Hemorrhagic risk of concomitant direct oral anticoagulants and fluoroquinolones: integration of pharmacovigilance and therapeutic drug monitoring

Background Direct oral anticoagulants (DOACs) and fluoroquinolone antibiotics (FQNs) are often co-prescribed. A pharmacokinetic interaction is plausible, as DOACs are P-glycoprotein (P-gp) substrates and several FQNs can bind to and affect P-gp activity. However, robust clinical evidence characterizing the associated hemorrhagic reporting signal remains limited. Methods We conducted an integrated pharmacoepidemiological and therapeutic drug monitoring (TDM) study. Disproportionality analysis was performed using FDA Adverse Event Reporting System (FAERS) data (2010–2025), calculating adjusted reporting odds ratios (adj. ROR) and Ω shrinkage values. Concurrently, a prospective observational cohort (n = 50) measured trough and peak plasma concentrations of dabigatran with and without levofloxacin or moxifloxacin. Results FAERS analysis identified a significant reporting signal for bleeding for dabigatran coadministered with FQNs (adj. ROR = 4.68, 95% CI: 3.41–6.55), particularly levofloxacin (adj. ROR = 6.12) and ciprofloxacin (adj. ROR = 3.84). No significant signals were found for rivaroxaban, apixaban, or edoxaban. Consistently, TDM showed significantly elevated peak dabigatran concentrations with levofloxacin (133.36 ng/mL, P = 0.008) and moxifloxacin (138.20 ng/mL, P < 0.001) compared to monotherapy (65.34 ng/mL), alongside a numerical trend towards more bleeding events. Conclusion This integrated analysis provides suggestive evidence for a pharmacokinetically plausible interaction that may increase the reporting odds of bleeding between dabigatran and certain FQNs (e.g., levofloxacin). Other DOACs appear safer with FQNs coadministration. For dabigatran patients requiring FQNs, alternative agents or enhanced monitoring should be considered.