A DOAC lowers events in atrial fibrillation at intermediate stroke risk

Design

Multicentre, open-label, adjudicator-masked, randomised trial (SINGLE-AF)

Population

1,803 adults with atrial fibrillation and CHA2DS2-VASc 1 (men) or 2 (women)

Primary outcome

Stroke, systemic embolism, major bleeding, or cardiovascular death at 24 months

Effect

0.5% vs 1.5%, hazard ratio 0.31 (95% CI 0.10–0.94), P=0.03

SINGLE-AF, an open-label Korean trial, randomised 1,803 patients with atrial fibrillation at intermediate stroke risk — a CHA2DS2-VASc score of 1 in men or 2 in women — to a direct oral anticoagulant or no anticoagulation, the group where guidelines give only a weak (class IIa) recommendation. Mean age was 60.4.

At 24 months the composite of stroke, systemic embolism, major bleeding or cardiovascular death occurred in 0.5% on a DOAC versus 1.5% without (hazard ratio 0.31, 95% CI 0.10–0.94; absolute difference 1.0 percentage point). Stroke alone was 0.3% versus 1.1%. Major bleeding was similar between groups and there were no cardiovascular deaths.

The relative reduction is large but the absolute event rates are very low, so the number needed to treat over two years is high. The trial supports offering a DOAC at this borderline threshold, while leaving room to weigh bleeding risk and preference in each patient.

• A DOAC cut the composite to 0.5% from 1.5% over two years in intermediate-risk atrial fibrillation.
• Stroke fell to 0.3% from 1.1%; major bleeding was similar in the two groups.
• The absolute benefit was about 1 percentage point, so many patients are treated for each event prevented.
• This supports, but does not mandate, anticoagulation at CHA2DS2-VASc 1 in men or 2 in women.

Why it matters

It puts randomised evidence under a threshold guidelines only weakly recommend, firming up a common borderline decision.

Don't overread it

A 69% relative reduction rests on 4 versus 13 events; the absolute gain is about 1 point over two years, and the trial was open-label in an East Asian population.

The statistics, in plain English

The hazard ratio of 0.31 looks dramatic, but with event rates near 1% the confidence interval is wide (0.10–0.94) and the real-world benefit per patient is modest — relative and absolute effects diverge here.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.