RESUMO
Introduction. In developing countries, a lack of decentralization of perinatal care leads to many high-risk births occurring in facilities that do not have NICU, leading to admission to a PICU. Objective. To assess SNAP II and SNAPPE II as predictors of neonatal death in the PICU. Methodology. A prospective study of newborns divided into 3 groups according to postnatal age: Group 1 (G1), of 0 to 6 days; Group 2 (G2) of 7 to 14 days; and Group 3 (G3), of 15 to 28 days. Variables analyzed were SNAP II, SNAPPE II, perinatal data, and known risk factors for death. The Hosmer-Lemeshow test and the receiver operating characteristics (ROC) curve were used with SPSS 17.0 for statistical analysis. An Alpha error <5% was considered significant. Results. We analyzed 290 newborns, including 192 from G1, 41 from G2, and 57 from G3. Mortality was similar in all 3 groups. Median SNAP II was higher in newborns that died in all 3 groups (P < 0.05). The area under the ROC curve for SNAP II for G1 was 0.78 (CI 95% 0.70-0.86), for G2 0.66 (CI 95% 0.37-0.94), and for G3 0.74 (CI 95% 0.53-0.93). The area under the ROC curve for SNAPPE II for G1 was 0.76 (CI 95% 0.67-0.85), for G2 0.60 (CI 95% 0.30-0.90), and for G3 0.74 (CI 95% 0.52-0.95). Conclusions. SNAP II and SNAPPE II showed moderate discrimination in predicting mortality. The results are not strong enough to establish the correlation between the score and the risk of mortality.
