Abstract
BACKGROUND: Identification of factors associated with antifungal utilization in neonatal, pediatric, and adult patient groups is needed to guide antifungal stewardship initiatives in academic medical centers. METHODS: For this hospital-level analysis, we analyzed antifungal use in hospitals across the United States of America, excluding centers only providing care for hematology/oncology patients. Analysis of variance was used to compare antifungal use between patient groups. Three multivariable linear regression models were used to determine independent factors associated with antifungal use in the neonatal, pediatric, and adult patient groups. RESULTS: For the neonatal, pediatric, and adult patient groups, 54, 44, and 60 hospitals were included, respectively. Total antifungal use was significantly lower in the neonatal patient group (14 days of therapy (DOT)/1000 patient days (PDs) versus 76 in pediatrics and 74 in adults, p < 0.05). There were no significant associations identified with total antifungal DOT/1000 PDs in the neonatal patient group (model R(2) = 0.11). In the pediatric patient group (model R(2) = 0.55), admission to immunosuppressed service lines and total broad-spectrum antibiotic use were positively associated with total antifungal use (coefficients of 1.95 and 0.41, both p < 0.05). In the adult patient group (model R(2) = 0.79), admission to immunosuppressed service lines, total invasive fungal infections, and total broad-spectrum antibiotic use were positively associated with total antifungal use (coefficients of 5.08, 5.17, and 0.137, all p < 0.05). CONCLUSIONS: Variability in antifungal use in the neonatal group could not be explained well, whereas factors were associated with antifungal use in the adult and pediatric patient groups. These data can help guide antifungal stewardship initiatives.