124 - Empiric antibiotic prescribing practices for late-onset sepsis in Canadian and American neonatal intensive care units

Poster Number: 124
Publication Number: 124.1881

Background: Late-onset sepsis (LOS) in the Neonatal Intensive Care Unit (NICU) is associated with significant morbidity and mortality. Vancomycin is often prescribed for empiric gram-positive coverage given concerns for coagulase-negative staphylococci (CoNS) and methicillin-resistant staphylococcus aureus (MRSA). Although empiric therapy is not standardized, there is a potential role for alternative regimens such as anti-staphylococcal penicillins (e.g. cloxacillin/oxacillin/nafcillin).

Objective: We aimed to better understand empiric antibiotic prescribing practices for LOS in order to inform antibiotic stewardship efforts.

Design/Methods: An online survey using REDCap was distributed through the Canadian Neonatal Network and Children’s Hospitals Neonatal Consortium to 71 NICUs across Canada (n=33) and the United States (n=38). Descriptive statistics were used to summarize survey results.

Results: The overall survey response rate was 52%. Most responses were from regional NICUs with on-site surgical capability (76%). Nine centers (24%) reported empiric vancomycin use in all LOS cases. Others reported empiric vancomycin use in cases of known MRSA colonization (57%), critically unwell patients (43%), previous CoNS sepsis (22%), current central lines (22%), central lines within 48 hours (19%), or multiple recent vascular access attempts (3%) (Figures 1, 2). In critically unwell patients with a central line, 76% and 27% of sites reported empiric vancomycin and anti-staphylococcal penicillin use for gram-positive coverage, respectively (Figure 3). This contrasts with responses for critically unwell patients without a central line, where 49% of sites reported empiric vancomycin and 32% reported empiric anti-staphylococcal penicillin use. For patients who are not critically unwell, these responses changed to 43% empiric vancomycin use and 51% empiric anti-staphylococcal penicillin use for those with a central line, and 16% empiric vancomycin and 52% empiric anti-staphylococcal penicillin use for those without a central line.

Conclusion(s): Empiric antimicrobial regimens for LOS varied greatly between centers. The decision to initiate vancomycin or an anti-staphylococcal penicillin for empiric gram-positive coverage was center- and case-dependent. This study demonstrates the lack of standardization of care for LOS in the NICU and has the potential to inform antimicrobial stewardship efforts in both Canada and the United States.