Frequency of Post-Treatment Recurrence After Short-Course Regimens for Drug-Resistant Tuberculosis: A Systematic Review and Meta-Analysis
In brief
About 1.2% had recurrent tuberculosis after short-course drug-resistant TB treatment
Across 6,954 patients in 25 studies, the pooled recurrence frequency after short-course treatment was 1.24%, though estimates varied between studies. Overall comparisons found no clear difference from longer regimens, while randomized trials suggested more recurrence with short courses only when follow-up time was not accounted for. Low-certainty evidence and inconsistent follow-up leave long-term durability uncertain.
- Journal
- Journal of clinical medicine (Q1)
- Published
- 23 October 2026
- Study design
- Systematic review / meta-analysis of RCTs
- Evidence level
- Level 1, High (CEBM 1a)
- Authors
- Roman S Ladyu, Anastasia I Gaida, Viktor A Mudrov, Platon I Eliseev, Ruslan O Khachetlov, Ksenia V Rudakova, et al.
- PMID
- 43030416
- DOI
- 10.3390/jcm15218273
Why clinicians should know about it
- Picked for Epidemiology (paper of the day, 15 November 2026): Synthesizes recurrence surveillance and comparative evidence after TB treatment
Abstract
Background/Objectives: Short-course regimens (SCRs) are increasingly used to treat drug-resistant tuberculosis (DR-TB), but data on post-treatment recurrence remain limited. We aimed to estimate the frequency of post-treatment recurrence after SCRs and compare recurrence between SCRs and comparator regimens. Methods: We conducted a systematic review and meta-analysis of clinical studies reporting treatment duration, post-treatment follow-up, and recurrent tuberculosis episodes after SCRs for DR-TB. The primary single-arm analysis used a binomial-normal random-effects generalized linear mixed model (GLMM) based on r/A, retaining zero-event cohorts without a continuity correction. Comparative analyses used pooled risk ratios (RRs) and person-time-adjusted incidence rate ratios (IRRs). Results: Twenty-five studies contributed 26 SCR cohorts. In the primary analysis, 98 recurrent tuberculosis episodes occurred among 6954 patients; the pooled proportion was 1.24% (95% confidence interval [CI], 0.79-1.93%; 95% prediction interval, 0.24-6.21%). The pooled proportion was 1.20% (95% CI, 0.70-2.03%) for fully oral SCRs, 1.57% (0.85-2.85%) for pretomanid-containing SCRs, 1.55% (0.93-2.56%) for SCRs lasting ≤ 39 weeks, and 1.95% (1.20-3.15%) for SCRs lasting ≤ 36 weeks. In the overall comparative analysis, no statistically significant difference in post-treatment recurrence was observed between SCRs and comparator regimens: the pooled RR was 1.35 (95% CI, 0.34-5.40), and the person-time-adjusted IRR was 1.25 (95% CI, 0.50-3.15). In the randomized controlled trial (RCT) subset, the pooled RR indicated a higher recurrence risk with SCRs (RR, 2.72; 95% CI, 1.25-5.89), whereas the corresponding person-time-adjusted estimate did not show a statistically significant difference (IRR, 1.66; 95% CI, 0.81-3.43). This divergence indicates that comparative estimates are sensitive to how post-treatment follow-up duration is accounted for. Meta-regression identified no statistically significant associations between recurrence frequency and the examined cohort-level clinical characteristics. The certainty of evidence, assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework, was low for the estimated recurrence frequency after fully oral SCRs, very low for the subgroup estimates, and low for the comparative estimates. Conclusions: These findings indicate a low frequency of post-treatment tuberculosis recurrence after SCRs and support the durability of treatment outcomes achieved with these regimens. However, substantial between-study heterogeneity and differences in post-treatment follow-up limit the interpretation of the pooled estimate as a uniform long-term recurrence risk. Further studies with standardized follow-up beginning at treatment completion are needed to provide more precise estimates of long-term outcomes after SCRs.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.