Comparing the delay with different anticoagulants before elective electrical cardioversion for atrial fibrillation/flutter

AIMS: To assess the impact of the introduction of direct oral anticoagulants upon the outcomes from elective electrical cardioversion for atrial fibrillation. METHODS: This is a retrospective comparison of delay to elective cardioversion with different anticoagulants. The data was gathered from a large regional hospital from January 2013 to September 2017. There were 3 measured outcomes: 1) the time in weeks from referral to the date of attempted electrical cardioversion; 2) the proportion of patients who were successfully cardioverted; and 3) the proportion of patients who remained in sinus rhythm by the 12 week follow-up. Time-to-cardioversion was non-parametrically distributed so was analysed with Kruskal-Wallis testing and Mann-Whitney-U testing. Maintenance of sinus rhythm was analysed using z-testing. RESULTS: 1,374 patients were submitted to cardioversion. The referrals for cardioversion were either from primary care or from cardiologists. At the time of cardioversion, 789 cases were anticoagulated on warfarin (W), 215 on apixaban (A) and 370 on rivaroxaban (R). All 3 cohorts were initially compared independently using Kruskal-Wallis testing. This demonstrated a significant difference in the delay (measured in weeks) between the A and W group (A = 7, W = 9, P<0.00001); the R and W group (R = 7, W = 9, P<0.00001) and no difference between R and A (A = 7, R = 7, P = 0.92). As there was no difference between the A and R groups, they were combined to form the AR group. The AR group was compared to the W group using Mann-Whitney-U testing which demonstrated a significant delay between the groups (AR = 7, W = 9, P<0.00001). Excluding patients with prior or unknown attempts of cardioversion (n = 791), the W patients (n = 152) were less successful in achieving sinus rhythm at cardioversion than the AR (n = 431) group (W = 95% vs AR = 99% P = 0.04). However at 12 weeks, incidence of sinus rhythm was significantly different (W = 40% vs AR = 49% P = 0.049). These groups were compared by z testing. At 12 weeks' follow-up there was no statistical difference in rate of adverse consequences between the AR group and the W group, but the rate of adverse consequences was too low to draw further conclusions. CONCLUSION: DOACs appear to significantly shorten the latency between the decision to cardiovert and the cardioversion procedure by at least 2 weeks compared to warfarin in a real-world setting. In this study, patients who had not previously been cardioverted who were anticoagulated with warfarin had a significantly lower probability of conversion to sinus rhythm and a significantly lower probability to remain in sinus rhythm at the 12 week follow-up compared to the combined apixaban and rivaroxaban group.