136 - Treating the Unknown: Patterns of Antibiotic Use in Culture-Negative Sepsis

Publication Number: 3130.136

Meghan V. Handley, University of Virginia School of Medicine, Gordonsville, VA, United States; Sara Manetta, University of Virginia School of Medicine, Charlottesville, VA, United States; Megan Elliott, University of Virginia School of Medicine, Charlottesville, VA, United States; Sarah Miller, UVA Childrens, Charlottesville, VA, United States; Jonathan R. Swanson, University of Virginia School of Medicine, Charlottesville, VA, United States

Background: Culture-negative sepsis is a common and elusive diagnosis in neonatal intensive care units (NICUs). When infants show clinical signs or laboratory abnormalities concerning for infection, antibiotics are often continued despite sterile cultures. Studies suggest that culture-negative sepsis may occur up to 16 times more frequently than culture-proven sepsis in the United States. Many providers worry about missing an undiagnosed infection in high-risk infants, which can increase mortality. Treatment guidance is clear for proven infections; however uncertainty remains with regards to how to treat unproven infections.. For these reasons, antibiotic stewardship in culture-negative suspected sepsis has become a focus for NICU quality improvement.

Objective: To describe and evaluate current prescribing practices and duration of antibiotic therapy for culture-negative suspected sepsis in a level IV NICU.

Design/Methods: We performed a retrospective analysis of all blood cultures obtained between July 2022 and September 2025. We excluded positive cultures, prophylactic antibiotic use, treatment for ureaplasma, and those transferred or who died during therapy. Repeat cultures during initial therapy were also excluded from analysis. For negative cultures, antibiotic duration in days was recorded. For those exceeding 48 hours, provider documentation was reviewed for rationale. Culture-negative early-onset ( < 72 hrs; EOS) or late-onset (LOS) was defined as clinical suspicion of sepsis with negative blood cultures.

Results: A total of 2418 blood cultures were obtained from 1401 patients; 51.6% were for EOS and 48.4% for LOS. Overall, 9.2% were positive (3.7% EOS, 14.8% LOS). Among negative cultures, 11.3% of EOS and 22.2% of LOS evaluations were associated with antibiotic treatment beyond 48 hours. For EOS treated >48 hours, common indications were presumed pneumonia (26%), clinical illness (25.6%), and maternal factors (17.8%). For LOS, presumed pneumonia (28.3%), and gastrointestinal pathology (24.5%) predominated. Compared with EOS, LOS evaluations were more often prolonged for GI pathology, cellulitis/skin infection, tracheitis, and UTI, whereas EOS courses were more often prolonged for maternal factors and laboratory abnormalities.

Conclusion(s): There is inconsistency in clinical reasoning for continuing antibiotics beyond 48 hours after negative EOS or LOS evaluations. Larger studies are needed to identify drivers of prolonged therapy and inform neonatal antibiotic stewardship guidelines. Additionally, units should develop local guidelines on treating infections without documented positive cultures.

Table 1: EOS vs LOS Demographics
Demographics per early onset sepsis (EOS) versus late onset sepsis (LOS) evaluations of all negative cultures within the study period.

Table 2: LOS Sepsis and Antibiotic Duration Demographics
Demographics per early and late onset sepsis evaluations who received less than 48 hours of antibiotics versus those who received greater than 48 hours of antibiotics. Those of lower gestational age, lower birth weight, and lower apgars were significant for receiving longer than 48 hours of antibiotics for both early onset and late onset sepsis evaluations with negative blood culture growth.

Table 3: Clinical reasoning EOS vs LOS with antibiotics greater than 48 hours duration
Clinical reasoning for continuation of antibiotics for greater than 48 hours comparing early onset versus late onset sepsis evaluations. For early onset evaluations, maternal factors were the most significant for increased antibiotic duration. Laboratory anomalies, GI pathologies, and clinical illness were also all significant for both EOS and LOS evaluations when determining antibiotic duration.