Background/Purpose: Neonatal sepsis remains a leading cause of neonatal mortality in the United States despite advances in intrapartum screening and antibiotic administration. Late onset sepsis (LOS) is defined as suspected or confirmed sepsis at greater than seventy-two hours of life. Patient presentation is often nonspecific, creating a low threshold for empiric treatment, which may lead to unnecessary antimicrobial administration and increased development of resistant organisms. Previously, there was not a standardized workflow for treating LOS in our neonatal intensive care unit (NICU) and special care nursery (SCN). This led to variations in practices and created an opportunity for care optimization. A pharmacy-driven order set for empiric therapies in the treatment of LOS was implemented at our institution, composed of screening
suggestions and empiric antibiotic recommendations based on specified patient risk factors. This review evaluated the appropriate use of this new order set and how effectively LOS is managed by evaluating appropriate antimicrobial selection and culture collection.
Methodology: This was a multi-center, IRB-reviewed determined exempt, retrospective cohort review conducted at the Moses Cone Memorial Hospital NICU and Alamance Regional Medical Center SCN in Greensboro and Burlington, NC, respectively. On November 1, 2024, a pharmacy-driven order set for empiric neonatal LOS treatment went live at both sites. For this evaluation, patients were included if they were admitted to either site and received antimicrobial agent(s) for the treatment of LOS. Patients were excluded if they received antimicrobial(s) for another known indication or if they received antimicrobial agent(s) but were not admitted to one of the study locations. Patients were then divided into a pre-group from December 1, 2023 through October 31, 2024 and a post-group from December 1, 2024 through October 31, 2025. The primary outcome was the composite of positive cultures covered by empiric antibiotics and adequate coverage of “culture negative” sepsis with empiric antimicrobials based on patient risk factors and suspected infection source. Secondary outcomes included number of instances in which 2 initial blood cultures were obtained, percentage of patients with positive blood cultures who had repeat blood cultures obtained, average days of therapy, use of the late onset sepsis order set, and whether appropriate doses of antimicrobials were utilized.
Results: 89 instances of LOS in 60 patients were included in the study. There were 51 instances of LOS included in the pre-order set group, and 38 instances included in the post-order set group. The primary composite outcome of positive cultures covered by empiric antibiotics and adequate coverage for culture negative sepsis was 71% in the post-order set group and 66.7% in the pre-order set group (p=0.659). In the post-order set group, providers were more likely to obtain a set of two blood cultures (instances with 2 initial blood cultures obtained 5.9% vs 57.9%, p<0.001). The pharmacy-driven order set was utilized in 55.3% of instances in the post-order set group. Other secondary outcomes were similar between groups.
Conclusions: The implementation of a pharmacy-driven order set at our institution resulted in a clinically significant increase in empiric coverage of culture negative sepsis and significant trend in increase in number of initial blood cultures collected. There was not a significant difference in the percentage of positive cultures covered by empiric antibiotics. Common reasons that empiric therapy did not cover culture negative sepsis included lack of anaerobic coverage in suspected infections of intraabdominal sources and lack of MRSA coverage in patients with MRSA risk factors. Potential limitations to this evaluation include limited sample-size, single-institution review, and variability of provider preference. Future directions include assessing barriers to utilization of the LOS order set and implementation of strategies to increase utilization.
