Duration of early empirical antibiotic exposure in very low birth weight infants with suspected early-onset sepsis and associated factors: a retrospective cohort study

Introduction Early antibiotic exposure in very low birth weight (VLBW) infants has been linked to alterations in gut microbiota and increased neonatal morbidity. However, data on actual treatment duration and associated factors remain limited. Objective To determine antibiotic exposure time, measured in hours of life, in VLBW infants empirically treated for early-onset sepsis (EOS) with negative blood cultures, and to identify factors associated with longer exposure. Methods A retrospective cohort study was conducted in a level IIIC neonatal intensive care unit between March 2021 and December 2023. VLBW infants who received empirical antibiotics for EOS risk and had negative blood cultures were included. Clinical, perinatal, and time-related variables were analyzed. Results Of 286 admitted VLBW infants, 161 met inclusion criteria. Empirical antibiotics were administered to 56.3% without proven infection. Median time from birth to antibiotic discontinuation order was 45 h for both ampicillin and gentamicin, and 42 h when considering blood culture incubation time. Initiation of antibiotics during weekends was associated with 0.46 additional days of therapy [CI (0.148, 0.762); p = 0.004]. Over the study period, antibiotic discontinuation occurred progressively earlier. Conclusions Although antibiotic discontinuation is increasingly prompt, VLBW infants remain exposed closer to 48 than to 36 h of life. Being born on weekends is associated with prolonged antibiotic exposure. Awareness of actual exposure times can guide further optimization of antibiotic stewardship in this high-risk population.