Extracorporeal membrane oxygenation for pediatric respiratory failure: five-year experience at the University of Pittsburgh.

OBJECTIVES: To describe the etiology, respiratory severity of illness, and outcome in patients with pediatric respiratory failure who were treated with extracorporeal membrane oxygenation (ECMO). To identify predictors of death, and to compare our morbidity and mortality rates with those rates of a previously reported series of patients with pediatric respiratory failure managed conventionally.

DESIGN: Survey, case series.

SETTING: Intensive care unit in a tertiary care pediatric hospital.

PATIENTS: Twenty-eight pediatric patients (3 wks to 20 yrs of age) who underwent ECMO for pediatric respiratory failure between 1985 and 1991.

MEASUREMENTS AND MAIN RESULTS: Thirteen (46%) of the 28 patients survived. The most common diagnoses were adult respiratory distress syndrome and nonspecific pneumonitis. Multiple organ system failure occurred in only four (14%) patients; most patients died of respiratory failure. The occurrence of persistent airleak during ECMO was significantly greater in nonsurvivors than in survivors. Furthermore, nonsurvivors had significantly less response to lung reexpansion maneuvers compared with survivors, as measured by a calculated compliance index (effective tidal volume/mean airway pressure x 100). The mortality rate was comparable with those rates of other published studies of conventionally managed and ECMO-treated patients with pediatric respiratory failure. Moreover, our patients appeared to exhibit more severe respiratory failure at the start of ECMO than those patients in other studies.

CONCLUSIONS: ECMO appears to be a rational therapy for patients with pediatric respiratory failure who are likely to die with continued conventional management. Recovery of lung function by the end of the first week of ECMO may be a favorable prognostic indicator. Persistent airleak may be a nonfavorable prognostic indicator.