Outcome-associated factors in pediatric patients treated with extracorporeal membrane oxygenator after cardiac surgery.

BACKGROUND: The use of the extracorporeal membrane oxygenator (ECMO) for postoperative cardiac patients has not resulted in the same high success rate as when ECMO is used for neonates with pulmonary hypertension or pulmonary failure. The reason for this is poorly understood. METHODS AND
RESULTS: We analyzed retrospectively all pediatric patients placed on ECMO after surgery for a congenital heart lesion between 1981 and 1995 (n = 64). Patients had a two-ventricular repair (A) or pulmonary blood flow supplied by an aortopulmonary shunt (B) or by a cavopulmonary connection (C). Indication for ECMO was unsatisfactory hemodynamics due to (1) ventricular dysfunction, (2) pulmonary failure, (3) pulmonary hypertension, or (4) a combination or (5) for unclear reasons. Hospital survival was related to these and other factors. Overall hospital survival was 33%; 42% of group A patients survived to discharge, whereas only 25% and 17% survived in groups B and C, respectively. Survival was unrelated to the indication for ECMO but appeared to be lower when ECMO was initiated in the operating room or > 50 hours after surgery. Except for one patient with pneumonia, no patient survived who was on ECMO for > 208 hours. ECMO discontinuation in nonsurvivors was due to neurological (30%) or multiple complications (39%), the lack of return of cardiac function (12%), or other reasons (15%).
CONCLUSIONS: This review suggests that the diagnosis of single ventricle, initiation of ECMO in the operating room or > 50 hours after surgery, and ECMO for > 208 hours are associated with patient nonsurvival. Noncardiac complications more frequently led to discontinuation of ECMO than did failure of the return of cardiac function.