Objectives: Fluoroquinolones (FQ) are one of the most prescribed broad-spectrum antibiotics and a cornerstone of tuberculosis (TB) treatment. TB patients may have had FQ resistance before treatment initiation. However, the association between pre-diagnostic FQ exposure and acquired FQ-resistant TB remains unclear.
Methods: A case-control study was conducted among all pulmonary TB patients in Shanghai during 2022-2023. Cases were TB patients who had possible acquired FQ resistance identified through whole-genome sequencing (WGS), while controls were FQ susceptible patients. Pre-diagnostic FQ prescriptions were extracted from the Shanghai Electronic Health Record (EHR) platform.
Results: Among 3496 patients, 7.4% had FQ-resistant TB, with 93.5% (243/260) phylogenetically inferred as possible acquired resistance. Multivariate analysis revealed FQ exposure was the strongest predictor of possible acquired FQ resistance with an aOR of 4.31 for a single prescription and 13.18 for multiple prescriptions. A nonlinear dose-response relationship between resistance probability and prescription number was found. Most prescriptions to cases were from non-TB-designated tertiary hospitals for non-TB respiratory diseases, with an exposure interval of ≥61 days prior to TB diagnosis.
Conclusion: Acquired resistance dominates FQ resistance in Shanghai TB patients. The dose-response relationship between pre-diagnostic FQ exposure and possible acquired resistance underscores the need for judicious FQ use.
Keywords: Electronic health records; Fluoroquinolones resistance; Tuberculosis; Whole-genome sequencing.
Copyright © 2025 The Author(s). Published by Elsevier Ltd.. All rights reserved.