Objective: To describe the epidemiology, critical care interventions, and mortality of children with pulmonary hypertension receiving extracorporeal membrane oxygenation.
Design: Retrospective analysis of prospectively collected multicenter data.
Setting: Data entered into the Extracorporeal Life Support Organization database between January 2007 and November 2018.
Patients: Pediatric patients between 28 days and 18 years old with a diagnosis of pulmonary hypertension.
Measurements and main results: Six hundred thirty-four extracorporeal membrane oxygenation runs were identified (605 patients). Extracorporeal membrane oxygenation support type was pulmonary (43.1%), cardiac (40.2%), and extracorporeal cardiopulmonary resuscitation (16.7%). The majority of cannulations were venoarterial (80.4%), and 30% had a pre-extracorporeal membrane oxygenation cardiac arrest. Mortality in patients with pulmonary hypertension was 51.3% compared with 44.8% (p = 0.001) in those without pulmonary hypertension. In univariate analyses, significant predictors of mortality included age less than 6 months and greater than 5 years; pre-extracorporeal membrane oxygenation cardiac arrest; pre-extracorporeal membrane oxygenation blood gas with pH less than 7.12, PaCO2 greater than 75, PaO2 less than 35, and arterial oxygen saturation less than 60%; extracorporeal membrane oxygenation duration greater than 280 hours; extracorporeal cardiopulmonary resuscitation; and extracorporeal membrane oxygenation complications including cardiopulmonary resuscitation, inotropic support, myocardial stun, tamponade, pulmonary hemorrhage, intracranial hemorrhage, seizures, other hemorrhage, disseminated intravascular coagulation, renal replacement therapy, mechanical/circuit problem, and metabolic acidosis. A co-diagnosis of pneumonia was associated with significantly lower odds of mortality (odds ratio, 0.5; 95% CI, 0.3-0.8). Prediction models were developed using three sets of variables: 1) pre-extracorporeal membrane oxygenation (age, absence of pneumonia, and pH < 7.12; area under the curve, 0.62); 2) extracorporeal membrane oxygenation related (extracorporeal cardiopulmonary resuscitation, any neurologic complication, pulmonary hemorrhage, renal replacement therapy, and metabolic acidosis; area under the curve, 0.72); and 3) all variables combined (area under the curve, 0.75) (p < 0.001).
Conclusions: Children with pulmonary hypertension who require extracorporeal membrane oxygenation support have a significantly greater odds of mortality compared with those without pulmonary hypertension. Risk factors for mortality include age, absence of pneumonia, pre-extracorporeal membrane oxygenation acidosis, extracorporeal cardiopulmonary resuscitation, pulmonary hemorrhage, neurologic complications, renal replacement therapy, and acidosis while on extracorporeal membrane oxygenation. Identification of those pulmonary hypertension patients requiring extracorporeal membrane oxygenation who are at even higher risk for mortality may inform clinical decision-making and improve prognostic awareness.