Tuberculosis is a chronic infectious disease caused by bacteria belonging to the Mycobacterium tuberculosiscomplex (MTBC), comprising mainly Mycobacterium tuberculosis, but also M. bovis,M. africanum,M. cannetii,M. orygis, and M. caprae [1][2]. These mycobacteria are characterized by a lipid-rich cell wall, which confers increased resistance to intracellular destruction mechanisms and allows them to persist for long periods within the host’s macrophages [3]-[5].

Despite scientific and therapeutic advances made during the 20th century, tuberculosis remains a major global public health problem [1][6]. The HIV pandemic has profoundly altered the epidemiology of the disease, increasing the risk of progression to active tuberculosis in infected individuals, while the COVID-19 pandemic has disrupted diagnostic and treatment services in many countries [7][8]. According to the World Health Organization, 10.8 million new cases and 1.25 million deaths were reported in 2023, making tuberculosis the second leading cause of infectious mortality worldwide [2][8][9].

While the introduction of molecular biology tools has led to significant advances in rapid diagnosis and the detection of certain resistances, their impact remains limited in low-resource countries [10]-[13]. Economic, organizational, and structural constraints on health systems hinder equitable access to these technologies [8][9]. In this context, it seems essential to rethink tuberculosis diagnostic strategies by favoring integrated, innovative approaches that are adapted to local realities [9]. The aim of this review is to analyze the limitations of current approaches and propose sustainable diagnostic strategies for developing countries.

This narrative review is based on a critical analysis of the international scientific literature, combined with the authors’ operational experience in microbiological diagnosis of tuberculosis in several resource-limited African settings. The methodological objective was to identify, evaluate, and put into perspective innovative diagnostic strategies likely to improve tuberculosis screening, diagnosis, and follow-up in developing countries, while taking into account the structural, economic, and human constraints specific to these contexts.

An in-depth literature search was conducted in the PubMed/MEDLINE, Scopus, and Web of Science databases to identify relevant publications published between January 2000 and December 2024. The search strategy combined free terms and MeSH descriptors related to tuberculosis and its diagnosis, including tuberculosis, diagnosis, molecular diagnostics, Xpert MTB/RIF, lipoarabinomannan, point-of-care testing, artificial intelligence, chest radiography, low-income countries, and sub-Saharan Africa. Boolean operators AND and OR were used to optimize the sensitivity and specificity of the search.

Publications were selected based on their scientific relevance and applicability to low- and middle-income country contexts. Original articles, systematic reviews, meta-analyses, observational studies, clinical trials, and international recommendations on tuberculosis diagnostic strategies published in English or French were included. Purely basic studies with no direct diagnostic implications were excluded, as were non-peer-reviewed documents, with the exception of strategic reports and guidelines issued by the World Health Organization.

The selected articles were read in full, and data were extracted on diagnostic performance (sensitivity, specificity, predictive value), turnaround time, technical and logistical requirements, costs, and operational feasibility in resource-limited settings. Particular attention was paid to the potential impact of these tools on national tuberculosis control strategies, particularly in rural and peripheral areas.

Finally, the data from the literature were compared with the authors’ field experience in reference laboratories and peripheral health facilities, enabling an integrative analysis of the strengths and limitations of existing approaches. This approach led to the proposal of an integrated and sustainable diagnostic strategy, combining conventional microbiological tools, molecular tests, biomarkers, artificial intelligence, and community strategies, formalized in the form of an algorithm adapted to the realities of low-resource countries.

In most low- and middle-income countries, tuberculosis diagnosis still relies heavily on conventional methods, particularly microscopy of smears stained with Ziehl-Neelsen or auramine. These techniques have the advantage of being inexpensive, simple to implement, and easily deployable on a large scale, which explains their central place in national tuberculosis control programs [2].

However, microscopy has significant limitations. Its sensitivity is low, particularly in paucibacillary patients, children, people living with HIV, and in extrapulmonary forms of tuberculosis [4][9]. In addition, its performance is highly dependent on the quality of samples, the expertise of staff, and the existence of quality assurance systems, which are often inadequate in resource-limited settings. The presence of non-tuberculous mycobacteria in certain regions can also lead to false-positive results, compromising the clinical relevance of the diagnosis [10].

Mycobacterial culture is the gold standard for definitive diagnosis [2]. It allows microbiological confirmation, identification of MTBC species, and testing for sensitivity to anti-tuberculosis drugs. However, this method has major limitations. The slow growth rate of mycobacteria means that results can take up to six to eight weeks to come back, delaying the initiation of appropriate treatment or therapeutic adjustment in cases of resistance [2][9].

In addition, culture requires high-level biosafety infrastructure, generally a P3 containment laboratory, to prevent the risk of nosocomial transmission [11][12]. These technical and financial requirements exceed the capabilities of many developing countries, sometimes leading to the initiation of anti-tuberculosis treatment based solely on clinical or radiological findings, with an increased risk of diagnostic errors, treatment failure, and selection of resistant strains [10].

Conventional tuberculosis diagnostic strategies, although essential, have shown their limitations in terms of sensitivity, speed, and accessibility in developing countries. Molecular biology tools have revolutionized diagnosis, but their impact remains constrained by economic and organizational factors.

The adoption of innovative, integrated diagnostic strategies adapted to local realities appears essential to improving tuberculosis control. By combining technological innovations, health system strengthening, and community-based approaches, it is possible to move toward earlier, more equitable, and more sustainable diagnosis, which is essential to achieving global tuberculosis control goals.

We would like to thank all our colleagues at the University of Bangui and the National Laboratory of Clinical Biology and Public Health who have contributed in one way or another to improving the quality of this work.

This study was conducted in accordance with the HELSINKI declarations on respect for human dignity.