---
title: "Predicting fluoroquinolone resistance in rifampicin-resistant TB: cross-country validation study"
id: "plos-medicine-0-predicting-resistance-to-fluoroquinolones-among-patients-with-rifampicin"
canonical_url: "https://medichelpline.com/clinical-feed/plos-medicine-0-predicting-resistance-to-fluoroquinolones-among-patients-with-rifampicin"
content_type: "clinical_feed_article"
specialty: "Infectious Disease"
source_name: "PLOS Medicine"
source_url: "https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965"
published_at: "2026-09-15T14:00:00.000Z"
evidence_level: "Journal Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# Predicting fluoroquinolone resistance in rifampicin-resistant TB: cross-country validation study
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/plos-medicine-0-predicting-resistance-to-fluoroquinolones-among-patients-with-rifampicin
- **Specialty:** [Infectious Disease](https://medichelpline.com/clinical-feed/infectious-disease.md)
- **Primary Source:** PLOS Medicine
- **Source URL:** [Original Journal Publication](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965)
- **Published At:** 2026-09-15T14:00:00.000Z
- **Evidence Rating:** Journal Feed
## Executive GIST (TL;DR)
- This multinational study evaluated whether routinely available clinical and demographic data can predict **fluoroquinolone (FQ) resistance** among patients with rifampicin-resistant or multidrug-resistant tuberculosis (RR/MDR-TB). - Data came from 5,175 RR-TB patients with FQ drug susceptibility testing (DST) results submitted to the TB Portals platform from eight countries in Eastern Europe and Central Asia (Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Romania, Ukraine) collected between 2012 and 2024. - Overall, 1,772 patients (34.2%) had FQ-resistant TB in the analytic dataset. - Three modeling approaches were compared: logistic regression, neural networks, and XGBoost, under three evaluation strategies: pooled multi-country models, within-country models (internal validation), and cross-country models (external validation on held-out countries). - Pooled models achieved moderate discrimination after optimism correction (AUROC ~0.70–0.72; AUPRC ~0.57–0.59 across algorithms). - Within-country models sometimes reached higher discrimination (AUROC and AUPRC up to ~0.8 in some countries), indicating better performance when models were trained and validated locally. - Cross-country external validation showed variable performance loss: for some held-out countries the drop in AUROC/AUPRC was negligible, while for others reductions exceeded 0.1, depending on the country and algorithm. - Predictors most consistently informative were **case definition** and **treatment-history–related variables**; demographic, comorbidity, social-risk, education, and employment variables had inconsistent contributions across countries and algorithms. - The dataset displayed relatively stable incidence of RR-TB and FQ resistance, which the authors note limits generalizability to settings with changing MDR-TB dynamics. - Main conclusions: readily available clinical/demographic data offer moderate ability to indicate FQ resistance when DST is unavailable, but models trained in one setting cannot be assumed to generalize elsewhere; locally informed models and expanded rapid DST access remain important.
## Clinical Analysis & Structured Key Points
[ Skip to main content ](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#main-content) Advertisement * [plos.org](https://plos.org/) * [Create account](https://community.plos.org/registration/new) * [Sign in](https://journals.plos.org/user/secure/login?page=%2Fplosmedicine%2Farticle%3Fid%3D10.1371%2Fjournal.pmed.1004965) * * About * Browse * Publish * [](https://journals.plos.org/plosmedicine/ "PLOS Medicine") * Search [advanced search](https://journals.plos.org/plosmedicine/search) * 0 [Save](https://journals.plos.org/plosmedicine/article/metrics?id=10.1371/journal.pmed.1004965#savedHeader) [Total Mendeley and Citeulike bookmarks.](https://journals.plos.org/plosmedicine/article/metrics?id=10.1371/journal.pmed.1004965#savedHeader) * 0 [Citation](https://journals.plos.org/plosmedicine/article/metrics?id=10.1371/journal.pmed.1004965#citedHeader) [Paper's citation count computed by Dimensions.](https://journals.plos.org/plosmedicine/article/metrics?id=10.1371/journal.pmed.1004965#citedHeader) * 8 [View](https://journals.plos.org/plosmedicine/article/metrics?id=10.1371/journal.pmed.1004965#viewedHeader) [PLOS views and downloads.](https://journals.plos.org/plosmedicine/article/metrics?id=10.1371/journal.pmed.1004965#viewedHeader) * 0 [Share](https://journals.plos.org/plosmedicine/article/metrics?id=10.1371/journal.pmed.1004965#discussedHeader) [Sum of Facebook, Twitter, Reddit and Wikipedia activity.](https://journals.plos.org/plosmedicine/article/metrics?id=10.1371/journal.pmed.1004965#discussedHeader) Open Access Peer-reviewed Research Article # Predicting resistance to fluoroquinolones among patients with rifampicin-resistant tuberculosis: A cross-country validation study * Tianfang Shao, Roles Conceptualization, Data curation, Formal analysis, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing Affiliation Department of Epidemiology of Microbial Diseases, Yale School of Public Health, New Haven, Connecticut, United States of America [ ![ORCID logo](https://journals.plos.org/resource/img/orcid_16x16.png) https://orcid.org/0009-0005-0338-9909 ](https://orcid.org/0009-0005-0338-9909 "ORCID Registry") ⨯ * Mariana R. Neves, Roles Investigation, Methodology, Writing – review & editing Affiliation Philip R. Lee Institute for Health Policy Studies, University of California San Francisco, San Francisco, California, United States of America ⨯ * Molly Franke, Roles Conceptualization, Investigation, Writing – review & editing Affiliation Department of Global Health and Social Medicine, Harvard Medical School, Boston, Massachusetts, United States of America [ ![ORCID logo](https://journals.plos.org/resource/img/orcid_16x16.png) https://orcid.org/0000-0002-4890-5728 ](https://orcid.org/0000-0002-4890-5728 "ORCID Registry") ⨯ * Carole Mitnick, Roles Conceptualization, Investigation, Writing – review & editing Affiliation Department of Global Health and Social Medicine, Harvard Medical School, Boston, Massachusetts, United States of America ⨯ * Jennifer Furin, Roles Conceptualization, Investigation, Writing – review & editing Affiliation Department of Global Health and Social Medicine, Harvard Medical School, Boston, Massachusetts, United States of America [ ![ORCID logo](https://journals.plos.org/resource/img/orcid_16x16.png) https://orcid.org/0000-0002-0825-7199 ](https://orcid.org/0000-0002-0825-7199 "ORCID Registry") ⨯ * Ted Cohen, Roles Conceptualization, Investigation, Writing – review & editing Affiliation Department of Epidemiology of Microbial Diseases, Yale School of Public Health, New Haven, Connecticut, United States of America [ ![ORCID logo](https://journals.plos.org/resource/img/orcid_16x16.png) https://orcid.org/0000-0002-8091-7198 ](https://orcid.org/0000-0002-8091-7198 "ORCID Registry") ⨯ * Reza Yaesoubi Roles Conceptualization, Funding acquisition, Investigation, Methodology, Supervision, Validation, Writing – review & editing * E-mail: reza.yaesoubi@ucsf.edu Affiliations Philip R. Lee Institute for Health Policy Studies, University of California San Francisco, San Francisco, California, United States of America, Department of Epidemiology and Biostatistics, University of California San Francisco, San Francisco, California, United States of America [ ![ORCID logo](https://journals.plos.org/resource/img/orcid_16x16.png) https://orcid.org/0000-0002-9276-5750 ](https://orcid.org/0000-0002-9276-5750 "ORCID Registry") ⨯ # Predicting resistance to fluoroquinolones among patients with rifampicin-resistant tuberculosis: A cross-country validation study * Tianfang Shao, * Mariana R. Neves, * Molly Franke, * Carole Mitnick, * Jennifer Furin, * Ted Cohen, * Reza Yaesoubi ![PLOS](https://journals.plos.org/resource/img/logo-plos-full-color.svg) x * Published: September 15, 2026 * * [Article](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965) * [Authors](https://journals.plos.org/plosmedicine/article/authors?id=10.1371/journal.pmed.1004965) * [Metrics](https://journals.plos.org/plosmedicine/article/metrics?id=10.1371/journal.pmed.1004965) * [Comments](https://journals.plos.org/plosmedicine/article/comments?id=10.1371/journal.pmed.1004965) * [Media Coverage](http://plos.altmetric.com/details/doi/10.1371/journal.pmed.1004965) * [Peer Review](https://journals.plos.org/plosmedicine/article/peerReview?id=10.1371/journal.pmed.1004965) * [Abstract](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#abstract0) * [Author summary](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#abstract1) * [Introduction](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#sec004) * [Methods](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#sec005) * [Results](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#sec013) * [Discussion](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#sec014) * [Supporting information](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#sec015) * [Acknowledgments](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#ack) * [References](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#references) * [Reader Comments](https://journals.plos.org/plosmedicine/article/comments?id=10.1371/journal.pmed.1004965) * [Figures](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965) [?](https://journals.plos.org/plosmedicine/s/accepted-manuscripts#loc-early-version) ## This is an uncorrected proof. ## Abstract ### Background Fluoroquinolones (FQs) are a cornerstone of most all-oral, shorter regimens endorsed by the World Health Organization for the treatment of rifampicin-resistant or multidrug-resistant tuberculosis (RR/MDR-TB). Knowledge of resistance to FQs can help guide regimen selection at the point of care. In settings where rapid testing for FQ resistance is unavailable, prediction models could support treatment decisions by identifying FQ resistance based on patient characteristics observable at the point of care. These prediction models have been typically developed and evaluated within a single country, and their generalizability across different geographic settings is unclear. ### Methods and findings We used data from 5,175 patients with RR-TB and available FQ drug susceptibility testing (DST) results submitted to the TB Portals, an open-access data-sharing platform curated by the National Institute of Allergy and Infectious Diseases, from eight countries (Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Romania and Ukraine) between 2012 and 2024. Among these patients, 1,772 (34.2%) had FQ-resistant TB. We developed prediction models for FQ resistance using logistic regression, neural networks, and XGBoost. Models were evaluated under three strategies: (1) _pooled models_ trained on multi-country data; (2) _within-country models_ trained and evaluated using internal validation; and (3) _cross-country models_ trained on subsets of countries and externally validated on held-out countries. Model discrimination was assessed using the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). Across the three algorithms, pooled models showed moderate optimism-corrected discrimination, with AUROC ranging from 0.70 to 0.72 and AUPRC ranging from 0.57 to 0.59. Within-country models demonstrated slightly better discrimination, with AUROC and AUPRC reaching 0.8 for some countries. Cross-country external validation showed that performance loss from using a model trained on external data could be negligible to >0.1 AUPROC or AUROC, depending on the country and algorithm. A limited set of predictors, including case definition and treatment-history-related variables, were among the most consistently informative predictors, whereas demographic, comorbidity, social-risk, education, and employment variables showed more variable contributions across countries and algorithms. A limitation of our study is that the incidence of RR-TB and FQ resistance was relatively stable in our analysis dataset. Hence, the results may not generalize to scenarios with marked changes in MDR-TB dynamics. ### Conclusions Predicting FQ resistance using demographic and clinical characteristics showed moderate ability to identify FQ resistance among patients with RR-TB, but their performance and predictor patterns varied across countries. Models developed for one or several countries cannot be assumed to generalize to other settings without rigorous external validation. These findings highlight the limitations of globally trained prediction models for RR/MDR-TB and underscore the need for locally informed prediction models to support clinical decision-making. ## Author summary ### Why was this study done? * Fluoroquinolones are key antibiotics used to treat rifampicin-resistant and multidrug-resistant tuberculosis (RR/MDR-TB), but they are not effective when the infecting strain is resistant to these drugs. * Rapid laboratory tests for fluoroquinolone resistance are not available in many settings, making it difficult for clinicians to choose the most appropriate treatment regimen at the time of diagnosis. * Prediction models based on patient characteristics have been proposed as an alternative, but most have been developed and evaluated in a single country, leaving it unclear whether they can be used reliably in other geographic settings. ### What Did the Researchers Do and Find? * We analyzed data from more than 5,000 patients with RR-TB from eight countries in Eastern Europe and Central Asia, about one-third of whom had fluoroquinolone-resistant tuberculosis. * We developed prediction models and compared models trained on data from multiple countries, models trained using data from individual countries, and models trained on data from multiple countries and tested in countries that were not used for model development. * Models achieved only moderate overall accuracy. Models developed within a country generally performed better than those applied to new countries, although the magnitude of this difference varied substantially across settings. * Previous tuberculosis treatment and disease history were among the most consistently useful predictors, whereas demographic and socioeconomic characteristics contributed less consistently across countries. * The incidence of RR-TB and FQ resistance was relatively stable in our analysis dataset, and our findings may not generalize to settings with changing MDR-TB dynamics. ### What do these findings mean? * Readily available clinical and demographic information can provide some indication of fluoroquinolone resistance when laboratory testing is unavailable, but prediction models alone are not sufficiently accurate to replace drug susceptibility testing. * Prediction models developed in one country or region should not be assumed to work equally well elsewhere without external validation. * These findings support the development and validation of locally adapted prediction models and reinforce the importance of expanding access to rapid drug susceptibility testing for patients with RR/MDR-TB. ## Figures ![Fig 2](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.g002) ![Fig 3](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.g003) ![Table 1](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.t001) ![Table 2](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.t002) ![Fig 1](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.g001) ![Fig 2](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.g002) ![Fig 3](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.g003) ![Table 1](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.t001) ![Table 2](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.t002) ![Fig 1](https://journals.plos.org/plosmedicine/article/figure/image?size=inline&id=10.1371/journal.pmed.1004965.g001) **Citation:** Shao T, Neves MR, Franke M, Mitnick C, Furin J, Cohen T, et al. (2026) Predicting resistance to fluoroquinolones among patients with rifampicin-resistant tuberculosis: A cross-country validation study. PLoS Med 23(9): e1004965. https://doi.org/10.1371/journal.pmed.1004965 **Academic Editor:** Amitabh Bipin Suthar, PLOS Medicine Editorial Board, UNITED STATES OF AMERICA **Received:** February 12, 2026; **Accepted:** September 2, 2026; **Published:** September 15, 2026 **Copyright:** © 2026 Shao et al. This is an open access article distributed under the terms of the [Creative Commons Attribution License](http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. **Data Availability:** The data used in this study are available through the TB Portals website ( ) following completion of a Data Use Agreement (DUA). TB Portals ( ), which is an open-access TB data resource supported by the National Institute of Allergy and Infectious Diseases (NIAID) Office of Cyber Infrastructure and Computational Biology (OCICB) in Bethesda, MD. These data were collected and submitted by members of the TB Portals Consortium ( ). We gratefully acknowledge the patients whose clinical, laboratory, and radiological data made this research possible, as well as the clinicians and site staff involved in their care and data collection. Investigators and other data contributors who originally submitted the data to the TB Portals did not participate in the design or analysis of this study. The study described was not registered, nor was a study protocol prepared. The final, cleaned dataset used for the analysis described here is available in S1 Data. The code used in the analysis can be accessed at ( ). A permanent archive is also available at Zenodo ( ). **Funding:** Research reported in this publication was supported by the National Institute of Allergy and Infectious Diseases (NIAID) of the National Institutes of Health (NIH) ( ) under Award Number R01AI177326 to RY. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. **Competing interests:** The authors have declared that no competing interests exist. **Abbreviations:** AUPRC, area under the precision-recall curve; AUROC, area under the ROC curve; DST, drug susceptibility testing; MDR/RR-TB, multidrug- or rifampicin-resistant TB; NIAID, National Institute of Allergy and Infectious Diseases; OCICB, Office of Cyber Infrastructure and Computational Biology ## Introduction Tuberculosis (TB) remains one of the most pressing global health threats. In 2024, an estimated 10.7 million people developed TB, and 1.23 million died from the disease [[1](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref001),[2](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref002)]. Despite progress in diagnosis and treatment, the burden of TB has been exacerbated by the rise of drug-resistant TB, particularly multidrug- or rifampicin-resistant TB (MDR/RR-TB) [[3](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref003)–[5](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref005)]. MDR/RR-TB is difficult to treat (with a success rate of 71% in 2022) and should be treated with second-line regimens, which include multiple antibiotics [[1](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref001),[6](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref006)]. The selection of antibiotics for treating MDR/RR-TB should ideally be guided by DST. However, access to rapid and comprehensive DST remains limited in many high-burden settings, resulting in substantial delays in identifying resistance to second-line agents such as fluoroquinolones (FQs) [[7](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref007)]. FQs are a cornerstone of most all-oral, shorter regimens endorsed by the World Health Organization for the treatment of RR/MDR-TB [[6](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref006)]. However, FQ resistance continues to challenge efforts to treat MDR/RR TB; in 2024, an estimated 18% of MDR/RR TB cases globally were resistant to any FQs tested [[1](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref001),[8](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref008)]. Recent studies have shown the potential of machine-learning approaches for predicting resistance to specific antibiotics using patient-level data. In Moldova, a clinical risk-based model for predicting resistance to FQs demonstrated reasonable discrimination (area under the receiver operating characteristic curve >0.8) using basic patient characteristics, including age, urban residence, and TB type [[9](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref009)]. Similarly, a study in Brazil developed prediction models for tuberculosis drug resistance using clinical and social characteristics from patients in São Paulo [[10](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004965#pmed.1004965.ref010)]. Together, these studies suggest that clinical and demographic information may contain usable signals for predicting resistance and that prediction models could enhance antibiotic selection in settings with limited or no access to DST. A key limitation of existing studies is that they develop and evaluate resistance prediction models within a single country. Hence, little is known about whether these m
## Related Clinical Research

- [Understanding Parasitic Identification in Delusional Parasitosis](https://medichelpline.com/clinical-feed/plos-one-21-bridging-perception-and-reality-parasitic-identification-in-delusional.md)
- [Simple sequence repeat INDELs in Mycobacterium tuberculosis influence antibiotic susceptibility](https://medichelpline.com/clinical-feed/biorxiv-4-variation-in-multiple-classes-of-simple-sequence-repeats-can-alter-drug.md)
- [Temporal patterns of tuberculosis-associated hyperglycaemia and effects on treatment outcomes in n](https://medichelpline.com/clinical-feed/plos-one-0-temporal-patterns-of-tuberculosis-associated-hyperglycaemia-and-its-effect-on.md)
- [Preclinical study finds lesional liability for PaBQU and DBQU regimens in hard-to-treat TB models](https://medichelpline.com/clinical-feed/biorxiv-14-comprehensive-preclinical-reevaluation-of-pabqu-and-dbqu-regimens-identifies.md)
- [Integrating TB Treatment into Community Pharmacies in Uganda: Human-Centered Design–Derived Implem](https://medichelpline.com/clinical-feed/medrxiv-6-implementation-strategies-for-integrating-tb-treatment-into-community.md)

## Navigation
- [← Back to Infectious Disease Feed](https://medichelpline.com/clinical-feed/infectious-disease.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.