Background: Neonatal exposure to antibiotics is linked to disrupted gut microbiome. We revised our nursery sepsis risk protocol to align with the American Academy of Pediatrics 2018 guidelines in order to further improve antibiotic stewardship in neonates at risk for early onset sepsis (EOS).
Objective: We aimed to reduce antibiotic administration by 25% in asymptomatic neonates at risk for EOS, including those with abnormal laboratory findings, within 6 months of new protocol administration. Secondary aims included reducing antibiotic duration and decreasing laboratory testing and lumbar punctures (LP).
Design/Methods: We conducted a retrospective cohort study of neonates ≥35 weeks gestation at risk for EOS in an urban academic medical center nursery. Pre-protocol period was 04/01/2018-04/30/2019 (cohort 1) and post-protocol was 05/01/2019-11/27/2019 (cohort 2). Inclusion criteria: neonates with maternal ICD 10 diagnosis of chorioamnionitis, prolonged rupture of membranes, and late preterm neonates exposed to group B streptococcus (GBS) with inadequate prophylaxis. Exclusion criteria: clinical illness at birth. Our revised sepsis risk protocol relied on clinical assessment and called for 1) holding off antibiotics in asymptomatic neonates with abnormal labs (complete blood count, C-reactive protein) 2) avoiding repeat labs if initial tests normal and 3) limiting LP to ill appearing neonates or those with bacteremia. Serial vitals and physical exams were performed. We discontinued antibiotics at 36-48 hours of negative cultures.
Results: We evaluated 188 neonates. Antibiotic use decreased by 69% (P 0.023). Number of blood tests and LP rates decreased by 25 % and 75% respectively (Table). Five infants received antibiotics; 3 neonates for clinical illness at 8-10 hours of life, while 2 were given antibiotics off protocol. Of the 29% with abnormal labs, only 1 became symptomatic and received antibiotics for > 48 hours. No positive blood or CSF cultures were identified. No re-admissions seen for sepsis within three days of hospital discharge.
Conclusion: Using our revised EOS protocol that relied on clinical assessment, we were able to effectively reduce antibiotic usage, LP rates, and number of lab tests in both term and late pre-term neonates. We conclude that abnormal labs do not reliably predict neonatal EOS and were not associated with blood culture positivity. Asymptomatic neonates with abnormal labs may be safely monitored off antibiotics.