Risk assessment of early-onset sepsis in extremely preterm infants: implications for antibiotic stewardship.

OBJECTIVE: To identify risk factors for early-onset sepsis (EOS) in extremely preterm infants. STUDY DESIGN: This retrospective study included 554 infants at 22-296/7 weeks' gestation at Parkland Hospital Dallas, Texas, from 2018 to 2023. Proven EOS was defined by positive blood culture within 72 hours after birth; suspected EOS by ≥5 days of antibiotics despite negative cultures. Remaining infants served as controls. Maternal, neonatal, delivery, and placental characteristics were compared. Low EOS risk was defined as cesarean delivery without labor, rupture of membranes (ROM) at delivery, and no chorioamnionitis. RESULTS: Prelabor preterm ROM and acute chorioamnionitis on placental pathology were more common among the infants with EOS than those with suspected sepsis and controls (P = 0.02; OR 5.5, 95% CI 1.43-21, P = 0.01, respectively). None of the infants with proven EOS were classified as low risk. CONCLUSION: Risk assessment for EOS should be promoted to enhance antibiotic stewardship.