Oral Anticoagulant Use for A-fib in Community Practice Appears Appropriate

 Download this article's Factoid (PDF & PPT for Gold Subscribers

In outpatients with atrial fibrillation (A-fib), the use of oral anticoagulants is high and appears to be appropriately more driven by stroke risk than by bleeding risk, according to a registry study published online June 11, 2013, ahead of print in Circulation: Cardiovascular Quality and Control.

Michael W. Cullen, MD, of the Mayo Clinic (Rochester, MN), and colleagues assessed community-based practices by looking at 9,957 outpatients with A-fib who were enrolled in the prospective Outcomes Registry for Better Informed Treatment of AF (ORBIT-AF) at 174 US centers from 2010 to 2011. At baseline, 76% of patients received oral anticoagulants, of whom 71% were on warfarin only and 4.9% on dabigatran only. A small minority (0.1%) received both drugs but not concomitantly.

Stroke and Bleeding Risk Both Influential

Overall, 71.2% had a CHADS2 score ≥ 2, indicating high stroke risk, and 26.2% had an ATRIA score ≥ 4, indicating high bleeding risk. Oral anticoagulant use was more prevalent in patients at increased stroke risk and less prevalent in those at increased bleeding risk (table 1).

Table 1. Oral Anticoagulant Use


(n = 2,394)

(n = 7,563)

P Value

CHADS2 Score
    ≥ 2





< 0.001

    ≥ 4





< 0.001

Abbreviation: OAC, oral anticoagulant.

The increase in anticoagulant use by stroke risk was progressive. The proportion receiving therapy was 52.5% among patients with CHADS2 scores of 0 and 80% among those with CHADS2 scores of 2 or more (P < 0.001). Moreover, use was higher in patients who had additional stroke risk factors.

Conversely, use declined progressively along with rising risk of bleeding, from 80.5% in patients with an ATRIA score of 3 to 63.8% in those with a score of 7 or higher (P < 0.001).

The interaction between stroke and bleeding risk proved to be significant (P = 0.012). However, the decision to prescribe anticoagulants was driven predominately by stroke rather than by bleeding risk as evidence by the increased use among those with low vs. high CHADS2 scores (table 2).

Table 2. Oral Anticoagulant Use: Interaction Between Stroke and Bleeding Risk


CHADS2 0-1
















Risk-Treatment Paradox Not Found

“One of the things that we had potentially expected to observe that we didn’t was the risk-treatment paradox in patients with stroke and heart attack,” said Dr. Cullen in a telephone interview with TCTMD. Previous studies have shown a contradiction, with acutely ill A-fib patients at the highest stroke risk not receiving oral anticoagulants even though they would derive the most benefit from treatment, he said.

Physicians may have been reluctant to prescribe the drugs because of existing comorbidities or because they were taking concomitant medications that already put them at increased risk of bleeding, Dr. Cullen explained.

Results Encouraging, but Selection Bias a Concern

According to Larry B. Goldstein, MD, of Duke University Medical Center (Durham, NC), the ORBIT-AF results are “encouraging because most other studies of oral anticoagulants show them being used half the time or less, and the percentages reported here are much higher in higher-risk patients.”

In a telephone interview, Michael D. Ezekowitz, MD, PhD, of Thomas Jefferson Medical College (Philadelphia, PA), noted that while the news is “very encouraging,” it raises the question of whether the patient population in this registry was more carefully selected than those seen in prior studies and thus may overestimate real-world use.

Dr. Goldstein noted that subjects in this study came mainly from cardiology, electrophysiology, and anticoagulation clinics, so the study population may not reflect the care of patients treated by general practitioners. While the registry included a large number of patients, “there’s a real concern about selection bias,” he acknowledged.

Low-Risk Patients Perhaps Overtreated

One surprising finding highlighted by Dr. Cullen is that 52.5% of the low stroke risk patients received oral anticoagulants. “The guidelines don’t recommend this,” he said, adding that the suggestion is that some of the lowest risk patients are being overtreated, an issue worthy of further investigation.

Dr. Goldstein agreed that this may be a concern but pointed out that the patients were being seen by subspecialty physicians and may have received anticoagulants for other reasons that are not apparent based just on CHADS2 score alone. In addition, Dr. Ezekowitz reported that cardiologists tend to use the CHA2DS2-VASc score when evaluating those at lowest risk of stroke.

Making decisions about anticoagulant use should “come down to individual patients,” said Dr. Cullen. While physicians have risk scores for guidance, patients and their caregivers may have reasons for not wanting anticoagulants, such as fear of bleeding based on prior experience, he stressed.

“Researchers need to conduct a prospective study looking at combined bleeding and stroke risk and evaluating a predefined treatment algorithm,” he said.


Cullen MW, Kim S, Piccini JP, et al. Risks and benefits of anticoagulation in atrial fibrillation: Insights from the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) registry. Circ Cardiovasc Qual Outcomes. 2013;Epub ahead of print.

Related Stories:

Oral Anticoagulant Use for A-fib in Community Practice Appears Appropriate

In outpatients with atrial fibrillation (A fib), the use of oral anticoagulants is high and appears to be appropriately more driven by stroke risk than by bleeding risk, according to a registry study published online June 11, 2013, ahead of
  • The study received financial support from Johnson &amp; Johnson.
  • Drs. Cullen and Goldstein report no conflicts of interest.
  • Dr. Ezekowitz reports serving as a consultant to Bayer/Johnson &amp; Johnson, Boehringer Ingelheim, Bristol-Myers Squibb/Pfizer, and Daiichi Sankyo.