By Michael H. Crawford, MD, Editor

SOURCES: Chao TF, et al. Should atrial fibrillation patients with 1 additional risk factor of the CHA2DS2-VASc score (beyond sex) receive oral anticoagulation? J Am Coll Cardiol 2015;65:635-642; Calkins H. Data strengthen to support recommending anticoagulant therapy for all atrial fibrillation patients with a CHA2DS2-VASc score ≥ 1. J Am Coll Cardiol 2015;65:643-644.

Several guidelines recommend using the CHA2DS2-VASc score for determining the risk of stroke in patients with atrial fibrillation (AF), but the treatment recommendations are not the same in all guidelines. The controversy arises at scores between 0-1 and the effect of female sex. Thus, the objective of this study was to assess the risk of stroke in both sexes with one risk factor being female sex, which is a one-point risk factor in the CHA2D52-VASc score. The study was conducted in Taiwan and was a retrospective analysis of their national health system database of more than 23 million enrollees. From 1996-2013, 354,649 patients with AF were identified, their risk score calculated, and their pharmacy records examined. Patients receiving antithrombotic medications (warfarin and any antiplatelet agents) were excluded, for a final population of 186,570, among which 12,935 men had a risk score of 1 and 7900 women had a score of 2. The primary endpoint was ischemic stroke confirmed by brain imaging. Among the men with a score of 1, 14% experienced a stroke over a 5-year follow-up, for an annual rate of 2.75%. Not all risk factors carried equal rates, ranging from 1.96% per year with vascular disease to 3.5% per year for ages 65-74 years. In the women with a risk score of 2, 15% experienced a stroke, for an annual rate of 2.6%, ranging from 1.9% per year for hypertension to 3.3% per year for ages 65-75 years. The authors concluded that oral anticoagulants are recommended for anyone with AF and one risk factor beyond female sex, given the increased risk of stroke observed.


This study is potential game changer. The American College of Cardiology/American Heart Association/Heart Rhythm Society (ACC/AHA/HRS) guidelines state that a patient with a CHA2D52-VASc score of 1 can be treated with antiplatelet agents (APA), oral anticoagulants (OAC), or nothing depending on consideration of other risks for stroke, bleeding, and patient preference (IIb). This study should move the recommendation for OAC in these patients to a IIa. The authors point out that the risk of stroke in patients with one risk factor beyond female sex (2-3% per year) far outweighed the risk of major bleeding (0.25-1.45% per year) or intracranial hemorrhage (0.23-0.5% per year) on one of the new OACs. Hence, they recommended that OACs be considered in everyone with one risk factor beyond female sex.

They also analyzed the predictive power of each risk factor in the CHA2D52-VASc score. Interestingly, in both sexes, ages 65-74 years was the most potent individual risk factor and diabetes was the second. Hypertension, heart failure, and vascular disease were less predictive. This is not surprising because these risks are not simple binary factors. Mild, well-controlled hypertension is not the same as severe untreated hypertension.

The major strengths of this study are that they identified a large number of patients with AF, with one risk factor beyond sex, who were not on APA or OAC therapy and were followed for an average of 5 years. There are weaknesses with this study. Since it was a retrospective database study, we don’t know what the bleeding risk would be in the patients if they were treated with APA or OAC. Also, we don’t know if the AF was permanent, persistent, or paroxysmal, which could make a difference. There is no imaging data on the left atrium. A patient with a few occurrences of AF per year and normal left atrial size would probably have a different risk profile than someone with permanent AF and a left atrium that was severely enlarged. In addition, it is not known if the patients had other vascular diseases that could have caused their stroke such as aortic or carotid or intracerebral artery disease. Finally, we are not given these patient’s HAS-BLED score, which might explain why they were not on APA or OAC therapy.

The editorial accompanying this report states that the guidelines should not be changed for one retrospective study, but that physicians should incorporate this data into the discussion with the patient. I would certainly want patients with AF and one risk factor beyond sex on APA and would strongly suggest they consider OAC.