A Comparison of Anticoagulation Strategies in Veno-venous Extracorporeal Membrane Oxygenation

Bleeding remains a major source of morbidity associated with veno-venous extracorporeal membrane oxygenation (VV-ECMO). Moreover, there remains significant controversy, and a paucity of data regarding the ideal anticoagulation strategy for VV-ECMO patients. All patients undergoing isolated, peripheral VV-ECMO between January 2009 and December 2014 at our institution were retrospectively reviewed. Patients (n = 123) were stratified into one of three sequential eras of anticoagulation strategies: activated clotting time (ACT: 160-180 seconds, n = 53), high-partial thromboplastin time (H-PTT: 60-80 seconds, n = 25), and low-PTT (L-PTT: 45-55 seconds, n = 25) with high-flow (>4 L/min). Pre-ECMO APACHE II scores, SOFA scores, and Murray scores were not significantly different between the groups. Patients in the L-PTT group required less red blood cell units on ECMO than the ACT or H-PTT group (2.1 vs. 1.3 vs. 0.9; p < 0.001) and patients in the H-PTT and L-PTT group required less fresh frozen plasma than the ACT group (0.33 vs. 0 vs. 0; p = 0.006). Overall, major bleeding events were significantly lower in the L-PTT group than in the ACT and H-PTT groups. There was no difference in thrombotic events. In this single-institution experience, a L-PTT, high-flow strategy on VV-ECMO was associated with fewer bleeding and no difference in thrombotic events than an ACT or H-PTT strategy.