Antibiotics are often necessary in high-risk patients such as neonates. However, exposure to broad-spectrum antibiotics has been associated with adverse neonatal outcomes. Variation in antibiotic use across neonatal intensive care units has been demonstrated on a regional level, but little is known about United States nationwide antibiotic use among hospitalized neonates. Prior studies have measured antibiotic use rates (AUR; antibiotic therapy days as a portion of days present) rather than antibiotic days of therapy per 1000 patient-days (DOT/1000pd), the preferred metric in antimicrobial stewardship practice.
Hospitals participating in the Vizient (formerly University HealthSystem Consortium) Clinical Database/Resource Manager with > 100 neonatal discharges from January to December 2016 were identified. Facility-level clinical outcomes, diagnoses and antibiotic utilization data from 118 hospitals were included. The primary antibiotic utilization metric was DOT per 1000 patient-days; AUR and antibiotic spectrum index (ASI) per antibiotic day were also evaluated according to previously published methods.
The number of neonatal discharges per facility in 2016 ranged from 228 to 15,773 (median 2578, interquartile range [IQR] 1314 – 3927). Of the 118 hospitals 94 (80%) provided care to neonates with birthweight less than 1500 grams, 77 (65%) performed major surgical procedures, 32 (27%) performed cardiac surgery, and 19 (16%) performed extracorporeal membrane oxygenation. Across all hospitals, there was 71-fold variation in antibiotic DOT/1000pd with range from 7.9 to 560.7 (median 271.1; IQR 181.7 – 347.5) and 85-fold variation in AUR with range from 0.4% to 34.1% of days present (median 16.7%; IQR 10.7% – 20.8%). The ASI per antibiotic day ranged from 2.0 to 7.4 (median 6.2; IQR 5.8-6.5).
There is substantial variation in antibiotic use among neonates hospitalized in academically affiliated United States centers. Variation in days of exposure is greater than variation in spectrum of activity per day of therapy. Understanding sources of variation in antibiotic use at the facility level will be important to provide informative benchmarking of neonatal antimicrobial management.
See more of: Poster Abstract Session