Isolated atrial flutter carries about 40% less stroke risk than fibrillation

Design

nationwide registry-linkage cohort study with ECG-based rhythm classification, Finland 2007–2018

Population

40,986 patients with new-onset atrial fibrillation or flutter — 30,261 fibrillation-only, 2,409 flutter-only, 8,316 both

Primary outcome

ischaemic stroke rate and adjusted incidence rate ratio

Effect

flutter-only IRR 0.60 (95% CI 0.49–0.74); both IRR 1.11 (95% CI 1.02–1.20) versus fibrillation-only

Clinical practice treats atrial flutter as atrial fibrillation for anticoagulation purposes, on the basis that the two coexist and the thromboembolic mechanism is assumed to be similar. This Finnish registry study tested the assumption with ECG documentation rather than diagnostic codes: 40,986 patients with new-onset atrial fibrillation or flutter between 2007 and 2018, classified from 532,041 digitally recorded ECGs into flutter-only, fibrillation-only, or both.

Ischaemic stroke rates were 1.1 per 100 patient-years in flutter-only, 1.9 in fibrillation-only and 2.1 in those with both. The adjusted incidence rate ratio for flutter-only was 0.60 (95% CI 0.49–0.74) against fibrillation-only, and 1.11 (95% CI 1.02–1.20) for those with both. The pattern held in analyses restricted to time off anticoagulation and before ablation, and across CHA2DS2-VA categories.

The flutter-only group was small — 2,409 patients — and the crucial caveat is clinical rather than statistical. Many patients recorded as flutter-only have undetected fibrillation; a single ECG classifies a rhythm at one moment, not a patient. This lowers the estimated risk in isolated flutter; it does not license stopping anticoagulation in someone with flutter and a CHA2DS2-VA score that would otherwise mandate it.

• Do not withhold anticoagulation in flutter on the strength of a rate ratio from an observational cohort.
• Where flutter is the documented rhythm and ablation is planned, this supports discussing the residual risk honestly rather than assuming it equals fibrillation.
• Extended rhythm monitoring before calling a patient flutter-only changes the question — coexisting fibrillation is common and carries the higher risk.
• Record which arrhythmia the ECG actually shows; registries and clinics alike blur the two.

Why it matters

It challenges the reflex that flutter and fibrillation are one condition for stroke prevention, without yet giving you licence to act differently.

Don't overread it

This was a registry cohort — it cannot show that withholding anticoagulation in flutter is safe.

The statistics, in plain English

An adjusted incidence rate ratio of 0.60 means roughly 40% fewer strokes per year of follow-up, with the interval 0.49 to 0.74 comfortably excluding no difference. But adjustment can only handle the variables recorded, and the biggest unmeasured variable here is whether the flutter-only patients really had no fibrillation. Mean follow-up was 1.2 years, which is short for accumulating stroke events, and 77.6% of the whole cohort started anticoagulation — so this is largely the risk that remains on treatment.

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.