ABSTRACT
Purpose: Anticoagulation during extracorporeal membrane oxygenation (ECMO) for COVID-19 can be carried out by direct or indirect thrombin inhibition. The former agent obviates monitoring of antithrombin III but differences in outcomes with either approach are uncertain. Methods: A retrospective, multi-center study was conducted. Adult patients (≥18 years old) placed on ECMO for severe respiratory or circulatory failure due to COVID-19 between March 1, 2020, to April 30, 2021, in the United States were included. Patient were divided in 2 groups based on the utilized anticoagulation agent during ECMO support: 1) direct thrombin inhibitor (DTi, e.g. bivalirudin and argatroban) and 2) indirect thrombin inhibitor (IDTi, e.g. unfractionated heparin). The primary outcome was in-hospital mortality after ECMO placement assessed by a time-to-event analysis. Results: Overall, 455 patients from 17 centers were placed on ECMO, of whom 44 were excluded due to no reported anticoagulation agent. DTi was used in 160 (age: 47±11, 28% female) cases and 251 patients received IDTi (age 47±12, 29% female). At 90 days, in-hospital mortality was 50% (DTi) and 61% (IDTi), p=0.08, (Figure). After adjustment for clinical covariates, the likelihood of in-hospital mortality was similar with DTi (aHR: 0.79, 95%CI 0.57-1.10, p=0.16) compared to IDTi. Noted prevalence of deep vein thrombosis (DTi 14%, IDHi 12%), ischemic stroke (DTi 2%, IDHi 3%), intracranial hemorrhage (DTi 11%, IDHi 10%) and bleeding requiring transfusion (DTi 71%, IDHi 83%) was comparable between groups. Conclusion: Anticoagulants that directly or indirectly inhibit thrombin are associated with similar outcomes during ECMO for COVID-19.