Implantation of cardiac rhythm devices without interruption of oral anticoagulation compared with perioperative bridging with low–molecular weight heparin

Abstract

Increasing numbers of patients requiring arrhythmia device implantation are taking warfarin. The common practice of warfarin interruption and perioperative bridging with heparin is associated with a high rate of postoperative hemorrhagic complications. We assessed the safety of device implantation without interruption of warfarin therapy. Three patient groups were studied: Group 1 consisted of 117 consecutive patients on long-term warfarin therapy with significant risk of thromboembolism (atrial fibrillation with CHADS(2) score > or =2, mechanical heart valve, recent venous thromboembolism) who underwent arrhythmia device implantation without interruption of warfarin. Group 2 was 117 patients who served as age- and sex-matched controls matched to procedure type not taking warfarin. Group 3 consisted of 38 similar thromboembolic risk historical control patients who underwent interruption of warfarin therapy and bridging with dalteparin before and 24 hours after surgery. Active fixation leads were used by subclavian or axillary vein puncture, with septal fixation in the ventricle in 56% of patients. Hemorrhagic and thromboembolic complications were assessed at discharge and at 7 and 30 days after surgery. During an 18-month period, 1,562 consecutive adult patients underwent heart rhythm device implantation or replacement. One hundred seventeen of the 447 patients on warfarin were considered high risk and remained on warfarin for their procedure. The mean international normalized ratio in group 1 patients was 2.2 +/- 0.4 (age 79 +/- 11 years, 73 male). Significant hematoma was noted in 9 patients (7.7%), and one required surgical revision (0.9%). Five group 2 patients (control) had significant hematomas (4.3%), none of which required revision (P = .41). In group 3, 9 patients developed significant hematomas (23.7%, P = .012), 3 of whom required reoperation (7.9%, P = .046). There were no deaths, thromboembolic events, cardiac tamponade, or hemothorax in any patient. The only risk factor for hematoma in the warfarin patients was the number of leads implanted. Arrhythmia devices can be implanted safely in patients with high thromboembolic risk without interruption of warfarin. This strategy may be associated with reduced risk of significant pocket hematoma compared with dalteparin bridging.