Oral anticoagulants cut events in the atrial fibrillation group guidelines leave undecided
SINGLE-AF in South Korea compared a direct oral anticoagulant with no anticoagulation among 1,803 people with atrial fibrillation and a CHA2DS2-VASc score of 1 in men or 2 in women. After 24 months the combined rate of stroke, systemic embolism, major bleeding and cardiovascular death was 0.5 per cent against 1.5 per cent, and major bleeding did not rise. Healio reports this as the first direct test of a class IIa guideline recommendation that rests on observational studies.
Science··Midday
A first direct comparison in the intermediate-risk band
SINGLE-AF at 18 centres in South Korea compared a direct oral anticoagulant with no anticoagulation among 1,803 people with atrial fibrillation and a CHA2DS2-VASc score of 1 in men or 2 in women. Those enrolled took apixaban 5 mg twice daily or rivaroxaban 20 mg once daily. Boyoung Joung told Healio that guidelines give anticoagulation a class IIa recommendation in this group based on observational studies with mixed results, and that SINGLE-AF is the first trial to test the question directly.[1], [2]
The combined event rate fell to one third
After 24 months the combined rate of stroke, systemic embolism, major bleeding and cardiovascular death was 0.5 per cent on the drug and 1.5 per cent on no anticoagulation. Stroke appeared in 0.3 per cent on a direct oral anticoagulant versus 1.1 per cent without; there was no difference in systemic embolism or major bleeding and no cardiovascular deaths in either group. Healio reports fewer major adverse events than no anticoagulation and no rise in major bleeding. Event counts were low in both arms, which limits how firmly the gap can be read.[1], [2]
Congress presentation and reimbursement hopes
The results were presented at the ESC Congress 2026 and published in the New England Journal of Medicine. Joung said the study may expand reimbursement in South Korea, where oral anticoagulation was not covered for men with a score of 1 or women with a score of 2 when the trial began about ten years ago. That score band matches the intermediate-risk group many guidelines leave without a firm treatment rule. Low event counts support the direction of the result but may call for broader follow-up.[2], [1]
