Impact of risk mitigation measures on oral fluoroquinolone prescribing: a multi-site population-based Canadian cohort study

Background — Use of fluoroquinolones (FQs), broad-spectrum antibiotics, has been linked to adverse health outcomes and resulted in safety warnings by regulatory agencies worldwide. We tested the effect of Canadian risk mitigation measures (RMMs) introduced in January 2017 on FQ prescription rates.

Methods — We conducted a retrospective multi-site cohort study using administrative data from six Canadian provinces. The cohort included adults (18 + years) with outpatient prescriptions for four oral systemic FQs (ciprofloxacin, levofloxacin, moxifloxacin, norfloxacin) between 2008 and 2022. Overall FQ prescription rates and percentage of FQ prescriptions for three antibiotic indications (acute bacterial sinusitis [ABS], acute exacerbation of chronic obstructive pulmonary disease [AECOPD], urinary tract infection [UTI]) were assessed before and after RMMs were introduced. Segmented generalized linear models were applied to monthly prescription rates and percentages for: (1) pre-RMM (January 2008-December 2016; reference), (2) post-RMM pre-COVID (January 2017-Feburary 2020), and (3) post-RMM within-COVID (March 2020-December 2022) segments. We estimated province-specific relative rates (RR) for the post-RMM segments and slope coefficients for pre- and post-RMM segments and their 95% confidence intervals (CIs) and pooled them using random-effects models.

Results — Crude annual FQ prescription rates decreased from 107.5 to 45.0 per 1,000 population over the study period; in the pre-RMM segment, age- and sex-adjusted rates decreased an average of 0.30 per 1,000 population per month (95% CI: 0.19–0.41); province-specific estimates of decrease ranged from 0.16 to 0.48. The pooled RR for the post-RMM pre-COVID segment was 0.50 (95% CI: 0.43–0.59); the post-RMM within-COVID segment pooled RR was similar (RR = 0.38; 95% CI: 0.29–0.50). The decline in percentage of FQ prescriptions post-RMM pre-COVID was largest for UTI (pooled RR = 0.32, 95% CI: 0.25–0.41), followed by ABS (RR = 0.41, 95% CI: 0.34–0.51) and AECOPD (RR = 0.51, 95% CI: 0.37–0.69), although there was variation across provinces.

Conclusions — Canadian RMMs for FQ use were associated with a decrease in prescription rates overall and for three indications, although rates had begun to decline before RMMs were introduced and the magnitude of decrease varied across provinces. Safety warnings tailored to specific indications and regional practices may be needed to address variation in FQ prescribing.