The early use of antibiotics in the neonatal period increases antimicrobial resistance . In 2015, the World Health Organization (WHO) identified antimicrobial resistance as an enduring public health threat and published a Global Action Plan  and in the same year it was estimated that 214,000 neonatal sepsis deaths each are attributable to resistant pathogens . The most commonly utilized medications in the Neonatal Intensive Care Unit (NICU) are antibiotics, with greater than 80% of all NICU admissions receiving antibiotics during their hospital course [4, 5].
Life-threatening infections in the NICU generally fall into the two broad categories of early-onset sepsis (EOS) and late-onset sepsis (LOS). A diagnosis of EOS occurs within the first 48–72 h of life with an incidence of 0.3–1.0 per 1000 live births, and a LOS diagnosis is after the initial 48–72 h with a higher rate of 2.2 per 1000 live births [5, 6].The underlying causative organisms for EOS and LOS vary, requiring different approaches to antibiotic stewardship.
Adverse outcomes of early antibiotic use in neonates
Antibiotic use in the NICU in the first week of life increases morbidity and mortality, results in mother and child separation, and increasing healthcare costs [7, 8]. Early or prolonged empiric antibiotic use for preterm neonates results in an increased risk of necrotizing enterocolitis (NEC), infections, and mortality [5, 9, 10]. For instance, a recent retrospective study of very low birth weight neonates (less than 1500 g and 32 6/7 weeks’ gestation) demonstrates for each additional day of antibiotics; there is a 24% increase in the risk for the development of sepsis, NEC, or death . Furthermore, utilization of antibiotics early and for a prolonged duration reduces the gut microbiota increasing antibiotic resistance and altering the immune system early in life [1, 11,12,13]. Alterations to microbes early in life increase the risk for autoimmune disorders, obesity, and allergic diseases greater than the two-fold risk for asthma [11, 13, 14].
Antibiotic stewardship programs in the NICU
Consequently, there is a need for robust antibiotic stewardship programs (ASPs), particularly in the neonatal population, who have underdeveloped immune systems. ASPs are hospital-based programs dedicated to improving antibiotic use by optimizing the treatment of infections and reducing adverse events associated with antibiotics. The Infectious Disease Society of America and the Society for Healthcare Epidemiology of America recommend that APS, compromised of appropriate antibiotic selection, proper dosing, therapy duration, as well as the route of administration, be implemented in the NICU to decrease inappropriate use and resistance . For instance, inappropriate and overuse of vancomycin results in high colonization rates leading to outbreaks of vancomycin-resistant enterococci (VRE) . However, the implementation of a vancomycin guideline resulted in a significant decrease in vancomycin use in two tertiary NICUs of 35% and 65%, demonstrating the applicability of ASPs in the NICU . Another retrospective chart review showed that implementing ASPs into the NICU could decrease total antibiotic days of therapy (DOT) by 18%, decrease targeted broad-spectrum antibiotic DOT by 70%, and decrease vancomycin use in two NICUs by 35% and 62% .
Implementing an ASP provides clinical management, surveillance interventions, and work strategies, including prescribing practices such as choosing the correct antibiotic, dosing, course duration, and monitoring, as well as educating healthcare workers . Following the implementation of ASPs, a tertiary NICU in the United Kingdom significantly reduced overall antibiotic use by 43% from a median of 347 antibiotic use days to 198 antibiotic use days per 1000 patient days, reduced median days of antibiotic use at discharge from three to two days, and decreased practice variations . Importantly, the use of an ASP in a level III-IV NICU resulted in a reduction from 543 to 380 DOT per 1000 patient days and decreased rates of late-onset sepsis from 11.4 to 6.5% without increasing readmission rates .
Despite the prevalence of ASPs, there continues to be significant variability in antibiotic use rate (AUR), defined as the number of patient days that neonates were exposed to antibiotics (1 or more) per 100 patient days. Alarmingly, misuse of antibiotics in the NICU may be as high as 20–50% . Analysis of 127 NICUs in California revealed a 40-fold variance in AUR from 2.4 to 97.1% correlating with the NICU level of care . Level III-IV NICUs (regional, community) had a 7 to 12-fold variance in AUR while the variation in AUR at level II (intermediate) NICUs was significantly higher at 31-fold .
The American Academy of Pediatrics (AAP) endorses the use of a multivariate risk calculator to guide initiation of antibiotics in infants > 34 weeks gestation at risk for early-onset sepsis with prompt discontinuation of antibiotics after 36–48 h if blood culture remains sterile . A more extensive cohort study of 204,485 neonates at 35 or greater week’s gestation utilization of the EOS calculator reduced antibiotic administration within the first 24 h of birth from 5.0 to 2.6% . Another meta-analysis review of 175,752 newborns applying the EOS calculator resulted in a lower relative risk for antibiotic therapy with no higher readmission rates .
Utilization of the NICU BacT/Alert microbial detection technology allows for detecting infection in as few as 24–36 h . In fact, in one review of 845 neonatal blood cultures, all gram-negative isolates were identified by 26 h and many as early as nine hours after inoculation . However, a study utilizing the National Antimicrobial Prescribing Survey Australian Database with 215 neonates from 39 different hospital audits revealed 22% of antibiotics were given beyond 48 h with 9% more than 72 h despite a confirmed infection in only 4.2% neonates .
The Center for Disease Control and Prevention (CDC)  developed seven standardized core elements of ASPs, including leadership, accountability, drug expertise, actions, tracking, reporting, and education. In a recent quality audit, none of the 143 participating NICUs report following all seven of these guidelines, while noting a median AUR of 17% . Incorporating an automatic stop order for antibiotics at 48-h in the electronic order set on admissions is one strategy shown to significantly reduce antibiotic use by 30–38%, creating a mandatory review and opt-in approach for use beyond 48-h .
A collaboration of 146 NICUs participating in the Choosing Antibiotics Wisely Campaign, a QI collaborative by Vermont Oxford Network online program with interactive web sessions, decreased the median AUR from 16.7 to 12.1%, a 34% relative risk reduction . Concurrently, participating NICUs increased use of the CDC seven ASP core measures of leadership 15.4–68.8%, accountability 54.5–95%, drug expertise 61.5–85.1%, actions 21.7–72.35%, tracking 14.7–78%, reporting 6.3–17.7% and education 32.9–87.2% highlighting the necessity for leadership engagement and new methods for reporting antibiotic practices.While many hospitals have detailed ASP protocols for adult patients, similar NICU protocols are not commonly implemented. Our study aimed to explore AUR changes in the NICUs of an IDN to expand primary factors driving change in antibiotic prescription practices in the NICU and provide a tool for reporting on ASPs.