Introduction: In atrial fibrillation (AF), a strategy of rhythm control based on cardioversion, by restoring sinus rhythm, may reduce the risk of stroke/systemic embolization (SE) and improve quality of life. However, randomized trials conducted so far have failed to show a benefit of cardioversion on hard endpoints. This study aims to investigate the prevalence of cardioversion and its association with clinical outcomes in patients from GARFIELD-AF registry.
Methods: GARFIELD-AF is a prospective, global registry of patients with recent-onset (<6 weeks) non-valvular AF (NVAF). Patients were enrolled in 32 countries between 2010 and 2016, patients with paroxysmal AF were excluded from the analysis. Comparisons were made between those receiving cardioversion at baseline and patients who had no cardioversion. Clinical endpoints, evaluated over 1 year, were: all-cause mortality, stroke or systemic embolism (SE) and major bleeding. An adjusted Cox proportional hazard model was utilized.
Results: The study cohort consisted of 23,919 patients; 2856 received cardioversion (11.9%). Patients who were treated with cardioversion were younger (65.5±11.7 years vs 70.6±11.3 years; p≤0.001), had a shorter time since AF diagnosis (1.7±1.6 weeks vs 2.0±1.7 weeks; p≤0.001), and were more often treated in a cardiology setting (73.9% vs 63.5%; p≤0.001). Event rates per 100 person years (95% CI) for all-cause mortality were 3.26% (2.65–4.01) (n=89) for cardioversion vs 5.40% (5.09–5.76) (n=1072) for no cardioversion, (Fig.1); stroke/SE rate 0.96% (0.65–1.40) (n=26) for cardioversion and 1.48% (1.32–1.65) (n=291) no cardioversion and major bleed 0.66% (0.42–1.05) (n=18) cardioversion vs 0.93% (0.80–1.07) (n=183) no cardioversion. Adjusted hazard ratios (95% CI) for cardioversion were 0.69 (0.53–0.90) for all-cause mortality, 0.92 (0.58–1.44) for stroke/SE and 0.82 (0.46–1.47) for major bleed.
Conclusion: Data from GARFIELD-AF show that cardioversion is used in a minority of patients with recent onset, non-paroxysmal NVAF. Patients who receive cardioversion have a lower risk of major events than those who did not. Our results, after adjustment, support the findings from randomized clinical trials, which suggest that there is no additional risk associated with cardioversion.