Global burden of disease due to rifampicin-resistant tuberculosis: a mathematical modeling analysis

Rifampicin-resistant tuberculosis (RR-TB) remains a major public health challenge, with an estimated 450,000 incident cases in 2021. Rifampicin is a cornerstone of first-line TB therapy; resistance leads to prolonged, more toxic, and often less effective treatment, with lower cure rates and higher mortality than drug-susceptible TB. Many TB patients are not tested for drug resistance, resulting in inappropriate therapy, treatment failure, and recurrence. Beyond acute disease, TB survivors experience ongoing disability, elevated mortality, income loss, and psychosocial sequelae, which appear more common and severe among those surviving RR-/MDR-TB. There is substantial heterogeneity across countries in RR-TB incidence, detection, and treatment capacity, and uneven adoption of improved diagnostics and regimens. This study aims to quantify the lifetime global burden (DALYs) due to incident RR-TB in 2020 across 192 countries, stratified by age, sex, HIV, and resistance status, including both disease-period and post-TB sequelae among survivors.

The paper synthesizes evidence indicating poorer outcomes for RR-/MDR-TB compared to drug-susceptible TB, including lower cure and higher mortality rates with historical long, toxic regimens. Many patients are not tested for resistance, increasing inappropriate treatment. Post-TB sequelae are well-documented: survivors have persistent disability, elevated long-term mortality, economic losses, and psychosocial impacts. Meta-analyses report greater lung impairment and functional limitations among MDR-TB survivors compared to drug-susceptible TB. RR-TB epidemiology varies by region, with the former Soviet Union experiencing high resistance due to historical treatment practices, and southern Africa facing high burden linked to TB-HIV co-epidemics. Despite advances in diagnostics (e.g., Xpert MTB/RIF Ultra) and new RR-TB regimens (e.g., BPaLM), uptake has been uneven. These findings motivate quantifying RR-TB’s lifetime DALY burden, including post-TB effects.

The authors adapted a previously published global cohort model to project lifetime health losses among individuals developing TB in 2020, stratified by country (n=192), age, sex, HIV status, and rifampicin resistance (RR vs RS). The model tracks cohorts over their remaining lifetimes to estimate excess deaths and disability causally attributable to TB and post-TB sequelae. For RR-TB, strata included: untreated, treated with inappropriate first-line regimen, and treated with appropriate RR-TB (second-line) regimen; for RS-TB, treated vs untreated. Key inputs: WHO estimates for TB incidence (by country, sex, age), treatment coverage from notifications, and HIV prevalence in TB from WHO/UNAIDS. Rifampicin resistance prevalence in new and previously treated cases was from drug resistance surveys and surveillance; RR-TB diagnosis and second-line treatment initiation fractions were derived from WHO country-reported data, with imputation where missing. Disease episode duration was based on WHO estimates by treatment and HIV status; disability weights used GBD values. Mortality during disease was modeled using mortality odds ratios by age (from Brazil’s notified TB case fatality data), HIV, treatment status, and regimen appropriateness for RR-TB, calibrated to country-specific case fatality rates for 2020. Survival for RR-TB inappropriately treated with first-line therapy was modeled as a weighted average of second-line treated and untreated RR-TB (wide prior; extreme values explored in sensitivity analyses). Post-TB mortality and disability among survivors were decomposed into causal effects of TB and non-causal confounding using a published approach: overall post-TB mortality rate ratios were drawn from systematic reviews; the causal component was estimated via excess COPD mortality attributable to TB; post-TB disability weights applied similarly. Elevated post-TB mortality and disability for RR-TB vs RS-TB were specified based on evidence of greater sequelae severity among RR-TB survivors; a conservative sensitivity analysis assumed equivalence to RS-TB. Outcomes included YLLs (premature mortality) and YLDs (non-fatal disability) during the disease episode and post-TB, summed to DALYs. Uncertainty was addressed via second-order Monte Carlo simulation (Latin hypercube sampling, 1000 draws) to produce 95% uncertainty intervals; sensitivity was assessed using PRCCs. Alternative specifications tested included longer RR-TB disease duration, varying case fatality for inappropriately treated RR-TB, higher untreated RR-TB fatality vs RS-TB, alternative post-TB mortality source, and equal post-TB sequelae for RR- vs RS-TB. Analyses used R 4.2.2.

- Global burden: Incident RR-TB in 2020 caused an estimated 6.93 million DALYs (95% UI: 5.52, 8.53), representing 5.4% (4.4, 6.3) of total TB DALYs from incident TB in 2020. - Composition of DALYs: 85.0% (79.8, 89.3) due to premature mortality (5.9 million YLLs), and 15.0% due to disability (1.1 million YLDs). 55.8% (46.0, 68.5) of RR-TB DALYs occurred during the disease episode (3.8 million), and 44.2% post-TB among survivors (3.1 million). - Per-case burden: Average 17.3 DALYs per RR-TB case (13.8, 20.6), which is 34.2% (11.8, 55.7) higher than for RS-TB. Excess RR-TB burden is mainly from higher mortality during and after the episode. - Regional distribution (Table 2): South-East Asia contributed 37.2% (2.57 million DALYs), Africa 24.8% (1.72 million), Western Pacific 14.0% (0.97 million), former Soviet Union countries 12.6% (0.87 million). High MDR/RR-TB burden countries accounted for 85.9% of global RR-TB DALYs. - Rates: Global RR-TB DALYs were 88.8 per 100,000 (70.7, 109.2). Highest per 100,000 in former Soviet Union countries (290.7) and southern Africa; Lesotho 773 (444, 1225) and South Africa 703 (449, 1030) per 100,000. - Country totals: India had the largest absolute burden at 1.7 million DALYs (1.3, 2.1), 24.1% of global; Russia 0.5 million (0.4, 0.6), 6.7%. - Demographics (Table 3): DALYs per case decreased with age; DALYs per 100,000 peaked in 35–44-year-olds (129.2). Men had higher population burden than women (109.2 vs 68.1 per 100,000), reflecting higher TB incidence. - HIV: RR-TB DALYs per 100,000 were 39.2 (30.2, 51.7) times higher in people with HIV than in HIV-uninfected; individuals with HIV accounted for 15.9% (12.8, 20.0) of RR-TB DALYs. DALYs per RR-TB case were higher among PLHIV (21.7) than HIV-uninfected (16.6). - Sensitivity: Key drivers of uncertainty included parameters for post-TB mortality and disability among survivors, RR-TB incidence prevalence, and case fatality for RR-TB inappropriately treated with first-line regimens. Alternative scenarios generally produced robust estimates; using alternative post-TB mortality data or assuming equal post-TB sequelae for RR- vs RS-TB reduced total RR-TB DALYs by ~15–16%.

The study quantifies the lifetime health loss due to incident RR-TB, demonstrating that RR-TB contributes substantially to global TB burden and that a large share (about 44%) arises after treatment among survivors. RR-TB incurs higher per-case DALYs than RS-TB, primarily because of elevated mortality during disease and in the post-TB period, consistent with evidence of more severe long-term sequelae after RR-/MDR-TB. Geographic patterns reflect underlying epidemiology and health system factors: former Soviet Union countries have high burdens due to historically high drug resistance, while southern African countries’ high RR-TB burden is intensified by HIV co-epidemics and health system constraints. These findings underscore the importance of comprehensive strategies that extend beyond acute treatment to include post-TB care and rehabilitation, and they support investments in RR-TB prevention, rapid resistance detection, and effective treatment regimens to reduce long-term morbidity and mortality.

RR-TB imposes a major, lifelong health burden globally, with nearly half of DALYs accruing among survivors. Quantifying this burden highlights priority regions and populations (former Soviet Union, southern Africa, men, and people living with HIV) and strengthens the case for expanded prevention, rapid diagnostics, optimized RR-TB regimens, and structured post-TB care and rehabilitation. Future research should use quasi-experimental designs to better isolate causal long-term effects of TB, evaluate interventions to preserve and restore lung function, and conduct policy analyses to identify the most impactful and cost-effective approaches across prevention, diagnosis, treatment, and rehabilitation for RR-TB.

Key limitations include: (1) reliance on country-level resistance surveillance and imputation for data gaps; (2) substantial uncertainty when extrapolating long-term (post-TB) mortality and disability, reflected in wider uncertainty intervals; (3) assumptions required to separate causal post-TB effects from confounding by comorbidities and socioeconomic factors; (4) limited evidence and greater uncertainty for pediatric RR-TB incidence and post-TB outcomes; and (5) assumptions about future access to care and interventions—scale-up of rehabilitation, improved diagnostics, and adoption of newer short-course, all-oral regimens (e.g., BPaLM) could reduce future RR-TB burden relative to current estimates. Sensitivity analyses indicate results are generally robust, though alternative post-TB parameters can reduce estimates by ~15–16%.