Antibiotic treatment of women with isolated intrapartum fever vs clinical chorioamnionitis: maternal and neonatal outcomes

Abstract

Clinical chorioamnionitis refers to the presence of maternal fever (≥38°C) and at least 2 clinical signs: (1) maternal tachycardia (>100 bpm), (2) fetal tachycardia (>160 bpm), (3) maternal leukocytosis >15,000/mm2, (4) purulent vaginal discharge, and (5) uterine tenderness. Few data exist to guide the appropriate management of women with isolated intrapartum fever in the absence of other clinical signs suggesting chorioamnionitis. This study compared maternal and neonatal infectious outcomes and microbiological outcomes between women with isolated intrapartum fever and women with clinical chorioamnionitis. This 10-year retrospective study included all the laboring women at our institution, at ≥34 weeks of gestation, with a singleton pregnancy and body temperature of ≥38.0°C, with or without other evidences of infection. According to our department protocol, women with isolated intrapartum fever received intravenous ampicillin, whereas women with clinical chorioamnionitis received intravenous ampicillin plus gentamicin. The primary outcome was puerperal endometritis, compared between women with isolated intrapartum fever (treated with ampicillin) and women with clinical chorioamnionitis (treated with ampicillin plus gentamicin). The secondary maternal outcomes consisted of (1) maternal clinical outcomes, such as cesarean delivery, surgical site infection, postpartum hemorrhage, and postpartum length of stay, and (2) microbiological studies, including positive chorioamniotic membrane swabs and blood culture. Among the secondary neonatal outcomes were early-onset sepsis, neonatal intensive care unit admission, and length of stay. Of note, 2 multivariate logistic regression models were created. A model aimed to predict puerperal endometritis controlled for gestational age of >41 weeks, diabetes mellitus, obesity, positive group B streptococcus status, rupture of membrane ≥18 hours, meconium staining, positive chorioamniotic membrane swabs, cesarean delivery, and empiric postdelivery antibiotic administration. A model aimed to predict neonatal early-onset sepsis controlled for gestational age of 34 to 37 weeks, positive group B streptococcus status, rupture of membrane ≥18 hours, and positive chorioamniotic membrane swabs. Overall, 458 women met the inclusion criteria. Compared with women with clinical chorioamnionitis (n=231), women with isolated intrapartum fever (n=227) had higher rates of puerperal endometritis (3.9% vs 8.8%; P=.03), early-onset sepsis (0.4% vs 4.4%; P=.005), positive chorioamniotic membrane swabs (46.3% vs 63.9%; P<.001), and ampicillin-resistant Escherichia coli (35.5% vs 48.9%; P=.033). The rate of group B streptococcus-positive chorioamniotic membrane swabs was similar between the groups. In a subanalysis of women with negative or unknown group B streptococcus status, the puerperal endometritis and neonatal early-onset sepsis rates were higher among women with isolated intrapartum fever than women with suspected chorioamnionitis (8.7% vs 3.3% [P=.041] and 4.1% vs 0% [P<.001], respectively). In 2 multivariate analysis models, among women with isolated intrapartum fever treated with ampicillin compared with those with clinical chorioamnionitis treated with ampicillin and gentamicin, the odds ratio of antibiotic treatment of endometritis was 2.65 (95% confidence interval, 1.06-6.62; P=.036), and the odds ratio of neonatal early-onset sepsis was 8.33 (95% confidence interval, 1.04-60.60; P=.045). Women with intrapartum fever, with or without other signs of infection, were at increased risk of maternal and neonatal complications. The use of ampicillin as a sole agent in isolated intrapartum fever might promote ampicillin-resistant E coli growth in the chorioamniotic membranes and consequently lead to puerperal endometritis and early-onset sepsis. In this context, a broad-range antibiotic should be considered.