Inappropriate antibiotic initiation for suspected urinary tract infection: a multicenter study comparing psychiatric vs. nonpsychiatric hospital admissions

Abstract

Differentiating between urinary tract infection (UTI) and asymptomatic bacteriuria is challenging in patients with altered mental status (AMS) and/or decompensated psychiatric illness. The purpose of this multicenter, retrospective cohort study was to identify the extent, patterns, and predictors of inappropriate antibiotic initiation for UTI (IAI-U) within an 18-hospital health system comprising of behavioral health units. Patient-specific factors were evaluated as predictors for IAI-U using descriptive, bivariate and multivariate regression with two-tailed alpha= 0.05. A total of 150 patients were included for data analysis, half of which were admitted for psychiatric evaluation. Almost half of patients (42.7%) had IAI-U, with a nearly even split between IAI-U occurring in the emergency department vs. following hospital admission. There was no statistical difference in IAI-U between patients admitted primarily for a psychiatric diagnosis compared to those with a non-psychiatric diagnosis (46.7% vs. 38.7% respectively, p = 0.32). Factors associated with IAI-U were female sex (OR 5.83, 95% CI: 1.73–19.71; p = 0.005) and being discharged to skilled nursing facilities (OR 11.77, 95%: CI 2.11–65.48; p = 0.005). This exploratory analysis suggests IAI-U occurs frequently in the various subgroups evaluated, even those where no association was found. These findings demonstrate an ongoing urgent need for interventions to reduce IAI-U across all sectors of inpatient health systems. Future large-scale studies are needed to fully evaluate these findings across health care systems and specialty acute care settings.