Epidemiology of anticoagulation for children supported by extracorporeal membrane oxygenation in the United States: A Pediatric Hospital Information System database study

Abstract

IntroductionDue to the risk of thrombosis, nearly all children supported by extracorporeal membrane oxygenation (ECMO) receive systemic anticoagulation. While heparin has traditionally been used, there are reports of increased use of direct thrombin inhibitors. We sought to describe the use of anticoagulation in children supported by ECMO in the United States using a large administrative database.MethodsWe performed a retrospective cohort study of children supported by ECMO within the Pediatric Health Information System (PHIS) database. Pediatric encounters involving ECMO from 2012 to 2020 were identified. Data regarding demographics, diagnoses, anticoagulation, complications, and outcomes were extracted for eligible encounters.ResultsEleven thousand five hundred ninety-five encounters that involved ECMO were identified. Fifty-four percent were male with an age range of 0–17 years and a median (IQR) age of 0 (0–2) years. Unfractionated heparin (UFH) only was used in 94% (95% CI: 93.6–94.5%) of encounters and UFH followed by bivalirudin in 5% (95% CI: 4.3–5.1%) of cases. There was a significant difference in the use of bivalirudin from 2012 to 2020 ( p < 0.001). Differences in anticoagulation regimens were observed between infants and children ( p = 0.004) and between those with and without cardiac indications for ECMO ( p < 0.001). Four percent (95% CI: 4.1–4.8%) of encounters were associated with diagnostic coding for thrombosis and differences in occurrence of thrombosis were observed between different anticoagulant regimens ( p < 0.001).ConclusionsThough the majority of children on ECMO in the United States receive heparin anticoagulation, there is an increase in use of direct thrombin inhibitors. Prospective studies must evaluate the efficacy of different anticoagulants in this patient population.