Incidence of intrauterine infections in neonatal intensive care unit: a long-term epidemiological analysis

Abstract

Aim. To improve epidemiological surveillance in the neonatal intensive care unit.Materials and Methods. We retrospectively examine case histories of 646 newborns delivered in Tyumen Perinatal Center and admitted to neonatal intensive care unit (310 with intrauterine infection and 336 without intrauterine infection). Predictors of intrauterine infection (odds ratio and 95% confidence interval) were determined by means of multivariate analysis and logistic regression analysis.Results. According to the multivariate analysis, past medical history of abortions, urinary tract infections, preterm birth during previous pregnancy, colpitis, vaginitis, leukocytosis, elevated C-reactive protein, anemia, and impaired fetal blood supply were among the predictors of intrauterine infection. Stepwise logistic regression identified 7 significant features associated with intrauterine infection: leukocytosis, anemia, pathological alterations in amniotic fluid, preterm birth, impaired fetal blood supply, newborn respiratory distress syndrome, and injury of the newborn central nervous system. Sensitivity and specificity of the prognostic model were 87.7% and 83.9%.Conclusion. Our prognostic model might be applicable for the prediction and verification of intrauterine infection during the prenatal and perinatal observation.