Background Direct current cardioversion (DCC) has been shown to be effective for the management of atrial fibrillation (AF) in the emergency department (ED). Pharmacological cardioversion was compared with a strategy including DCC on patients with uncomplicated, recent-onset (<48 h) AF managed in a short observation unit (SOU).

Methods A prospective observational study was undertaken over a period of 13 months in two institutions. A DCC-centred protocol was applied to 171 AF cases in a hospital (DCC-cohort) and pharmacological cardioversion to 151 AF cases in another hospital (P-cohort). Patients remaining in AF after 24 h were admitted. The outcomes were rate of discharge in sinus rhythm, length of stay in the ED-SOU, rate of hospitalisation and complications of treatment. Data collected were analysed according to Student t test and χ² statistics.

Results Discharge in sinus rhythm was achieved in 159/171 cases in the DCC-cohort and 77/151 cases in the P-cohort (93% vs 51%; number needed to treat (NNT) 2.4; 95% CI 2.0 to 3.1, p<0.001), whereas mean length of stay was 7+7 h in the DCC-cohort and 9+6 h in the P-cohort (p=0.43). Eleven cases from the DCC-cohort and 67 from the P-cohort were admitted (admission rate 6% vs 44%; NNT 2.6; 95% CI 2.2 to 3.5, p<0.001). Three short-term complications occurred in the DCC-cohort and five in the P-cohort (2% vs 3%, p=0.59). Two strokes were registered in the DCC-cohort during 6-month follow-up (p undefined).

Conclusions Electrical cardioversion of recent-onset AF in the SOU is safe, effective and reduces hospitalisations. Further studies are needed to identify the most cost-effective strategy for the management of AF patients in emergency settings.