Perioperative Anticoagulation in Atrial Fibrillation: A Bridge to Nowhere

Authors: Douketis JD et al., N Engl J Med 2015 Aug 27; 373:823

Perioperative “bridging” anticoagulation in warfarin-treated AF patients didn’t prevent thromboembolism but did cause excess major bleeding.

In the U.S., many warfarin-treated patients with atrial fibrillation (AF) who undergo elective surgery receive perioperative bridging anticoagulation, despite a paucity of evidence supporting this practice. In this randomized trial, researchers evaluated >1800 warfarin-treated adults with AF and CHADS₂ (congestive heart failure, hypertension, age ≥75,diabetes, stroke) scores ≥1 who were undergoing elective invasive procedures or surgery.

Patients were bridged with subcutaneous dalteparin (Fragmin) or placebo from 3 days before their procedures until 24 hours before their procedures; bridging was resumed at 1 or 2 days after procedures and was continued until patients achieved full warfarin anticoagulation. Patients with recent arterial or venous thromboembolism (within 3 months) or mechanical heart valves were excluded.

Arterial thromboembolism within 30 days did not differ in the two groups (0.3% and 0.4%). Significantly more patients in the bridging group than in the placebo group experienced major bleeding (3.2% vs. 1.3%; number needed to harm [NNH], 53) and minor bleeding (21% vs. 12%; NNH, 11). Incidences of acute myocardial infarction, venous thromboembolism, and death were similar in the two groups.

Comment Forgoing bridging anticoagulation is safe in many AF patients who require surgery. However, the average CHADS2 score in this study was 2.3, and 90% of patients underwent minor procedures (e.g., endoscopic, dermatologic, or dental procedures, or cardiac catheterization), so whether the results apply to patients with higher CHADS2scores or to those undergoing major surgery is uncertain. Additionally, we don’t know whether the results apply to the subgroup of patients with histories of transient ischemic attacks or stroke. These findings are consistent with American College of Chest Physicians’ guidelines (Chest 2012; 141 (2 Suppl):e326S), which recommend no bridging for patients at low risk for thromboembolism. In fact, those guidelines recommend no interruption of warfarin for some dermatologic, ophthalmologic, and dental procedures that apparently were included in this trial.