2026 Southeastern Residency Conference

Background/Purpose: Early-onset sepsis (EOS) is a systemic infection occurring within the first week of life and remains a major contributor to neonatal morbidity and mortality. Clinical manifestations of EOS are often nonspecific. Empiric broad‑spectrum antibiotics are frequently initiated in suspected infection based on maternal or neonatal risk factors. Literature supports discontinuation of empiric antibiotic therapy if blood cultures remain negative at 36 – 48 hours. Despite support for discontinuation, providers often continue antibiotics due to concerns of missed infection or presumed clinical improvement due to initiation of therapy. Prolonged antibiotic exposure in neonates is associated with adverse outcomes including disruption of gut microbiome, necrotizing enterocolitis, and development of antimicrobial resistance. This study aimed to evaluate the prevalence of prolonged empiric antibiotic therapy for culture‑negative EOS.

Methodology: This was a multi-centered, retrospective chart review which evaluate neonates delivered within the Northeast Georgia Health System and admitted to the NICU between January 1, 2023, and December 31, 2024. Patients were included if blood cultures were obtained within the first 72 hours of life, and at least one dose of ampicillin plus gentamicin was administered. Patients with positive blood cultures within the first 72 hours were excluded, as well as any patient that was discharged and readmitted within the first 72 hours of life. The primary objective was to evaluate the prevalence of prolonged empiric antibiotic therapy (> 48 hours) despite negative cultures. The secondary objective was to evaluate the prevalence of treatment failure, defined as discontinuation and reinitiation of any antibiotic within the first 14 days of life.

Results: A total of 125 neonates met inclusion criteria. Of these, 94 neonates (75%) received standard empiric therapy, and 31 neonates (25%) received prolonged empiric therapy despite negative blood cultures. Baseline characteristics were similar between groups, with no statistically significant differences observed for gestational age, birth weight, gender, or mode of delivery. The mean total days antibiotic exposure was 3 days in the standard group compared to 7 days in the prolonged group (P < 0.001). Treatment failure occurred in 9 of 94 neonates (10%) in the standard group and 3 of 31 neonates (10%) in the prolonged group, with no significant difference between groups (P = 0.613).

Conclusions: Of the 125 neonates included in this retrospective chart review, (25%) received empiric antibiotics beyond the consensus recommended 36 to 48 hour rule‑out period despite negative cultures. Prolonged empiric therapy did not reduce treatment failure when compared to the standard group. There was a clinically significant increase in antibiotic exposure in the prolonged group. Further studies are needed to evaluate the long-term effects of empiric treatment for early onset sepsis. These findings also highlight opportunities for antimicrobial stewardship initiatives aimed at reinforcing evidence-based, consensus recommended use of empiric antibiotics.