Paediatrica Indonesiana (Jan 2017)

Predictive value of Score for Neonatal Acute Physiology and Perinatal Extension II for neonatal mortality in Sanglah Hospital, Denpasar, Indonesia

  • I Gede Ketut Aryana,
  • I Made Kardana,
  • I Nyoman Adipura

DOI
https://doi.org/10.14238/pi56.5.2016.257-61
Journal volume & issue
Vol. 56, no. 5
pp. 257 – 61

Abstract

Read online

Background Neonatal mortality, which is largely caused by severe illness, is the biggest contributor to overall infant mortality. The World Health Organization (WHO) estimated that 4 million neonates die yearly worldwide, often due to severe infection and organ system immaturity. Neonates with severe illness require treatment in the neonatal intensive care unit (NICU), in which a reliable assessment tool for illness severity is needed to guide intensive care requirements and prognosis. Neonatal disease severity scoring systems have been developed, including Score for Neonatal Acute Physiology and Perinatal Extension II (SNAPPE II), but it has never been validated in our setting. ObjectiveTo study the prognostic value of SNAPPE II as a predictor of neonatal mortality in Sanglah Hospital, Denpasar, Indonesia. Methods This prospective cohort study was conducted in the NICU of Sanglah Hospital, Denpasar from November 2014 to February 2015. All neonates, except those with congenital anomaly, were observed during the first 12 hours of admission and their outcomes upon discharge from the NICU was recorded. We assessed the SNAPPE II cut-off point to predict neonatal mortality. The calibration of SNAPPE II was done using the Hosmer-Lemeshow goodness-of-fit test, and discrimination of SNAPPE II was determined from the receiver-operator characteristic (ROC) curve and area under the curve (AUC) value calculation. ResultsDuring the period of study, 63 children were eligible, but 5 were excluded because of major congenital abnormalities. The SNAPPE II optimum cut-off point of 37 gave a high probability of mortality and the ROC showed an AUC of 0.92 (95%CI 0.85 to 0.99). The Hosmer-Lemeshow goodness-of-fit test showed a good calibration with P = 1.0 Conclusion The SNAPPE II has a good predictive ability for neonatal mortality in Sanglah Hospital, Denpasar, Indonesia.

Keywords