OBJECTIVE: To examine the use of plasma lactate levels to predict mortality and neurodevelopmental outcome of neonates treated with extracorporeal membrane oxygenation. DESIGN: Prospective cohort study. SETTING: Two level III neonatal intensive care units in Canada and the United States. PATIENTS: Seventy-four neonates requiring extracorporeal membrane oxygenation in two neonatal intensive care units from 1994 to 1996. INTERVENTIONS: Differences in clinical and biochemical measurements, including serial lactate levels between three outcome groups (early deaths, adverse survivors, and normal survivors) were compared using analysis of variance. We also examined the predictive relationship between plasma lactate levels and the outcome at neonatal intensive care unit discharge and at 18-24 months postnatal age by backward, stepwise regression and Fisher's exact test. MEASUREMENTS AND MAIN RESULTS: Fifteen (20%) neonates died before neonatal intensive care unit discharge (early deaths), with seven additional deaths before follow-up, which are included in the adverse survivors group. Among 49 early childhood survivors (22 +/- 7 months), 27 were disabled or delayed with Mental and Performance Developmental Indices of 70 +/- 21 and 72 +/- 22, respectively. Early deaths had higher plasma lactate levels and were more acidemic than adverse and normal survivors, who were not different from each other (p <.05). Plasma lactate and the lowest arterial pH independently predicted 42% of the variance of the outcome ( p<.001). A peak lactate level of >or=25 mM predicted early mortality (sensitivity, 47%; specificity, 100%; positive and negative predictive values, 100% and 88%, respectively; p<.001), whereas a level of >or=15 mM predicted adverse outcome (sensitivity, 35%; specificity, 91%; positive and negative predictive values, 89% and 38%, respectively; p<.05). The predictability of plasma lactate was significantly improved in 45 neonates without congenital diaphragmatic hernia or lethal anomalies (sensitivity of 100% for early mortality, negative predictive value of 63% for adverse outcome). CONCLUSIONS In addition to assessing tissue oxygenation, plasma lactate may facilitate the decision-making process by providing early predictive information about the outcome of neonates treated with extracorporeal membrane oxygenation.
OBJECTIVE: To examine the use of plasma lactate levels to predict mortality and neurodevelopmental outcome of neonates treated with extracorporeal membrane oxygenation. DESIGN: Prospective cohort study. SETTING: Two level III neonatal intensive care units in Canada and the United States. PATIENTS: Seventy-four neonates requiring extracorporeal membrane oxygenation in two neonatal intensive care units from 1994 to 1996. INTERVENTIONS: Differences in clinical and biochemical measurements, including serial lactate levels between three outcome groups (early deaths, adverse survivors, and normal survivors) were compared using analysis of variance. We also examined the predictive relationship between plasma lactate levels and the outcome at neonatal intensive care unit discharge and at 18-24 months postnatal age by backward, stepwise regression and Fisher's exact test. MEASUREMENTS AND MAIN RESULTS: Fifteen (20%) neonates died before neonatal intensive care unit discharge (early deaths), with seven additional deaths before follow-up, which are included in the adverse survivors group. Among 49 early childhood survivors (22 +/- 7 months), 27 were disabled or delayed with Mental and Performance Developmental Indices of 70 +/- 21 and 72 +/- 22, respectively. Early deaths had higher plasma lactate levels and were more acidemic than adverse and normal survivors, who were not different from each other (p <.05). Plasma lactate and the lowest arterial pH independently predicted 42% of the variance of the outcome ( p<.001). A peak lactate level of >or=25 mM predicted early mortality (sensitivity, 47%; specificity, 100%; positive and negative predictive values, 100% and 88%, respectively; p<.001), whereas a level of >or=15 mM predicted adverse outcome (sensitivity, 35%; specificity, 91%; positive and negative predictive values, 89% and 38%, respectively; p<.05). The predictability of plasma lactate was significantly improved in 45 neonates without congenital diaphragmatic hernia or lethal anomalies (sensitivity of 100% for early mortality, negative predictive value of 63% for adverse outcome). CONCLUSIONS In addition to assessing tissue oxygenation, plasma lactate may facilitate the decision-making process by providing early predictive information about the outcome of neonates treated with extracorporeal membrane oxygenation.