Hemodynamic and Echocardiographic Predictors of Mortality in Pediatric Patients on Venoarterial Extracorporeal Membrane Oxygenation: A Multicenter Investigation

Background: Venoarterial extracorporeal membrane oxygenation (ECMO) supports patients with advanced cardiac dysfunction; however, mortality occurs in a significant subset of patients. The authors performed a multicenter, prospective study to determine hemodynamic and echocardiographic predictors of mortality in children placed on ECMO for cardiac support.Methods: Over 8 years, six heart centers prospectively assessed echocardiographic and hemodynamic vari-ables on full and minimum ECMO flow. Sixty-three patients were enrolled, ranging in age from 1 day to 16 years. Hemodynamic measurements included heart rate, vasoactive inotropic score, arteriovenous oxygen difference, pulse pressure, and lactate. Echocardiographic variables included shortening fraction, ejection fraction (EF), right ventricular fractional area change, outflow tract Doppler-derived stroke distance (veloc-ity-time integral [VTI]), and degree of atrioventricular valve regurgitation. Patients were stratified into two groups: those who were able to wean within 48 hours of assessment and survived without ventricular assist devices or orthotopic heart transplantation (successful wean group) and those with unsuccessful weaning. For each patient, variables were compared between full and minimum ECMO flow for each group.Results: Thirty-eight patients (60%) formed the unsuccessful group (two with ventricular assist devices, four with orthotopic heart transplantation, 24 deaths), and 25 constituted the successful wean group. At minimum flow, higher EF (53 & PLUSMN; 16% vs 40 & PLUSMN; 20%, P = .0094), less mitral regurgitation (0.8 & PLUSMN; 0.9 vs 1.4 & PLUSMN; 0.9, P = .0329), and lower central venous pressure (12.0 & PLUSMN; 3.9 vs 14.7 & PLUSMN; 5.4 mm Hg), along with higher VTI (9.0 & PLUSMN; 2.9 vs 6.8 & PLUSMN; 3.7 cm, P = .0154), correlated successful weaning. A longer duration of ECMO (8 vs 5 days, P < .0002) was associated with unsuccessful weaning. Multivariate logistic regression predicted minimum-flow EF and VTI to independently predict successful weaning with cutoff values by receiver operating characteristic analysis of EF > 41% (area under the curve, 0.712; P = .0005) and VTI > 7.9 cm (area under the curve, 0.729; P = .0010). Conclusions: Diminished VTI or EF during ECMO weaning predicts the need for orthotopic heart transplanta-tion or ventricular assist device support or death in children on ECMO for cardiac dysfunction. Increased post -wean central venous pressure or mitral regurgitation along with a prolonged ECMO course also predicted these adverse outcomes. These measurements should be used to help discriminate which patients will require alternative methods of circulatory support for survival. (J Am Soc Echocardiogr 2023;36:233-41.)