GARFIELD-AF Analysis Digs Into Impact of Bleeding in Newly Diagnosed A-fib

But don’t lose site of the fact that the benefits of oral anticoagulation outweigh the risks, one expert cautions.

GARFIELD-AF Analysis Digs Into Impact of Bleeding in Newly Diagnosed A-fib

PARIS, France—Bleeding, particularly when it’s major, has a detrimental prognostic impact in patients newly diagnosed with atrial fibrillation, a new analysis from the GARFIELD-AF investigators affirms.

Patients who had any type of bleeding—including minor, clinically relevant nonmajor, and major bleeds—were more likely to die in the first year of follow-up compared with those who remained free of bleeding. Patients with major bleeds had an especially high risk, with more than a quarter dying within that time period.

Of note, risks of major bleeding, intracranial hemorrhage, and death were higher in patients treated with vitamin K antagonists (VKAs) rather than non-VKA oral anticoagulants (NOACs), confirming findings from the pivotal stroke prevention trials of the newer agents, Jean-Pierre Bassand, MD (University of Besançon, France), reported recently at the European Society of Cardiology Congress 2019.

“The conclusion should be: please prescribe NOACs instead of vitamin K antagonists if you can because . . . the rate of intracranial hemorrhage will be reduced by 50% and that will have a major impact on global mortality,” he said in a discussion following his presentation.


Bassand noted that even though bleeding’s negative impact on outcomes is known for many conditions, it is less well defined for patients with A-fib.

To explore the issue, he and his colleagues turned to the GARFIELD-AF registry, a prospective investigation involving more than 1,000 sites in 35 countries. The current analysis included 52,080 patients with newly diagnosed A-fib and follow-up of 1 year. The vast majority of patients did not have any ISTH-defined bleeding during that span, but there were 625 major, 525 clinically relevant nonmajor, and 1,103 minor bleeds.

The proportions of various types of antithrombotic regimens used were relatively consistent across the bleeding categories, with greater use of VKAs and less use of antiplatelet therapy alone in patients with versus without bleeding.

The chances of having a devastating bleed are lower than the chance of stroke in every risk category except for people who shouldn’t be on a blood thinner to begin with. William Lewis

The rate of major bleeding gradually increased from patients who were on antiplatelet monotherapy up to those who were taking a VKA plus an antiplatelet. Overall, major bleeding was about 50% more likely with VKAs than with NOACs (HR 1.51; 95% CI 1.00-2.28).

One out of every nine patients who bled in the first year died, including 26.2% of those with a major bleed, 6.9% of those with clinically relevant nonmajor bleeding, and 5.2% of those with a minor bleed. The respective hazard ratios for 1-year mortality across those three categories—compared with nonbleeders—were 8.03 (95% CI 6.58-9.80), 2.58 (95% CI 1.79-3.72), and 1.53 (95% CI 1.07-2.19).

Mortality risk was higher in patients taking a VKA versus a NOAC (HR 1.23; 95% CI 1.11-1.46).

Throughout the first year, intracranial hemorrhage accounted for 20.6% of all deaths in patients who bled. That includes 39.7% of deaths in the first 30 days and 5.0% deaths in the last 11 months of the follow-up period, when noncardiovascular causes like malignancy and respiratory failure took on greater importance. As seen in multiple prior studies, the risk of intracranial bleeding was about twice as high with VKAs relative to NOACs (HR 1.98; 95% CI 1.22-3.20).

Don’t Ignore the Benefits

Commenting for TCTMD, William Lewis, MD (The MetroHealth System, Cleveland, OH), said the analysis does not add much to what’s already known about bleeding in patients with A-fib.

“We know that people who bleed are at higher risk. There have been several studies that demonstrate that [association with] mortality and we know that tools such as the GARFIELD-AF tool can be used to predict mortality risk, bleeding risk, and stroke risk,” said Lewis, who is chair of the clinical work group for the American Heart Association’s Get With The Guidelines-AFIB program. 

He cautioned, however, against allowing the study’s focus on bleeding risks to distract from the positive effects of oral anticoagulants. “When you look at the data out there, the chances of having a devastating bleed are lower than the chance of stroke in every risk category except for people who shouldn’t be on a blood thinner to begin with,” he said, referring to patients with very low scores on risk prediction tools.

Even in patients who have had a major GI or intracranial bleed while on oral anticoagulation, Lewis pointed out, studies have shown that restarting anticoagulation—as opposed to avoiding it for good—is associated with better outcomes.

Though “primum non nocere” has been a guiding principle in medicine, Lewis said, “we’ve gotten to the point where we sometimes have to take the risk of a small amount of harm to gain a major therapeutic [benefit] . . . for patients.”

Todd Neale is the Associate News Editor for TCTMD and a Senior Medical Journalist. He got his start in journalism at …

Read Full Bio
  • Bassand J-P. Early and late prognostic implications of bleeding in patients with newly diagnosed atrial fibrillation: real world evidence from the GARFIELD-AF registry. Presented at: ESC 2019. August 31, 2019. Paris, France.

  • The GARFIELD-AF registry is funded by an unrestricted research grant from Bayer AG.
  • Bassand and Lewis report no relevant conflicts of interest.