Cardioversion frequency is a strong independent predictor of stroke and systemic embolism (SSE) and all-cause mortality in nonanticoagulated patients with paroxysmal atrial fibrillation (AF), according to a study published in the Annals of Medicine.

Researchers conducted a retrospective analysis evaluating the risk for thromboembolic complications based on the frequency of cardioversion for acute AF in patients not taking oral anticoagulant (OAC) therapy. Data were collected from the FinCV study database.

Eligible participants were aged older than 18 years and had a diagnosis of AF and cardioversion for acute (<48 hours) AF episode from 3 Finnish hospitals between 2003 and 2010. The analysis included 2074 patients who were not using long-term (>4 weeks) OAC therapy after cardioversion and had long-term follow-up data available.

Continue Reading

Data were collected until initiation of long-term OAC therapy, AF was considered permanent, or death. Follow-up ended on the date of the last entry in a patient’s record with no OAC therapy in the medication if no endpoint events occurred.

The participants were classified into 2 groups based on the frequency of cardioversions during the follow-up, which were the high cardioversion frequency (mean interval between cardioversions ≤12 months) and low cardioversion frequency (interval >12 months) groups. A total of 663 participants are in the high cardioversion frequency group (median age, 66 [IQR, 58-73] years; 39.5% women), and 1411 are in the low cardioversion frequency group (median age, 59 [IQR, 50-68] years; 32.5% women).

The median cardioversion interval is 40.8 (IQR, 21.9-88.7) months for patients in the low cardioversion frequency group and 5.3 (IQR, 2.9-8.3) months for those in the high frequency group.

A total of 5.2% of patients had SSEs (97 ischemic strokes and 10 peripheral embolisms) during the median follow-up of 5.4 years (IQR, 1.9-10.6 years) when they were not using OAC therapy. Twenty-seven SSEs occurred in the high cardioversion frequency group and 80 SSEs in the low cardioversion frequency group, with linearized SSE rates per 100 patient-years of 1.82 and 0.67, respectively.

Cardioversion frequency is an independent predictor for SSE (hazard ratio [HR], 2.87; 95% CI, 1.47-5.64; P =.002) at 3 years after adjustment for CHA2DS2-VASc score, according to Cox regression analysis. Competing risk analysis shows that cardioversion frequency (subdistribution HR [sHR], 2.70; 95% CI, 1.38-5.31; P =.004) and increasing CHA2DS2-VASc score (sHR, 1.87; 95% CI, 1.66-2.11; P <.001) are independent predictors for SSE at 3 years.

During follow-up, 170 deaths occurred, 7.0% of which were within 30 days of SSE. The all-cause mortality rate per 100 patient-years is 9.0 in the high cardioversion frequency group and 1.4 in the low frequency group. Cardioversion frequency (HR, 2.54; 95% CI, 1.73-3.74; P <.001) and CHA2DS2-VASc score (HR, 1.50; 95% CI, 1.39-1.61; P <.001) are confirmed as independent predictors for mortality in Cox regression analysis.

Among the limitations in this retrospective study, the findings only apply for patients with AF who visit an emergency department owing to acute symptomatic AF episodes and are treated with cardioversion. Follow-up is also short in patients with frequent cardioversions.

“This study is the first to describe a significant association between symptomatic paroxysmal AF burden and clinical outcome in patients without cardiovascular implantable electronic devices,” the study authors wrote. “Both adjusted Cox regression and competing risk analysis revealed [cardioversion] frequency to be a strong independent predictor of both SSE and all-cause mortality during 3-year follow-up. Most importantly, patients at intermediate SSE risk based on CHA2DS2-VASc score 1 and infrequent symptomatic AF attacks had a very low SSE risk, which is a clinically useful finding when considering OAC therapy for these patients.”

Disclosure: Some of the study authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of authors’ disclosures. 


Jaakkola S, Kiviniemi TO, Jaakkola J, et al. Frequency of cardioversions as an additional risk factor for stroke in atrial fibrillation – the FinCV-4 study. Ann Med. Published online May 20, 2022. doi: 10.1080/07853890.2022.2077430