380 - Decreasing Antibiotic Usage in Preterm Neonates with Culture Negative Late Onset Sepsis

Poster Number: 380
Publication Number: 380.1750

Background: Late onset sepsis (LOS), defined as invasive infection presenting after 72 hours of life, poses a threat to the preterm population. It is associated with higher mortality rate and morbidities like prolonged hospitalization and neurodevelopmental impairment. Clinical manifestations of LOS is variable for preterm infants and diagnostic laboratory markers are not well established. Due to these limitations, empiric antibiotics are commonly initiated for clinical suspicion of LOS. However, studies have shown that as high as 91% of preterm infants with LOS have negative blood cultures. The duration of antibiotics remains controversial for culture negative LOS (CN-LOS). Inappropriate duration of antibiotic use is associated with adverse outcomes such as toxicity, invasive Candida infection, and antibiotic resistant pathogens. Recent studies and QI initiatives have suggested the feasibility of a shorter antibiotic duration of 5 days for CN-LOS.

Objective: Our primary aim was to decrease the duration of antibiotics for CN-LOS in preterm infants ≥ 30 weeks gestational age (GA) by 20% within 1 year. Balancing measures that we evaluated included: (1) antibiotic re-initiation and (2) infant death within 7 days of antibiotic end date.

Design/Methods: The antibiotic stewardship committee at a Level 4 NICU developed a LOS evaluation guideline that standardized the evaluation and management of LOS. We defined CN-LOS as ≥ 2 persistent or worsening symptoms of sepsis and/or persistently abnormal infectious labs (CRP >1.6mg/dL, IT ratio >0.2, ANC < 1000 cells/µL, and/or platelet < 100K/µL). Infants diagnosed with CN-LOS were re-evaluated on day 5 for possible discontinuation of antibiotics. QI interventions included NICU staff education, guideline revisions, and information/data distribution. Baseline data from Nov - Dec 2022 were compared to the intervention data from Jan - Oct of 2023 over 3 PDSA cycles.

Results: In a 12-month period, 239 infants born at ≥30 weeks GA underwent evaluation for suspected LOS. 71 infants had known etiology for infection while 14 infants had CN-LOS. Remainder were evaluated for LOS but antibiotics were discontinued after 2 days with lack of evidence of sepsis. The average duration of antibiotics for CN-LOS decreased from 8 days to 5 days, representing a 37.5% decrease. Of those who received 5 days of antibiotics for CN-LOS, no infant required re-initiation of antibiotics or experienced death.

Conclusion(s): We demonstrated that with close clinical and laboratory monitoring, CN-LOS in preterm infants born at ≥30 weeks GA can be successfully treated with a 5-day duration of antibiotics.