Abstract
Introduction: This study aims to describe laboratory and clinical factors associated with thrombotic events during prolonged pediatric extracorporeal membrane oxygenation. Methods: A secondary analysis of a multi-center prospective study performed between 2012 and 2014. Patients under the age of 19 years that received extracorporeal membrane oxygenation for at least 4 days of therapy were included (n = 385). Univariable analysis and binomial regression were performed to evaluate predictive factors of single and multiple thrombotic events. A posteriori scoring tool was created to categorize thrombotic event severity. Results: Over 39% of children receiving prolonged ECMO experienced a thrombotic event (TE). Binomial regression demonstrated an association between higher transfused platelet volume (mL/kg) (OR 1.04, CI: 95% 1.01–1.06, p = 0.003), Anti-Xa (OR 5.38, CI: 95% 1.22–23.8, p = 0.026) and aPTT (OR 1.01, CI: 95% 1.00–1.02, p = 0.032) the day prior to TE. Patients experiencing multiple TEs were associated with higher platelet transfusion volume (mL/kg) (OR 1.08, CI: 95% 1.05–1.12, p =< 0.001), antithrombin III (OR 1.03, CI: 95% 1.01–1.04, p = 0.001) and aPTT (OR 1.02, CI: 95% 1.01–1.03, p = 0.009). Patients experiencing multiple thrombotic events had a higher risk of 28-day mortality based on a cumulative clot severity score >4 (OR 2.37 (CI: 95% 1.32–4.24). Conclusions: Current lab tests show limited sensitivity to predict these events the day prior in a vulnerable patient group, leading to potential ECMO circuit failures. Patients with multiple thrombotic events during ECMO therapy face increased mortality risks, highlighting the need for dynamic reporting tools like clot severity scores and detailed documentation of interventions to enhance understanding and improve outcomes. © The Author(s) 2024.