Outcome prediction in Greek neonatal intensive care units using a score for neonatal acute physiology (SNAP)

Abstract

This study was undertaken to evaluate the performance of the score for neonatal acute physiology (SNAP) in Greece, to examine the predictive power of SNAP calculated during the 12 hours after admission in comparison with customarily calculated SNAP during the first 24 hours, and to assess SNAP during the second 12 hours from admission as a measure of response to treatment. A total of 579 newborns admitted to three neonatal intensive care units (NICUs) from two cities in Greece were enrolled in the study; SNAP was determined during the first 12 hours, the second 12 hours, and the first 24 hours from admission to the NICU and calculated using an algorithm based on deviations from normal values of 26 physiologic parameters. All three variants of SNAP were powerful predictors of vital status at discharge, as well as of duration of stay among survivors. A five-point increase in SNAP in the first 12 hours corresponds to a more than twofold ratio in the odds for death, whereas a five-unit difference in SNAP from the second 12 hours corresponds to a more than threefold ratio. The combined 24-hour score was similar to that for the first 12 hours. A considerable advantage of SNAP was its independence from more traditional predictors of neonatal death, notably gestational age, birth weight, and Apgar score. The combination of all of these predictors improved further the overall predictive potential. SNAP is a useful tool in medical research and can be applied in different population groups. Its independence from birth weight underlines its added value to predict fatality ratios. Moreover, the results of the present study indicate that SNAP can be estimated without loss of predictive efficiency during the first 12 hours from admission to the NICU, whereas SNAP during the second 12 hours adequately reflects the effectiveness of early medical interventions.