Abstract
INTRODUCTION: Preterm infants are commonly treated with antibiotics on admission to the neonatal unit as part of routine care. We aimed to identify infants <32 weeks' gestation at low risk of early-onset sepsis (EOS) in whom antibiotics could be safely withheld. METHODS: This retrospective cohort study included infants <32 weeks' gestation admitted between January 2012 and June 2022. Data were extracted from electronic databases. Low risk for EOS (LR) was defined as caesarean section delivery, rupture of membranes <1 h prior to birth, no preterm labour and no features of maternal chorioamnionitis. Maternal and neonatal characteristics and neonatal outcomes were compared between LR and not low risk (NLR) infants. IBM SPSS Statistics (Version 29) was used for data analysis. RESULTS: There were 3,285 infants included in the analysis of which 1,035 (31.5%) were LR and 2,250 (68.5%) NLR. No LR infants had culture-confirmed EOS compared with 35 (1.6%) NLR infants. Antibiotics were commenced in the first 48 h of life in 794 (76.7%) LR and 2,159 (96.0%) NLR infants (p < 0.001) and continued for ≥5 days in 226/782 (28.8%) LR and 603/2,107 (28.6%) NLR infants, despite negative blood cultures. There was no difference in mortality or late-onset sepsis between LR and NLR infants. CONCLUSION: Simple clinical parameters available at birth can be used to identify very preterm infants at lower risk of EOS in whom withholding empiric antibiotics could be considered.