Outcome of infants requiring cardiopulmonary resuscitation before extracorporeal membrane oxygenation.

BACKGROUND/PURPOSE: Cardiopulmonary resuscitation (CPR) is reported to be used in a significant number of neonates before initiation of extracorporeal membrane oxygenation (ECMO). This report establishes the incidence of infants who require CPR before initiation of ECMO and elucidates survival rates and long-term neurological outcomes. In addition, the authors sought prognostic factors that could reliably predict survival or long-term neurological outcome before initiating ECMO support. METHODS: The Extracorporeal Life Support Organization (ELSO) registry provided data on 839 neonates who received CPR before ECMO from January 1989 to April 1995. Supplemental questionnaires on 414 infants were returned from 64 ECMO centers regarding details of the CPR event and subsequent neurological development. One-year neurological evaluations were provided on 112 infants. Data were analyzed for statistical significance using chi2, multiple logistic regression, and Kruskal Wallis one-way analysis of variance as applicable, with significance set at P < .05. RESULTS: The incidence of infants requiring pre-ECMO CPR was 13.1%. Infants who received pre-ECMO CPR had a survival rate of 60.8% versus 81.6% survival for infants who did not require CPR (P < .00001). Analysis of ELSO registry data showed survival was significantly associated with primary diagnosis, location of CPR, last pH level before ECMO, and the presence of intraventricular hemorrhage. Questionnaire data on 12-month neurological assessment showed 63% had no impairment and 4% were graded as severely impaired. Twelve-month neurological outcome was significantly associated only with primary diagnosis. CONCLUSION: A survival rate of 60.8% with good neurological outcome at 12 months in 63.4% of infants suggests that CPR alone should not be a contraindication to placing a neonate on ECMO.

Differentiation of segmented and band neutrophils during the early newborn period.

This prospective study evaluated the degree of inter-reader variability in the identification of segmented and band neutrophils from blood smears of full-term, healthy neonates. Wide inter-reader differences of band neutrophil identification and the immature to total neutrophil ratio were observed. Because of poor correlation between evaluators of the same blood smear, the clinical utility of the manual differential leukocyte count in the evaluation of neonates is limited.