Ablation for atrial fibrillation beat a sham procedure by 2.6 points, which is nothing

Guidelines recommend catheter ablation for symptom relief in patients with atrial fibrillation [s1]. Until now nobody had reported it against a sham procedure. PVI-SHAM-AF, published in The Lancet on 30 August, did, and the difference between a real ablation and a convincing imitation of one was 2.6 points on a 100-point quality-of-life scale, with a confidence interval crossing zero [s1].

What the trial did

PVI-SHAM-AF was a double-blind, multicentre, randomised trial run at nine sites in Germany and Poland [s1]. Patients aged 18 or older with symptomatic paroxysmal or persistent atrial fibrillation were randomised 2:1 to catheter ablation or a sham procedure, using an automated online randomisation system with variable block sizes stratified by site [s1].

The primary endpoint was the between-group difference in change from baseline to six months in the Atrial Fibrillation Effect on Quality-of-life (AFEQT) summary score [s1]. The pre-specified primary analysis was intention-to-treat, including all randomised patients, with missing data handled by multiple imputation [s1]. The trial is registered as NCT05119231 and twelve-month follow-up is ongoing [s1].

What it found

Between 12 November 2021 and 3 November 2025, 1,199 patients were invited and 262 consented and were randomised: 173 to catheter ablation and 89 to sham [s1]. Median age was 67 (IQR 62–73), and 134 (51%) were female [s1]. Median follow-up was 184 days (IQR 181–191) [s1].

At six months, the mean AFEQT summary score had risen from 61.3 (SD 20.1) to 81.1 (16.6) in the ablation group, and from 59.2 (19.0) to 74.9 (19.5) in the sham group [s1]. The Hodges–Lehmann estimate of the between-group difference in change was 2.6 (95% CI −2.7 to 8.0; p = 0.36) [s1].

The number to sit with is not 2.6. It is 15.7 — the improvement in the sham arm, from 59.2 to 74.9 [s1].

One death occurred in each group, neither considered related to the study procedure [s1]. Serious adverse events adjudicated as related or possibly related to the procedure occurred in ten patients: six in the ablation group and four in the sham group, one of which was an ischaemic stroke in the sham arm [s1]. The authors' interpretation is that catheter ablation did not demonstrate superiority over a sham procedure for improving atrial fibrillation-related quality of life at six months [s1]. The trial was funded by Helios Gesundheit [s1].

What a sham controls for, and what it does not

The trial was accompanied by a linked Comment in the same issue of The Lancet titled "Rethinking success after atrial fibrillation ablation: when sham is more than a placebo" [s2], and by a separate commentary in Clinical Research in Cardiology asking what the trial tells us about ablation and symptoms [s3].

The framing in that first title is the right one. A sham procedure in an invasive trial does not only control for expectation. Both arms in PVI-SHAM-AF received the whole apparatus of a procedure: referral to a specialist centre, a pre-procedural workup, a hospital admission, sedation, femoral access, and a period of structured follow-up during which someone was paying attention to their symptoms and their rhythm control. Patients in an ablation pathway also frequently have medication reviewed, alcohol and weight discussed, and anticoagulation optimised. Any of that can move a symptom score.

So the 15.7-point improvement in the sham arm [s1] is not a measure of gullibility. It is a measure of everything about an ablation pathway that is not the ablation.

What the trial does not overturn

Three limits matter.

It measured symptoms, not events. The endpoint was quality of life at six months [s1]; the trial was not designed or powered for stroke, heart failure hospitalisation, progression to persistent atrial fibrillation or death, and it says nothing about them.

It is small, and the imbalance is deliberate. 262 patients randomised 2:1 gives 89 people in the control arm [s1]. The 95% confidence interval runs from −2.7 to 8.0 [s1], which excludes a large benefit but is compatible with a modest one. "Did not demonstrate superiority" is not "demonstrated equivalence."

And it recruited over four years to reach 262 patients from 1,199 invited [s1]. Patients who consent to a trial in which they might receive a sham ablation are not a random sample of patients offered an ablation, and the direction of that selection is not knowable from the paper.

Why this result was worth generating

Sham-controlled trials of procedures are rare because they are hard to fund, hard to recruit for and ethically demanding. They are also the only design that can separate a procedure's mechanical effect from everything that surrounds it, when the measured outcome is how a patient says they feel — the outcome most sensitive to knowing what was done to you.

This trial does not show that ablation does not work. It shows that at six months, on a symptom scale, most of the improvement patients experience is attributable to something other than the pulmonary vein isolation itself.

What to watch

Twelve-month follow-up is ongoing [s1]. If the between-group gap widens with time — plausible if arrhythmia recurrence separates the arms after the early post-procedural period — the six-month result will read as a timing problem. If it does not, the case for ablation will rest on rhythm and event outcomes rather than on symptoms.

This article describes trial results. It is not medical advice, and nothing here should be used to decide for or against a procedure.

Sources

• Catheter ablation for symptomatic atrial fibrillation (PVI-SHAM-AF): a randomised, double-blind, sham-controlled, multicentre trial, The Lancet, 30 August 2026
• Rethinking success after atrial fibrillation ablation: when sham is more than a placebo, The Lancet, 30 August 2026
• PVI-SHAM-AF: what does this trial tell us about catheter ablation and symptoms in patients with atrial fibrillation?, Clinical Research in Cardiology, 30 August 2026