Comparison of metrics of neonatal intensive care unit antibiotic use

To compare temporal trends, variation, and correlations between antibiotic use metrics across U.S. neonatal intensive care units (NICUs) and assess associations with mortality.

We conducted a retrospective cohort study of 438,156 infants admitted to 272 NICUs from 2017 to 2021 using the Premier Health Database. Antibiotic use rate (AUR), days of therapy (DOT), and antibiotic spectrum index (ASI) per 1,000 patient or therapy days were calculated both cumulatively by year and at the center level. Mixed-effects models adjusted for center-level characteristics were used for all analyses.

All three metrics declined over time: AUR by 16.8%, DOT by 19.0%, and ASI by 2.5%. AUR and DOT were highly correlated (r = 0.989, P < 0.001), while ASI showed weaker correlations with AUR (r = 0.247, P < 0.001) and DOT (r = 0.338, P < 0.001). None were significantly associated with center-level mortality. ASI had the least variability, indicating more uniform antibiotic selection and lower center-level discriminatory value.

DOT and AUR were comparable measures of antibiotic consumption, both showing significant declines. ASI exhibited the least variability, reflecting more consistency in antibiotic selection. The similarity in dispersion and decline between AUR and DOT suggests that neonatal antibiotic exposure is primarily influenced by initiation and discontinuation decisions rather than regimen complexity. Given its ease of calculation, AUR may be the most practical metric for evaluating the impact of antibiotic stewardship interventions at the center level.