OBJECTIVE: To identify and to estimate the risk factors associated to neonatal mortality in Taubaté, São Paulo, Brazil. METHODS: This case-control study enrolled live births in the city of Taubaté during 2003. Live birth data and death records were obtained from São Paulo Health Department. Neonatal deaths were cases and babies born alive in the same day of cases were the controls. A single data file was created by linkage approach. Dependent variable was neonatal death. Independent variables were those related to socio-demographic characteristics and prenatal care (maternal age, years in school, parity, previous stillbirths, prenatal care), as well as the biological ones (birthweight, gender, gestational age, congenital defects and Apgar score). Logistic regression was used to identify and to estimate the risk factors associated to neonatal death. The variables with p<0.20 were introduced in the model and maintained if p<0.05, by SPSS 10.0. RESULTS: 392newborns with 34 neonatal deaths were studied. There were 198 males (50%), 60 low birth weight (15%) and 51 preterm infants (13%). Mean maternal age was 26 years. The variables "low birth weight", "Apgar score <8" and "congenital defects" were significantly associated to neonatal deaths. The model explained 58% of neonatal deaths, with an accuracy of 93%. CONCLUSIONS: The risk factors associated to neonatal deaths show the importance of low birth weight prevention and a good prenatal and delivery care, in order to improve neonatal vitality at birth. The model had a good accuracy and a high power to explain the neonatal death.
neonatal mortality; birth defects; infant; low birth weight; risk factors
