Association Between Fluoroquinolones and Uveitis, Retinal Detachment and Aortic Aneurysm or Dissection: An Application of Self-controlled Study Designs with Active Comparators

Saturday, August 27, 2022

8:30 AM – 8:45 AM CEST

Location: C1M1

Publication Number: 140

Background: Non-interventional studies have led to concerns over the safety of fluoroquinolone antibiotics. Whether observed adverse events are causally related, or due to confounding by indication is unclear..

Objectives: To estimate the association and investigate the plausibility of a causal effect of fluoroquinolones on uveitis, retinal detachment, and aortic aneurysm or dissection.

Methods: We applied self-controlled study designs to control for time-invariant confounding and used cephalosporins as an active comparator to try to control for time-varying confounding. We conducted separate self-controlled case series (SCCS) for acute uveitis and retinal detachment and a case-time-control (CTC) study for hospitalisation with aortic aneurysm or dissection. We included patients with a first outcome recorded in UK Clinical Practice Research Datalink Aurum primary care data or in linked hospital records (Hospital Episode Statistics Admitted Patient Care) between the 1st April 1997 and 31st June 2019. To account for prescribing trends in the CTC study we matched each case to three controls on calendar date, primary care practice, and birth year. We included first outcome only in each study. Risk windows were 60 days following each prescription in the SCCS and 1-60 days and 91-150 days preceding outcome in the CTC study. We fitted conditional Poisson models adjusting for age and calendar year for the SCCS and a conditional logistic model for the CTC study. We used the interaction term approach of Hallas et al. 2021 to estimate ratios relative to active comparator.

Results: We identified 68,612, 23,701 and 84,886 eligible individuals respectively with uveitis, retinal detachment, and aortic aneurysm or dissection. The adjusted rate ratio for fluoroquinolone exposure and uveitis was 1.08 (95% CI 0.97-1.21) and for retinal detachment was 1.04 (0.84-1.29). We found an increased odds of aortic aneurysm or dissection (adjusted odds ratio [aOR] 1.44, 1.28-1.62) with fluoroquinolone exposure. Relative to cephalosporin there was no evidence for an association with uveitis (aRR 1.00, 0.87-1.15), retinal detachment (aRR 1.06, 0.80-1.41) or aortic aneurysm or dissection (aOR 0.99, 0.85-1.15).

Conclusions: We found no evidence of an association between fluoroquinolones and retinal detachment or uveitis and the harmful association we observed with aortic aneurysm or dissection appears to be driven by time varying confounding. Given that fluoroquinolones are prescribed for acute infections, disentangling the effects of the drug from the infection is vital. To account for both time-invariant and time-varying confounding active comparator self-controlled studies appear to be useful for this class of medication.

Presenting Author(s)

Jeremy Brown, MSc

PhD student
London School of Hygiene and Tropical Medicine
London, United Kingdom

Slides

Co-Author(s)

KW

Kevin Wing

Department of Non-Communicable Disease Epidemiology, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, United Kingdom
SE

Stephen Evans, MSc

Professor
Department of Medical Statistics, London School of Hygiene and Tropical Medicine, United Kingdom
CL

Clémence Leyrat

Department of Medical Statistics, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, United Kingdom
KM

Kathryn Mansfield

London School of Hygiene & Tropical Medicine
London, England, United Kingdom
LS

Liam Smeeth

London School of Hygiene and Tropical Medicine, United Kingdom
Angel Wong

BSc PhD
London School of Hygiene and Tropical Medicine
London, United Kingdom
DY

David Yorston

Tennet Institute of Ophthalmology, Gartnavel General Hospital, Glasgow, UK, United Kingdom
NG

Nicholas Galwey

Research Statistics, GSK, Stevenage, UK, United Kingdom
ID

Ian Douglas, PhD

Professor
Department of Non-Communicable Disease Epidemiology, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, United Kingdom