Outcomes of uninterrupted vs interrupted Periprocedural direct oral Anticoagulants in atrial Fibrillation ablation: A meta-analysis

Background Studies indicate that uninterrupted anticoagulation (UA) is superior to interrupted anticoagulation (IA) in the periprocedural period during catheter ablation of atrial fibrillation. Still IA is followed in many centers considering the bleeding risk. This meta-analysis compares interrupted and uninterrupted direct oral anticoagulation during catheter ablation of atrial fibrillation. Methods A systematic search into PubMed, EMBASE, and the Cochrane databases was performed and five studies were selected that directly compared IA vs UA before ablation and reported procedural outcomes, embolic, and bleeding events. The primary outcome of the study was major adverse cerebro-cardiovascular events. Results The meta-analysis included 840 patients with UA and 938 patients with IA. Median follow-up was 30 days. Activated clotting time (ACT) before first heparin bolus was significantly longer with UA (P = .006), whereas mean ACT was similar between the two groups (P = .19). Total heparin dose needed was significantly higher with IA (mean, -1.61; 95% CI, -2.67 to -0.55; P = .003). Mean procedure time did not vary between groups (P = .81). Overall complication rates were low, with similar major adverse cerebro-cardiovascular event (P = .40) and total bleeding (P = .55) rates between groups. Silent cerebral events (SCEs) were significantly more frequent with IA (log odds ratio, -0.90; 95% CI, -1.59 to -0.22; P I-2, 33%). Rates of major bleeding, minor bleeding, pericardial effusion, cardiac tamponade, and puncture complications were similar between groups. Conclusions UA during atrial fibrillation ablation has similar bleeding event rates, procedural times, and mean ACTs as IA, with fewer SCEs.

Automated summary

This meta-analysis compared interrupted and uninterrupted direct oral anticoagulation during catheter ablation of atrial fibrillation and found that uninterrupted anticoagulation was similar to interrupted anticoagulation in terms of bleeding event rates, procedural times, and mean activated clotting times, but had fewer silent cerebral events.