- Design
- Two cohort studies in UK primary-care databases with inverse probability weighting
- Population
- About 3.9 million adults without diabetes prescribed fluoroquinolones or cephalosporins
- Primary outcome
- Incident peripheral neuropathy within 60 days
- Effect
- aHR 0.96 (95% CI 0.77–1.20) at first prescription; tendon rupture aHR 1.98 (1.55–2.53)
Using two large UK primary-care databases (1997–2019), researchers compared new peripheral neuropathy within 60 days of a fluoroquinolone or a cephalosporin prescription in almost 3.9 million adults without diabetes or prior neuropathy, weighting for differences between the groups.
The association was close to null at the first prescription (pooled aHR 0.96, 95% CI 0.77–1.20) and across all prescriptions (aHR 1.05, 0.92–1.20). As a check, the same method found the known link with tendon rupture (aHR 1.98, 1.55–2.53), which suggests it could detect a real effect.
Fluoroquinolone neuropathy rests largely on case reports and a regulatory warning. This study, the largest of its kind, offers some reassurance that a short course does not commonly cause new neuropathy. It does not rule out rare idiosyncratic cases, and it does not change the other reasons to use fluoroquinolones sparingly — tendon injury, aortic events, C. difficile and resistance. In India, where fluoroquinolones are widely used, stewardship remains the main reason for restraint.
- Do not assume new neuropathy after a fluoroquinolone course is drug-caused without seeking other causes
- Keep fluoroquinolones for indications where they are clearly needed — tendon and aortic risks remain
- Still stop the drug and assess if neuropathic symptoms appear during treatment
- Rare idiosyncratic cases are not excluded by a population study
Why it matters
It weakens one of the frequently cited harms of fluoroquinolones while leaving the others intact.
Don't overread it
This is observational and cannot exclude rare cases or residual confounding; regulatory warnings remain in place.
The statistics, in plain English
A hazard ratio of 0.96 with an interval from 0.77 to 1.20 means no detectable difference, and a large increase is unlikely. The tendon rupture result is a positive control: it shows the study design could find a known harm.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for neurology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free