With one of the highest tuberculosis (TB) incidence rates globally and a significant TB-HIV co-infection burden, Lesotho faces a critical public health challenge. The TB TRIAGE+ project leverages cutting-edge technologies, such as AI-powered chest X-ray analysis, and community-based approaches to bridge diagnostic gaps. These groundbreaking strategies aim to improve access to TB screening and reduce transmission in remote areas.
A high TB burden with a risk of HIV co-infection
Lesotho, a small, landlocked country in Southern Africa, is facing significant public health challenges, particularly with regard to tuberculosis (TB) and HIV co-infection. The country has the second highest TB incidence rate in the world, exacerbated by a high prevalence of HIV, which compromises the immune system and increases susceptibility to TB disease. The World Health Organization (WHO) Global TB report 2024 estimates that Lesotho has an incidence rate of around 664 cases per 100,000 population, and this incidence is all the more alarming given that 50% of the TB patients are co-infected with HIV. In particular, the number of undetected TB cases remains a major public health concern because it results in prolonged infectiousness and disease transmission, increased risk of suffering and death due to delayed treatment initiation, serious health sequelae and catastrophic financial consequences.
Health infrastructure and challenges
The healthcare system in Lesotho is primarily public, with a network of clinics and hospitals providing services to the population. However, the system often struggles with limited resources, inadequate infrastructure, and a shortage of healthcare personnel, particularly in rural areas. This makes it challenging to conduct effective TB screening and ensure adequate linkage to care.
In addition, the TB epidemic in Lesotho is fuelled by several societal factors. High rates of unemployment and poverty, as well as the mobility patterns associated with labour migration contribute to the spread of TB. Many people temporarily migrate to neighbouring countries to work, where they may be exposed to TB, and then return home, facilitating the transmission of the disease in communities.
The TB TRIAGE+ project
Within this context, the TB TRIAGE+ project was implemented to address critical diagnostic gaps and improve TB detection in communities using innovative, cost-effective tools. A screening strategy that uses an inexpensive, rapid screening test first, followed by a more expensive, time-consuming confirmatory test if positive, could significantly reduce diagnostic costs, increase early access to diagnosis and improve outcomes. This underscores the potential of low-cost, rapid screening tests in improving TB detection, especially in hard-to-reach areas where under-reporting and under-diagnosis are prevalent due to barriers to care and inadequate diagnostic tools. This approach aligns with global efforts to end the TB epidemic by 2030 by enhancing access to care in underserved regions.
The main trial called TB TRIAGE+ TRIAL focuses on evaluating the effectiveness and cost-effectiveness of two community-based screening approaches using either digital chest X-ray systems equipped with artificial intelligence (AI)-powered tuberculosis detection software (CAD4TB, Delft Imaging Systems, NL) alone or in combination with C-reactive protein (CRP) testing to improve early detection of TB. A total of 20,460 participants were enrolled in the study, with 13,473 of whom were from Lesotho. Results from this study will help in identifying the most (cost-)effective TB screening strategy. Results are currently being analysed.
This project is part of the European and Developing Countries Clinical Trials Partnership 2 (EDCTP2) program supported by the European Union (grant number: RIA2018D-2498; TB TRIAGE+). The TB TRIAGE+ TRIAL is conducted by the Swiss Tropical and Public Health Institute (Swiss TPH) in collaboration with SolidarMed in Lesotho and the Human Sciences Research Council (HSRC) in South Africa, supported by ITM Antwerp, Radboud University Nijmegen, Charité Berlin and University Hospital Basel.
Additional research highlights of the ongoing TB TRIAGE+ project
Computer-Aided Detection (CAD) using digital chest X-rays and CRP as triage tests: Results from a TB TRIAGE+ ACCURACY study showed that CAD4TB is accurate as a triage test for patients from areas with a high TB and HIV burden attending health facilities with TB symptoms. Further refinement of CAD thresholds is, however, needed to optimise its use in diagnosing TB in different populations. The accuracy results for CRP disqualify it as a stand-alone TB triage test. However, its diagnostic value for TB screening, for example in combination with other screening tests, requires further research.
Bayesian latent class analysis: Another innovative approach was the use of Bayesian latent class analysis to reduce both verification and reference bias in TB screening studies. This method allowed more accurate estimation of test performance even when imperfect reference standards are used, which is common in community-based screening.
HIV-TB integration: In regions with high HIV prevalence, the trial also examined the integration of advanced HIV disease (AHD) care into a community-based TB screening campaign. Point-of-care CD4 testing and provision of an AHD care package alongside TB screening proved feasible and helped streamline care for individuals with advanced HIV, improving overall health outcomes. The inclusion of TB screening in HIV care protocols should be expanded, given its effectiveness in early detection.
Integration of screening for non-communicable diseases (NCDs): As part of the TB TRIAGE+ TRIAL community health campaign, screening for diabetes mellitus and hypertension was carried out and referrals made to the nearest health facility. This intervention also served as a cross-sectional study of disease prevalence in regions where data on NCD is still very limited.
Outlook
In the final phase of the TB TRIAGE+ project, a series of analyses will be completed, e.g. on hardware dependency of CAD, assessment of subclinical TB, comparison of CD4 tests, or spatial mapping of NCD prevalence.
The primary focus, however, is the real-world analysis of the TB TRIAGE+ TRIAL on the effectiveness and cost-effectiveness of screening strategies that include either CAD only or CAD in combination with CRP. These results will be presented to both national and international stakeholders and have the potential to inform policy.
