Periprocedural anticoagulation management of patients with nonvalvular atrial fibrillation

OBJECTIVE: To estimate the 3-month cumulative incidence of thromboembolism (TE), bleeding, and death among consecutive patients with nonvalvular atrial fibrillation (AF) who were receiving long-term anticoagulation therapy and were referred to the Thrombophilia Center at Mayo Clinic for periprocedural anticoagulation management. PATIENTS AND METHODS: In a prospective cohort study of consecutive patients receiving long-term anticoagulation therapy who were referred to the Thrombophilia Center for periprocedural anticoagulation management over the 7-year period, January 1, 1997, to December 31, 2003, 345 patients with nonvalvular AF were eligible for inclusion. Warfarin was stopped 4 to 5 days before and was restarted after surgery as soon as hemostasis was assured. The decision to provide bridging therapy with heparin was individualized and based on the estimated risks of TE and bleeding. RESULTS: The 345 patients with AF (mean ± SD age, 74±9 years; 33% women) underwent 386 procedures. Warfarin administration was not interrupted for 44 procedures. Periprocedural heparin was provided for 204 procedures. Patients receiving heparin were more likely to have prior TE (43% vs 24%; P<.001) and a higher CHADS₂ (congestive heart failure, hypertension, age, diabetes, stroke) score (2.2 vs 1.9; P=.06). Four patients had 6 episodes of TE (3 strokes and 3 acute coronary episodes; TE rate, 1.1%; 95% confidence interval, 0.0%-2.1%). Nine patients had 10 major bleeding events (major bleeding rate, 2.7%; 95% confidence interval, 1.0%-4.4%). There were no deaths. Neither bleeding nor TE rates differed by anticoagulant management strategy. CONCLUSION: The 3-month cumulative incidence of TE and bleeding among patients with AF in whom anticoagulation was temporarily interrupted for an invasive procedure was low and was not significantly influenced by bridging therapy.