OUTCOME PREDICTION IN GREEK NEONATAL INTENSIVE-CARE UNITS USING A SCORE FOR NEONATAL ACUTE PHYSIOLOGY (SNAP)

Objectives. This study was undertaken to evaluate the performance of t he score for neonatal acute physiology (SNAP) in Greece, to examine th e predictive power of SNAP calculated during the 12 hours after admiss ion in comparison with customarily calculated SNAP during the first 24 hours, and to assess SNAP during the second 12 hours from admission a s a measure of response to treatment. Methodology. A total of 579 newb orns admitted to three neonatal intensive care units (NICUs) from two cities in Greece were enrolled in the study; SNAP was determined durin g 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 devi ations from normal values of 26 physiologic parameters. Results. All t hree variants of SNAP were powerful predictors of vital status at disc harge, as well as of duration of stay among survivors. A five-point in crease in SNAP in the first 12 hours corresponds to a more than twofol d 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. T he combined 24-hour score was similar to that for the first 12 hours. A considerable advantage of SNAP was its independence from more tradit ional predictors of neonatal death, notably gestational age, birth wei ght, and Apgar score. The combination of all of these predictors impro ved further the overall predictive potential. Conclusions. SNAP is a u seful tool in medical research and can be applied in different populat ion groups. Its independence from birth weight underlines its added va lue to predict fatality ratios. Moreover, the results of the present s tudy indicate that SNAP can be estimated without loss of predictive ef ficiency 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.