- Design
- Retrospective target trial emulation (cloning-censoring-weighting) at 6 Chinese centres, 2019 to 2024
- Population
- 431 patients with atrial fibrillation and prior ischaemic stroke; 207 had early catheter ablation
- Primary outcome
- Major adverse cardiovascular events over 36 months
- Effect
- Absolute risk reduction 11.17% (95% CI 5.81% to 17.36%); 1.54 MACE-free months gained (0.40 to 2.85)
A target trial emulation of 431 patients aged 18 to 85 with atrial fibrillation and a previous ischaemic stroke at six Chinese centres compared early catheter ablation (207 patients, within a one-year grace period) with medical therapy. Median follow-up was 3.0 years.
At 36 months, early ablation was associated with 1.54 more months free of major adverse cardiovascular events (95% CI 0.40 to 2.85) and an absolute risk reduction of 11.17% (95% CI 5.81% to 17.36%). In a matched subset of 250, ablation was linked to better modified Rankin scores (adjusted odds ratio 0.31, 0.19 to 0.51) and, in a post hoc analysis, lower odds of cognitive impairment on the AD8 screen.
The authors rate it Class III evidence and note retrospective design and residual confounding. Patients chosen for ablation are usually fitter and younger, so the gap may partly reflect who was selected.
- Discuss rhythm control options with the cardiology team for suitable patients with AF and a previous stroke
- Do not delay anticoagulation or risk factor control while considering ablation
- Treat the cognitive and functional findings as exploratory
- Randomised data are needed before this changes referral policy
Why it matters
Randomised ablation trials have excluded or underrepresented patients with prior stroke, so observational data fill the gap.
Don't overread it
Selection into ablation probably favours healthier patients, and the cognitive result was post hoc.
The statistics, in plain English
Target trial emulation tries to mimic a randomised trial using records, by cloning patients into each strategy and weighting them. It improves on a naive comparison but cannot adjust for unmeasured differences such as frailty or patient preference.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for neurology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free