirin and an anticoagulant like warfarin, they may not feel empowered to make a change. Even in the anticoagulation clinic they may say, ‘Well, if the doctor put them on it, they must want them on it.’ And instead we may need to think about strategies to empower that nurse to ask and raise the question with the physician: ‘Hey, does that patient really need to be on aspirin or could we stop?’”

We may need to think about strategies to empower that nurse to ask and raise the question with the physician: ‘Hey, does that patient really need to be on aspirin or could we stop?’ Geoffrey D. Barnes

The next questions, Barnes continued, are “not just how can we reduce the risk or will we reduce the risk, but will that actually lead to fewer bleeding events? And those are the key questions that will come next.”

Also unanswered is the question of whether these findings would be similar in patients taking the newer direct oral anticoagulants.

“I would guess that we'll see similar rates of aspirin use among patients who are on DOACs, but who don't have an indication for aspirin,” Barnes predicted. “What I don't know about is, will we see the same increased risk of bleeding? I presume so, because the more blood thinners you're on, the more you're on, although we do know that many of the DOACs have lower bleeding risks than warfarin, especially apixaban—there's been a number of studies that have shown that. So how apixaban plus aspirin will ultimately come out with bleeding risk remains to be seen.”

The authors hope to look into this but didn’t have the length of follow-up for the newer oral anticoagulants that they had for warfarin at these clinics. “We're still trying to build up that cohort so we can look at that question,” Barnes said. “But I think it's a really good question.”

  • Barnes reports consulting for Pfizer/Bristol-Myers Squibb, Portola, and Janssen and receiving grant support from Pfizer/Bristol-Myers Squibb, Blue Cross Blue Shield of Michigan, and the National Heart, Lung, and Blood Institute.

We Recommend