763

Abstract

Introduction: To prevent thrombus formation within the extracorporeal membrane oxygenation (ECMO) circuit, systemic anticoagulation with unfractionated heparin (UFH) is recommended. Patients undergoing ECMO may have many causes of thrombocytopenia including but not limited to heparin-induced thrombocytopenia (HIT), critical illness, and ECMO itself. Methods: An IRB-approved, retrospective chart review was conducted from January 2011 through June 2013 in all adult patients on ECMO without a prior history of HIT. Patients were included if they received ECMO for at least five days and UFH within 24 hours of ECMO support. The indication and duration of ECMO, total heparin exposure, diagnostic testing for HIT, and platelet counts were recorded for each patient. Data are presented as median (IQR). A p<0.05 was considered significant. Results: There were 119 patients who met inclusion criteria. Twenty-three patients (19%) had an enzyme-linked immunosorbent assay (ELISA) test for HIT evaluation. Baseline platelet count in patients tested for HIT was 215 x 10^9/L (98, 277) compared to 178 x 10^9/L (133, 259) in those not tested; p=0.85. Patients evaluated for HIT had a significantly lower platelet count within the first three days of ECMO, 69 x 10^9/L (22, 126) versus 87.5 x 10^9/L (63, 149); p=0.04. The lowest platelet count on the day of ELISA testing was 43 x 10^9/L (26, 73), representing up to 71% reduction from baseline. One patient had clinically diagnosed HIT without thrombotic complications, supported by a positive serotonin release assay. There was no difference between the groups in total exposure to heparin, duration of ECMO, percent of patients investigated for clot, or presence of new thrombus; p=0.95, p=0.10, p=0.10, and p=0.42, respectively. Conclusions: The evaluation of HIT occurred in a small percentage of patients receiving ECMO and UFH. The patients that had ELISA testing exhibited lower platelet counts with a similar duration of ECMO, baseline platelet count, and total heparin exposure. Clinically diagnosed HIT was rarely detected in this cohort.