Contribuição dos escores de gravidade SNAP II, SNAP-PE II, CRIB e CRIB II como preditores de mortalidade neonatal em um hospital universitário

Abstract

Objective: To assess the contribution of SNAP II, SNAP-PE II, CRIB and CRIB II severity scores as predictors of neonatal mortality in newborns admitted to an NICU in a given period. Methods: Prospective cohort study that involved newborns admitted in the NICU of Santa Maria University Hospital between August, 2018 and Jully, 2019. The exclusion criteria were: occurrence of death before 12 hours of life, presence of malformations or genetic syndromes incompatible with life (Trisomy of chromosome XIII and XVIII, hydranencephaly and corpus callosum agenesis) and newborns admitted from other hospitals. Results: 160 newborns were included with a median birth weight of 1835 grams and a gestational age of 34 weeks and 1 day, being 81.88% preterm infants. The mortality rate was 10%, being significantly higher in babies with birth weight below 1000 grams (p<0.0001). The four score applied were significantly higher in newborns who died in comparison with the survivors. SNAP II sensitivity was 81.2% and specificity 38.9%, with a positive predictive value was 95.7% and negative 17.9%. SNAPPE-II sensitivity was 87.5%, specificity 53.5%, positive predictive value 96.8% and negative 20%; for CRIB, the sensitivity and specificity were 66.7%, the positive predictive value was 82.7% and the negative 30%; and for CRIB II, the sensitivity was 83.3%, specificity 47.6%, positive predictive value 77.8% and negative 53.3%. The area under ROC curve was 0.70 for SNAP II and 0.84 for SNAP-PE II to the total number of newborns assessed (p=0.023). For those with birth weight <1500g, the areas were for CRIB 0.72, CRIB II 0.80, SNAP II 0.64 and SNAP-PE II 0.76 (p=0.046). Conclusion: Considering all the babies studied SNAP-PE II was the best predictor of mortality. For babies with birth weight <1500 grams CRIB II showed the best prediction performance.