The following are 10 key points from this review of bridging anticoagulation:

1. Each year, 15-20% of patients on chronic oral anticoagulants undergo an invasive procedure or surgery requiring temporary interruption of the oral anticoagulant.
2. Most guidelines recommend three important principles for managing periprocedural anticoagulation.
   • Oral anticoagulants should not be interrupted for low bleeding risk procedures.
   • Patients at highest risk for thromboembolism without excessive bleeding risk should consider bridging. Conversely, those at low thromboembolism risk should not be bridged.
   • Intermediate risk cases should be managed by individuals considering patient- and procedure-specific risks for bleeding and thromboembolism.
3. An important first step is to confirm the indication for oral anticoagulation. In some cases, oral anticoagulants may no longer be needed. In other cases, a recent thromboembolism (such as an acute deep venous thrombosis) suggests that interruption of anticoagulation should be avoided or postponed.
4. Low bleeding risk procedures, for which anticoagulation should not be interrupted, include most dermatologic surgeries, orthopedic surgeries, pacemaker and defibrillator implantation, endovascular interventions, cataract surgery, and dental procedures.
5. Rates of periprocedural thromboembolism and bleeding vary by indication and choice of anticoagulant. In general, the rate of thromboembolism without bridging anticoagulation is quite low (one estimate at 0.53%). Even patients with mechanical heart values have low rates of thromboembolism in recent studies. Patients with left ventricular assist devices are commonly treated with oral anticoagulants. However, bleeding is much more common than clotting in these patients.
6. Contemporary bridging practices are quite variable and often without regard for thromboembolism risk. Use of bridging "just to be safe" is producing preventable adverse bleeding events without thrombosis prevention.
7. The recently published BRIDGE trial (N Engl J Med 2015;373:823-33) found that bridging in atrial fibrillation patients did not prevent thromboembolism (0.3-0.4%), but was associated with increased major (3.2% vs. 1.3%, p = 0.005) and minor bleeding (20.9% vs. 12%, p = 0.001). However, this trial did not enroll many atrial fibrillation patients at highest risk of thromboembolism (CHADS₂ of 5 or 6) or with other indications for oral anticoagulation (e.g., venous thromboembolism or mechanical heart valves).
8. The ongoing PERIOP2 (NCT00432796) study is exploring the role of bridging anticoagulation for patients with mechanical heart values and atrial fibrillation.
9. Clinicians can consider the use of the BleedMAP (prior bleeding, mechanical mitral valve, active cancer, and low platelets) score to estimate bleeding (and thromboembolism) rates.
10. Direct oral anticoagulants (dabigatran, rivaroxaban, apixaban, and edoxaban) do not require heparin bridging given their shorter half-life. They may also provide an oral alternative to low molecular weight heparin; however, few studies with this management strategy have been reported.

Keywords: Arrhythmias, Cardiac, Anticoagulants, Atrial Fibrillation, Blood Platelets, Cardiac Surgical Procedures, Defibrillators, Dermatologic Surgical Procedures, Heart-Assist Devices, Heart Failure, Heparin, Low-Molecular-Weight, Orthopedic Procedures, Thromboembolism, Venous Thromboembolism, Venous Thrombosis

< Back to Listings