Outcome predictors of pediatric extracorporeal cardiopulmonary resuscitation.

Extracorporeal cardiopulmonary resuscitation (ECPR) allows clinicians to potentially rescue pediatric patients unresponsive to traditional cardiopulmonary resuscitation (CPR). Clinical and laboratory variables predictive of survival to hospital discharge are beginning to emerge. In this retrospective, historical cohort case series, clinical, and laboratory data from 31 pediatric patients (<21 years of age) receiving ECPR from March 2000 to April 2006 at our university-affiliated, tertiary-care children's hospital were statistically analyzed in an attempt to identify variables predictive of survival to hospital discharge. Seven patients survived to hospital discharge (23%), and 24 patients died. Survival was independent of gender, age, and CPR duration. ECPR survival was, however, associated with a lower pre-ECPR phosphorus concentration (P = 0.002) and a lower pre-ECPR creatinine concentration (P = 0.05). A classification tree analysis, using, in part, a pre-ECPR phosphorus concentration threshold and a CPR ABG base excess concentration threshold, yielded a 96% nominal accuracy of predicting survival to hospital discharge or death. A large, multicenter, prospective cohort study aimed at validating these predictive variables is needed to guide appropriate ECPR patient selection. This study reveals the potential survival benefit of ECPR for pediatric patients, regardless of CPR duration prior to ECPR cannulation.