Global prevalence of extensively drug-resistant tuberculosis among adults with bacteriologically confirmed tuberculosis, 2015-2025: a systematic review and meta-analysis with assessment of HIV co-infection

Abstract

Background: Although the progress has been made under the World Health Organization(WHO) End TB Strategy, extensively drug-resistant tuberculosis (XDR-TB) remains an important challenge to global TB control. Patients with XDR-TB face prolonged treatment options with associated drug toxicity, and the outcomes remain poor. Understanding the prevalence of XDRTB is essential for assessing its global burden and informing surveillance and control strategies. Methods: A systematic review and meta-analysis were performed to estimate the global prevalence of XDR-TB among adults with bacteriologically confirmed TB. PubMed was searched for studies published between January 2015 and December 2025. Observational studies reporting XDR-TB prevalence using either the pre-2021 or revised 2021 WHO definitions were included. Pooled prevalence was estimated using the Freeman-Tukey double-arcsine transformation with restricted maximum likelihood estimation (REML). Between-study heterogeneity was assessed using the I² and the association between the proportion of previously treated patients and XDR-TB prevalence was evaluated by meta-regression. Data on PLHIV and mortality were summarized descriptively. Subgroup analyses examined study settings, geographic region, age group, and treatment history and diagnostic methods. Results: Among 42 studies,159,371 individuals with bacteriologically confirmed tuberculosis, among whom 4,500 XDR-TB cases were included in the analysis. The pooled prevalence of XDR-TB was 1.48% (95% CI: 0.68%-2.55%), with high between-study heterogeneity (I² = 98.9%). After the exclusion of one outlier study, the pooled prevalence decreased to 1.11 % (95% CI: 0.60%-1.74%) with persistent high heterogeneity (I² = 98.6%). Among studies using the pre-2021 WHO definition, the pooled prevalence was 1.10% (95% CI: 0.57%-1.79%; I² = 98.7%). Only three studies applied the revised 2021 WHO definition, with prevalence estimates ranging from 0.37% to 2.43% which was insufficient for comparison between definition groups. Evidence on HIV-associated XDR-TB remained limited across the included studies. XDR-TB prevalence was significantly higher among previously treated TB cases (2.71%, 95% CI: 1.04%- 5.01%) than among new TB cases (0.38%, 95% CI: 0.06%-0.90%; p = 0.0014). Meta regression revealed a significant positive association between the proportion of previously treated patients and XDR-TB prevalence (coefficient = 0.0216, p = 0.0137). Mortality outcomes were reported only in two studies. Prevalence varied significantly by study setting and region. In contrast, no statistically significant differences were observed by age group or diagnostic methods. Conclusion: Despite significant variation among studies and contexts, the frequency of XDRTB among individuals with bacteriologically confirmed TB was comparatively low. Previously treated TB patients had substantially higher XDR-TB prevalence than new TB cases (2.71% vs 0.38%) indicating that treatment history is an important indicator for drug-resistant TB surveillance and the need to focus TB control strategies on this group. The limited availability of accessible research restricted the interpretation of HIV-specific prevalence, mortality, and studies using the revised 2021 WHO definition. To address these evidence gaps. additional research and more studies using the consistent definitions and reporting of HIVstratified outcomes are needed. Keywords: extensively drug-resistant tuberculosis; XDR-TB; tuberculosis; prevalence; systematic review; meta-analysis; drug-resistant tuberculosis