Estimation of individual neonatal survival using birthweight and gestational age: a way to improve neonatal care.

The study was conducted to determine the combined effect of birthweight and gestational age at birth on neonatal mortality using individually-identified livebirths. Logistic regression was used for studying the interactive effect of birthweight and gestational age on the individual probability of neonatal death. All livebirths from Chile in 2000 were included in a linked file. Odds ratio models for birthweight and gestational age were developed for each sex. The probability of neonatal death by sex was presented using contour plots. The models were statistically significant, and odds ratios were different and non-linear for the effects of birthweight and gestational age. Contour plots of constant neonatal mortality according to birthweight and gestational age were presented; they were similar for each sex. A single graph for both sexes that estimates the survival potential of infants born too early or too small would improve neonatal care in developing countries.