Analysis of tuberculosis prevalence surveys in Africa: evidence to change the status quo in case finding.

Introduction: Lesotho has among the world's highest TB prevalences and the national program finding only 50% of incident cases. Lesotho underwent its first TB prevalence study in 2019 in reviewing this study, the authors recognized that the standard measures for screening and diagnosis TB often failed to detect the presence of culture positive pulmonary TB. This review seeks to determine if the systematic under detection of culture positive TB by standard methods is true among prevalence surveys in Africa. Methods: We conducted a review to identify TB prevalence surveys conducted in Africa from 2011 to 2024. Seventeen surveys meeting inclusion criteria were identified. For each survey, we extracted the proportion of culture-confirmed TB cases that were also identified by symptom screening, chest X-ray (CXR), sputum smear microscopy, and GeneXpert MTB/RIF. We additionally tabulated national HIV prevalence, HIV co-infection rates reported by national TB programs, and HIV co-infection rates documented in the prevalence surveys themselves. Results: TB prevalence across the 17 surveyed countries ranged from 119 per 100,000 population in Rwanda to 737 per 100,000 in South Africa, with Lesotho (581), Tanzania (590), Zambia (638), Kenya (558), Nigeria (524), Uganda (401), Namibia (465), Malawi (452), Ghana (356), Eswatini (352), Mozambique (334), Zimbabwe (344), Gambia (212), Sudan (183), and Ethiopia (183) recording intermediate burdens. Symptom screening missed an average of 50% of culture-confirmed pulmonary TB cases across all surveys. Smear microscopy missed approximately 60% of bacteriologically confirmed cases. GeneXpert, while more sensitive, still missed an average of 30-35% of culture-confirmed cases even under rigorous study conditions. In contrast, CXR demonstrated concordance exceeding 80% with culture-confirmed TB across almost all surveyed countries. HIV co-infection among TB patients in prevalence surveys was consistently lower than rates reported in national TB programs, suggesting systematic underrepresentation of HIV-positive individuals in survey populations. Conclusion: Symptoms-based screening and GeneXpert alone are insufficient to detect the majority of people with active, bacteriologically confirmed TB. Chest X-ray, with markedly higher sensitivity than symptom screening or molecular testing under real-world survey conditions, should be repositioned as a primary screening tool in high-burden settings.