Revisiting the diagnostic criteria of clinical chorioamnionitis in preterm birth

Objective To re-evaluate the utility of the conventional criteria for clinical chorioamnionitis in the prediction of early-onset neonatal sepsis (EONS) in preterm birth. Design Retrospective cohort study. Setting Seoul, Republic of Korea. Sample A total of 1468 singleton births between 24 and 34 weeks due to preterm labour (n = 713) or preterm prelabour rupture of membranes (n = 755). Method We evaluated three diagnostic categories of clinical chorioamnionitis: Criteria 1, conventional criteria; Criteria 2, combination of any three conventional parameters without prerequisite fever; Criteria 3, Criteria 1 plus positive maternal Creactive protein and neutrophil left-shift into minor criteria. EONS included proven or suspected sepsis within 7 days following birth. Neonatal morbidity and mortality of EONS were also reviewed. Main outcome measures Diagnostic performance of three combinations. Results The prevalence of EONS was 13.8%. Among 203 cases of EONS, maternal manifestation of clinical chorioamnionitis by criteria 1 was evident in only one out of seven, indicating 15.3% sensitivity for EONS prediction. However, with application of criteria 2, sensitivity significantly increased to 34.0%, while compromising specificity from 92.3% to 78.7%. Criteria 3 showed similar diagnostic performance compared with criteria 1 (sensitivity 16.7%, specificity 91.6%). Overall, neonatal mortality and neonatal composite morbidity in EONS were 14.9% and 67.8%, respectively, and there was no difference in neonatal morbidity and mortality between neonates whose mothers showed fever as a sign of clinical chorioamnionitis and those whose mothers did not. Conclusion The renouncement of fever as a prerequisite for the criteria of clinical chorioamnionitis could increase sensitivity for the identification of EONS, a serious outcome of preterm birth.