Association of Hospital Structure and Complications With Mortality After Pediatric Extracorporeal Membrane Oxygenation.

OBJECTIVES: Extracorporeal membrane oxygenation is increasingly utilized to provide cardiopulmonary support to critically ill children. Although life-saving in many instances, extracorporeal membrane oxygenation support is associated with considerable morbidity and mortality. This study evaluates the effect of extracorporeal membrane oxygenation complications and extracorporeal membrane oxygenation hospital characteristics on mortality in neonates and children supported with extracorporeal membrane oxygenation.

DESIGN: Retrospective analysis of administrative data.

SETTING: Data from 31 U.S. states included in 2012 Healthcare Cost and Utilization Project Kids' Inpatient Database.

PATIENTS: Children treated with extracorporeal membrane oxygenation.

INTERVENTIONS: None.

MEASUREMENTS AND MAIN RESULTS: Study subject were identified using International Classification of Diseases, 9th Edition Clinical Modification code 39.65 and classified into six diagnostic categories: 1) cardiac surgery, 2) non-surgical heart disease, 3) congenital diaphragmatic hernia, 4) neonatal respiratory failure, 5) pediatric respiratory failure, and 6) sepsis. Demographics, hospital characteristics, and outcome information were used in a multivariate logistic regression analysis to determine factors associated with mortality. We identified 1,465 children treated with extracorporeal membrane oxygenation. Overall mortality was 40% (591/1,465). Mortality was independently associated with diagnosis (heart disease: odds ratio, 1.7; p = 0.01; congenital diaphragmatic hernia: odds ratio, 5.1; p < 0.001; and sepsis odds ratio: 2.4; p = 0.003 compared with neonatal respiratory failure) time from hospital admission to extracorporeal membrane oxygenation of more than 10 days (odds ratio, 4.5; p < 0.001) and extracorporeal membrane oxygenation complications (renal [odds ratio: 5; p < 0.001] and neurologic [odds ratio, 1.4; p = 0.03] injury). In addition, hospitals with bed size less than 400 had higher mortality (odds ratio, 1.4; p = 0.02). In patients with any extracorporeal membrane oxygenation complication, probability of mortality was lower for extracorporeal membrane oxygenation patients in larger hospitals, 38% (95% CI, 37-39) versus 44% (95% CI, 43-46) with p value of less than 0.001.

CONCLUSIONS: Extracorporeal membrane oxygenation mortality was significantly associated with patient diagnosis, time to extracorporeal membrane oxygenation initiation, extracorporeal membrane oxygenation complications, and extracorporeal membrane oxygenation hospital bed size. Improved survival in larger hospitals supports centralization of extracorporeal membrane oxygenation services to larger centers.