(A-325) Use of fluoroquinolones and the risk of seizures: a population-based cohort study

Background: Fluoroquinolones can exert excitatory effects on the central nervous system. Our population-based cohort study assessed whether use of fluoroquinolones is associated with an increased risk of seizures.

Methods: We used German claims data from the InGef database (2016–2023) to assemble a cohort of patients initiating fluoroquinolones or amoxicillin. Patients were followed using a 30-day intention-to-treat approach. Inverse-probability-of-treatment-weighting based on propensity scores incorporating many potential confounders was used for confounding control. Cox models estimated hazard ratios (HRs) and 95% confidence intervals (CIs) of incident seizures associated with fluoroquinolones versus amoxicillin. Secondary analyses stratified by age, sex, prior epilepsy, and individual fluoroquinolones. All analyses were repeated using macrolides as an alternate comparator to control for residual confounding.

Results: The study cohort included 3,250,916 patients (mean age, 51 years; 46% female) initiating treatment with fluoroquinolones or amoxicillin. Compared to amoxicillin, fluoroquinolones were not associated with the risk of seizures (HR, 1.10; 95% CI, 0.995-1.23). There was potential effect modification by age with an increased risk among older adults (HR, 1.29; 95% CI, 1.07-1.47), and by prior epilepsy with an increased risk among patients without a diagnosis (HR, 1.34; 95% CI, 1.11-1.61). Using macrolides as comparator led to no associations with the risk of seizures overall (HR, 0.89; 95% CI, 0.72-1.09) and in the different subgroups.

Conclusions: Our large cohort study showed that fluoroquinolones were not associated with the risk of seizures. Risk increases in specific subgroups are possibly a result of bias and do not seem to reflect a true effect.