Abstract
This study sought to analyze the frequency of preterm births and identify the respective risk factors and their evolution over a decade (2001-2005-2010) in a city in São Paulo state. It is a time-series study using data from the Live Birth Information System. Univariate and multiple logistic regression were used to identify factors associated with preterm births, and linear regression was used to evaluate the time-series tendency of such factors in the period. To avoid underestimation, a correction factor was applied to the preterm frequencies obtained. A discrete increase in preterm birth was observed: 12.5%, 12% and 13.2%. After adjusting for confounding factors, maternal age equal to or higher than 35 years and cesarean sections were associated with higher chances for preterm births and increased over the decade. The number of births in high-risk hospitals and of women with fewer than seven prenatal consultations decreased during the period, while multiple pregnancies remained stable. Compared to vaginal births, cesarean sections doubled the chance of preterm birth. Among the five factors associated with preterm birth, three (cesarean sections, births in high-risk hospitals and fewer than seven prenatal consultations) can be modified by actions taken by health care services.
Key words
Premature; Pregnancy; Epidemiology; Information systems