Identifying neonatal risk factors and predicting neonatal deaths in Georgia

Abstract

A predictive model which identifies 38% of the neonatal deaths by isolating only 11% of the births was formulated. It utilized existing vital records data for 230,585 single live births which occurred in Georgia for the period 1974-1976, to describe factors recorded on the birth certificate which are determinants of neonatal mortality on a statewide basis. Maternal age, race, premature birth, spontaneous abortion or stillbirth, complications related to pregnancy, complications not related to pregnancy, and type of birth (multiple gestation) were found to be statistically significant. It is proposed that private physicians evaluate patients on the basis of this model to facilitate estimates of risk of a neonatal death. In addition, public health physicians could utilize this model to make informed decisions in regard to the management of public health programs directed at the care of pregnant women.