Utilization and outcomes of neonatal cardiac extracorporeal life support: 1996-2000

Abstract

Objectives: Extracorporeal life support for neonatal respiratory failure has decreased, but utilization and outcome of cardiac extracorporeal life support are not well characterized. Among neonates born 1996-2000, our objects were to evaluate changes in utilization and outcome of cardiac extracorporeal life support and characterize correlates of survival.

Design: Retrospective analysis of Extracorporeal Life Support Organization Registry data.

Setting: Intensive care units participating in the ELSO registry.

Patients: Patients placed on extracorporeal life support for center-specified "cardiac support" at </=30 days of age from 1996 to 2000. Patients with hypoplastic left heart syndrome were also analyzed separately.

Interventions: None.

Measurements and main results: Patient characteristics and correlates of survival to discharge or transfer were analyzed by chi-square, Student's t-test, and logistic regression analysis. Neonates placed on cardiac extracorporeal life support increased from 112 in 1996 to 200 in 2000 (total n = 740). Overall survival was 34.2%: 28% for hypoplastic left heart syndrome and 35.4% for nonhypoplastic left heart syndrome. For the overall group, no significant correlations were found between survival and year on extracorporeal life support, multiple runs, or diagnosis of hypoplastic left heart syndrome. Diagnoses of transposition of the great arteries (p = .03) or persistent pulmonary hypertension of the neonate (p = .004) and extracorporeal life support at <3 days (p = .003) were associated with higher survival. Survivors had fewer mean extracorporeal life support hours (125.5 +/- 121.4 vs. 159.0 +/- 127.6, p = .0006). Logistic regression confirmed significant bivariate findings. A total of 118 hypoplastic left heart syndrome patients were reported from 1996 to 2000. Extracorporeal life support at >15 days was associated with improved survival among hypoplastic left heart syndrome patients (p = .03), and survivors had fewer mean extracorporeal life support hours (89.3 +/- 52.3 vs. 147.5 +/- 129.7, p = .015). Logistic regression showed that only greater number of hours on extracorporeal life support was independently associated with nonsurvival.

Conclusions: Neonatal cardiac extracorporeal life support use increased substantially from 1996 to 2000, with survival to discharge or transfer in more than one third of patients. Hypoplastic left heart syndrome was not associated with nonsurvival. Fewer hours on extracorporeal life support, diagnoses of persistent pulmonary hypertension of the neonate and transposition of the great arteries, and extracorporeal life support at <3 days were associated with survival.