Advanced Patient Care Pharmacist -- NICU Nationwide Children's Hospital Columbus, Ohio, United States
Background: Sepsis is a leading cause of neonatal morbidity and mortality in the neonatal intensive care unit (NICU). Optimal duration of antibiotic therapy for Gram-negative bacteremia is not fully known. In 2018, the Neonatal Antimicrobial Stewardship Programs at the 2 study sites recommended treating uncomplicated Gram-negative bacteremia for 7 days if follow-up blood culture was sterile. Objective: To evaluate the efficacy and safety of short (≤8 days) versus long (≥10 days) antibiotic treatment duration for Gram-negative bacteremia among infants in the NICU Design/Methods: Retrospective cohort analysis of infants treated for Gram-negative bacterial sepsis at 7 NICUs of Nationwide Children’s Hospital, Columbus, OH and 1 NICU at University of Texas Health San Antonio, TX from 1/2016-5/2022. Infants were identified by review of the microbiology report of positive blood cultures. Patients were excluded if they had polymicrobial bloodstream infection, meningitis/osteomyelitis/endocarditis, or died before completion of antibiotic therapy. The primary outcome was recurrence of bacteremia with the same organism within 14 days of stopping antimicrobial therapy. Secondary outcomes were emergence of Gram-negative multi-drug resistant organisms (MDRO) and mortality. Pertinent demographic data were obtained from the electronic medical records. Results: 75 infants (39, short course; 36, long course) were included in the study. The median gestational age was 33 weeks (IQR: 26-37) and 34 weeks (IQR: 27-37) while median birth weight was 1930 grams (IQR: 783, 2849) and 1947 grams (IQR: 1078, 2765) in the short and long course groups, respectively. Early-onset sepsis ( < 72 hours of age) occurred in 10 and 5 infants in the short and long course groups, respectively. Twenty-seven (69%) and 30 (83%) infants had a central venous catheter at the onset of bacteremia in the short and long course groups, respectively. Escherichia coli was the most common pathogen in both groups (28 [72%], short course; 17 [47%], long course). The median time to first effective antibiotic administration was 2 hours (IQR: 1-3) in both groups. There were 2 recurrences of bacteremia, both in the long course group. Five infants had a subsequent infection with an MDRO and all were in the long duration group.
Conclusion(s): In this multi-site study, treatment failure and Gram-negative MDROs occurred only among infants who received ≥10 days of antibiotic therapy for uncomplicated Gram-negative bacteremia. Shorter antibiotic duration (≤8 days) appears to be a safe intervention that will reduce antibiotic use in the NICU.
