What are the risks of stopping the NAOC

The short version: stopping a NOAC after successful cardioversion in a 69-year-old man in sinus rhythm is a reasonable option, but the evidence is observational and the decision depends on your stroke risk score.

The best evidence you have comes from a 2010 multicenter study of 3,355 patients who had successful atrial fibrillation ablation [2]. Among the 2,692 who stopped oral anticoagulation 3 to 6 months after the procedure, the annual ischemic stroke rate was 0.03 per 100 patient-years. That is about 3 strokes per 10,000 people per year. Among the 663 who stayed on anticoagulation, the rate was 0.23 per 100 patient-years, or about 23 per 10,000 per year. The difference was statistically significant (p=0.049). Major hemorrhages were far more common in the group that stayed on anticoagulation: 0.98 per 100 patient-years versus 0.02 per 100 patient-years, a 49-fold difference (p<0.0001) [2].

A 2020 single-center study of 796 patients who stopped anticoagulation 3 months after ablation tells a more cautious story [4]. Among patients who had atrial fibrillation recurrence, the thromboembolism rate was 1.62 per 100 patient-years. Among those who stayed in sinus rhythm, it was 0.33 per 100 patient-years. The adjusted hazard ratio for recurrence was 4.49 (95% CI 1.38 to 14.59) [4]. This means the risk of stopping depends heavily on whether the atrial fibrillation actually stays gone.

The 2010 study also looked at patients with CHADS2 scores of 2 or higher, meaning moderate to high stroke risk. Among the 347 patients in that subgroup who stopped anticoagulation, zero ischemic strokes occurred during follow-up [2]. This is the most important single number in the evidence. It suggests that even patients with higher risk scores may be safe to stop after successful ablation, but the sample is small and the study is observational, not randomized.

The evidence has real limits. Both studies are retrospective and observational, not randomized controlled trials [2, 4]. The 2010 study's authors explicitly say their conclusion needs confirmation from prospective randomized trials [2]. The 2020 study is single-center with only 12 total thromboembolic events, making the risk estimates imprecise [4]. Neither study tested NOACs specifically; they studied warfarin and other oral anticoagulants. The mechanism is the same, but the absolute numbers may differ.

My call: stopping a NOAC after successful cardioversion with sustained sinus rhythm is supported by observational evidence showing very low stroke risk and dramatically lower bleeding risk, even in patients with moderate-to-high CHADS2 scores. But the evidence is not from randomized trials, the studies followed patients after ablation rather than cardioversion, and the risk rises sharply if atrial fibrillation returns. Confidence: moderate.