BackgroundThere is a paucity of data comparing vitamin K antagonists (VKAs) to direct oral anticoagulants (DOACs) at the time of cardiac implantable electronic device (CIED) surgery. Furthermore, the best management of DOACs (interruption vs continuation) is yet to be determined. ObjectivesThis study aimed to compare the incidence of device-related bleeds and thrombotic events based on anticoagulant type (DOAC vs VKA) and regimen (interrupted vs uninterrupted). MethodsThis was an observational multicenter study. We included patients on chronic oral anticoagulation undergoing CIED surgery. Patients were matched using propensity scoring. ResultsWe included 1,975 patients (age 73.8 ± 12.4 years). Among 1,326 patients on DOAC, this was interrupted presurgery in 78.2% (n = 1,039) and continued in 21.8% (n = 287). There were 649 patients on continued VKA. The matched population included 861 patients. The rate of any major bleeding was higher with continued DOAC (5.2%) compared to interrupted DOAC (1.7%) and continued VKA (2.1%) (P = 0.03). The rate of perioperative thromboembolism was 1.4% with interrupted DOAC, whereas no thromboembolic events occurred with DOAC or VKA continuation (P = 0.04). The use of dual antiplatelet therapy, DOAC continuation, and male sex were independent predictors of major bleeding on a multivariable analysis. ConclusionsIn this large real-world cohort, a continued DOAC strategy was associated with a higher bleeding risk compared to DOAC interruption or VKA continuation in patients undergoing CIED surgery. However, DOAC interruption was associated with increased thromboembolic risk. Concomitant dual antiplatelet therapy should be avoided whenever clinically possible. A bespoke approach is necessary, with a strategy of minimal DOAC interruption likely to represent the best compromise.