Risk of collagen-related disorders and neurological events among patients with uncomplicated urinary tract infection following short treatment with fluoroquinolones: a cohort study

Abstract

Studies of fluoroquinolone (FQ) safety across indications show increased collagen/neurological adverse event (AE) risk, yet patients still receive FQs for uncomplicated urinary tract infections (uUTIs). This retrospective, cohort study investigated the risk of collagen/neurological AEs of special interest (AESIs) with short-term FQ use versus standard-of-care antibiotics (trimethoprim–sulfamethoxazole [SXT], nitrofurantoin [NTF]) among female outpatients with uUTIs. This study was conducted between December 2009 and 2019 using Optum’s de-identified Clinformatics Data Mart Database. Adjusted absolute risks were calculated for composite/collagen/neurological AESIs (Kaplan–Meier cumulative hazards, after applying stabilized inverse probability of treatment weighting [sIPTW]). Adjusted hazard ratios were generated (sIPTW Cox proportional hazard modeling). Overall, 954,777 patients were included: FQ ( n = 386,537 [40.5%]); SXT ( n = 237,120 [24.8%]); NTF ( n = 314,585 [32.9%]). Adjusted absolute risk range for collagen/neurological AESIs was <1%–4.5%. The hazard (95% CI) of tendon rupture was 25% higher with FQ versus SXT (1.25 [1.00–1.57]; P = 0.0497). Patients receiving FQ had lower hazard of neurological (0.95 [0.93–0.97]; P < 0.0001), central nervous system (0.85 [0.80–0.89]; P < 0.0001), and peripheral nervous system (0.96 [0.93–0.98]; P = 0.0016) AESIs versus NTF. Following a short treatment duration, FQs were associated with increased risk of tendon rupture versus SXT and reduced risk (adjusted hazard ratios) of neurological AESI versus NTF. Individual patient risk and consequences for known uncommon, yet serious, AEs need to inform appropriate antibiotic choice in treating uUTIs. Patient profile, efficacy, microbiome impact, safety, and surveillance should inform antibiotic selection for uUTI management, in accordance with guidelines.