Catheter ablation for atrial fibrillation did not improve AF-related quality of life any more than a fake version of the procedure, according to the PVI-SHAM-AF trial published in The Lancet on August 30, 2026 and presented that day at the ESC Congress in Munich. The trial randomized 262 patients at nine hospitals in Germany and Poland. Ablation was clearly better at stopping the abnormal rhythm, but both groups reported feeling about equally better six months later.
Key takeaways
- The trial randomized 262 patients at nine hospitals in Germany and Poland to a real ablation or a sham version of it.
- Quality of life rose sharply in both groups, and the gap between them was 2.6 points on a 100 point scale, which is not a reliable difference.
- Serious procedure-related events were similar: six after ablation, four after the sham.
What the study found
Of 1,199 patients invited between November 2021 and November 2025, 262 agreed and were randomized 2 to 1: 173 to ablation, 89 to sham. Median age was 67, and 51% were female. Median follow-up was 184 days.
The sham was built to be indistinguishable. Patients got conscious sedation for the same length of time in the same setting, and vascular access was obtained, but no catheter was placed, and they were not told which one they got.
The main measure was the AFEQT questionnaire, a summary score of AF-related quality of life. Scores climbed in both arms. The ablation group went from 61.3 to 81.1, a gain of 19.8 points. The sham group went from 59.2 to 74.9, a gain of 15.7 points. The estimated between-group difference was 2.6 points, and the range the data supports runs from 2.7 points worse with ablation to 8.0 points better (95% CI -2.7 to 8.0). The p value was 0.36, meaning a gap this size would turn up by coincidence roughly one time in three.
One death occurred in each group, neither judged related to the procedure. Serious events related or possibly related to the procedure hit six patients after ablation and four after the sham.
Dr. Kumar’s take
The number to sit with is 15.7. That is how much the sham group improved on its own, from 59.2 to 74.9, without a catheter ever going in. People who got sedation, a needle stick in the groin, and nothing else ended up reporting almost as much improvement as the people who got the real thing. Whatever patients feel after an ablation, most of it is not the burn.
This is what happens whenever a procedure is tested against a convincing fake of itself. The mechanism works, and the mechanism is not what the patient is reporting.
One caveat cuts against reading this too hard. Only 262 of the 1,199 patients invited agreed to enroll. Trial investigator Professor Nikolaos Dagres said willingness to accept a sham “may have selected patients with different symptom burden, treatment expectations or preferences.” The people who felt worst likely refused the coin flip and went straight to ablation.
The practical consequence is that the consent conversation has to change. “This will likely fix your rhythm” is supportable. “This will make you feel better” is now a claim the strongest available evidence does not back.
What it means for you
If you have symptomatic AF and an ablation has been proposed, this is not a reason to cancel it. Ablation still did what it is designed to do: it kept the abnormal rhythm away better than the sham did.
It is a reason to ask two questions: what is my AF burden now, and what number is the procedure supposed to move? A procedure aimed at a measurable rhythm target is on firm ground. One sold on the promise that you will feel dramatically better rests on evidence that just failed its test.
It is also worth attacking what drives AF symptoms outside the lab. Professor Rolf Wachter pointed to placebo effect and optimization of therapies as likely explanations for the sham group’s gains. Untreated sleep apnea is one of the biggest levers, and I have written before about whether CPAP keeps AFib from coming back after an ablation. Alcohol and weight belong in that same conversation.
The rhythm and the symptom are not the same target, and treating one does not automatically deliver the other.