Ablation for atrial fibrillation was no better than a sham on the outcome it is prescribed for

PVI-SHAM-AF is the trial the field has needed and largely avoided. Nine sites in Germany and Poland randomised 262 adults with symptomatic paroxysmal or persistent atrial fibrillation, 2:1, to catheter ablation or a sham procedure, with patients and assessors blinded. The primary endpoint was change in the AFEQT quality-of-life score at six months — the outcome for which guidelines recommend ablation in the first place.

Both arms improved markedly. The ablation group went from a mean score of 61.3 to 81.1; the sham group went from 59.2 to 74.9. The between-group difference was 2.6 points, with a confidence interval from -2.7 to 8.0 and a P value of 0.36. One death occurred in each arm, neither procedure-related. Procedure-related serious adverse events occurred in six ablation patients and four sham patients, including one stroke — in the sham arm.

The finding is not that ablation does nothing. Patients who had it improved by twenty points. The finding is that patients who had a convincing sham improved by nearly sixteen, so most of the benefit attributed to pulmonary vein isolation is attributable to undergoing a procedure, being observed, and expecting to be better. That is a hard result to carry into a consent conversation, and it is the one the evidence now supports.

• Quality of life, not stroke or mortality — this trial says nothing about those
• Only 262 of 1,199 invited patients consented, so the participants are self-selected
• Six months of follow-up; the twelve-month data are still to come
• The sham arm's improvement is the finding, not the ablation arm's failure
• Symptom-driven ablation now needs a different conversation about expected benefit

The statistics, in plain English

A difference of 2.6 with a confidence interval from -2.7 to 8.0 means the data are compatible with ablation being slightly worse than sham and with it being meaningfully better; the study cannot distinguish them. What makes this more than an underpowered null is the size of the sham response — a 15.7-point improvement in patients who received no ablation. A trial without a sham arm would have recorded ablation's 19.8-point gain and called it efficacy.

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.