Abstract
We aimed to analyze the impact of histological chorioamnionitis (HCA) in the presence of preterm premature rupture of the membranes (PPROM) on obstetric and neonatal outcomes, and its possible predictability. A retrospective cohort analysis of PPROM cases (20-37 weeks) was conducted comparing the patients with and without HCA, seeking a predictive model of HCA using logistic regression. A total of 295 cases of PPROM were selected, of which 72 (24.4%) had HCA. The group with HCA had a shorter latency period and a greater number of clinical and laboratory criteria in the evolution. The group with HCA had a worse comparative result and presented: lower gestational age at delivery, lower average birth weight, lower Apgar scores, longer neonatal hospitalization, worse maternal clinical conditions and, higher rates of stillbirth, low birth weight (LBW), very low birth weight (VLBW), complications in pregnancy and childbirth, and cesarean delivery due to fetal distress or chorioamnionitis. A predictive model for HCA was developed, with the following independent variables: abdominal pain (odds ratio [OR] = 11.61), uterine activity (noticeable contractions on physical exam) (OR = 5.97), fever (OR = 5.77), latency > 3 days (OR = 2.13), and C-reactive protein (OR = 1.01). With this model, an adequate receiver operating characteristic curve was found, with an area under the curve of 0.726, and some HCA probability curves were constructed for different clinical situations. In this novel study, we present a non-invasive predictive model, with clinical and laboratory variables, which may help in decision-making in a patient with PPROM.