Abstract
BACKGROUND: The FDA issued a "black box" warning regarding risks of fluoroquinolones in 2008 with updates in 2011, 2013, and 2016. OBJECTIVE: To examine antimicrobial use in hospital-treated UTIs from 2000 to 2020. DESIGN: Cross-sectional study with interrupted time series analysis. PARTICIPANTS: Patient encounters with a diagnosis of UTI from January 2000 to March 2020, excluding diagnoses of renal abscess, chronic cystitis, and infection of the gastrointestinal tract, lungs, or prostate. MAIN MEASURES: Monthly use of fluoroquinolone and non-fluoroquinolone antibiotics were assessed. Fluoroquinolone resistance was assessed in available cultures. Interrupted time series analysis examined level and trend changes of antimicrobial use with each FDA label change. KEY RESULTS: A total of 9,950,790 patient encounters were included. From July 2008 to March 2020, fluoroquinolone use declined from 61.7% to 11.7%, with similar negative trends observed in inpatients and outpatients, age ≥ 60 and < 60 years, males and females, patients with and without pyelonephritis, and across physician specialties. Ceftriaxone use increased from 26.4% encounters in July 2008 to 63.6% of encounters in March 2020. Among encounters with available culture data, fluoroquinolone resistance declined by 28.9% from 2009 to 2020. On interrupted time series analysis, the July 2008 FDA warning was associated with a trend change (-0.32%, < 0.001) and level change (-5.02%, p < 0.001) in monthly fluoroquinolone use. CONCLUSIONS: During this era of "black box" warnings, there was a decline in fluoroquinolone use for hospital-treated UTI with a concomitant decline in fluoroquinolone resistance and rise in ceftriaxone use. Efforts to restrict use of a medication class may lead to compensatory increases in use of a single alternative agent with changes in antimicrobial resistance profiles.