Implementing AI-enabled chest X-ray for community-based integrated screening for tuberculosis, chronic respiratory diseases, and cardiovascular diseases in Nigeria.
Authors
Affiliations (9)
Affiliations (9)
- TB/HIV Department, Catholic Caritas Foundation of Nigeria, Abuja, Nigeria. [email protected].
- TB/HIV Department, Catholic Caritas Foundation of Nigeria, Abuja, Nigeria.
- PRIME Data Nexus, Enugu, Nigeria.
- National Tuberculosis and Leprosy Control Programme, Abuja, Nigeria.
- Department of Internal Medicine, Alex Ekwueme Federal University Teaching Hospital Abakaliki, Abakaliki, Ebonyi State, Nigeria.
- Centers for Disease Control and Prevention, Atlanta, GA, USA.
- Stop TB Partnership, TB REACH, Geneva, Switzerland.
- Zankli Research Centre, Bingham University, Nasarawa, Nigeria.
- Centre for Tuberculosis Research, Liverpool School of Tropical Medicine, Liverpool, UK.
Abstract
Low- and middle-income countries face a growing dual burden of communicable and non-communicable diseases, while services remain vertically organised. In Nigeria, tuberculosis (TB) services are established at primary care level, whereas access to cardiovascular disease (CVD) and chronic respiratory disease (CRD) care remains limited in rural settings. Artificial intelligence (AI) enabled chest X-ray can integrate TB screening with identification of other cardiopulmonary abnormalities at community level. We describe the screening outcomes and referral cascade of a community-based, AI-enabled integrated programme. We conducted a non-randomised descriptive study using routinely collected programme data from five Local Government Areas in Ebonyi and Nasarawa States, January 2023 to December 2024. Community outreach used portable digital chest X-ray with AI software to screen individuals aged six years and above. People with presumptive TB underwent Xpert MTB/RIF (Mycobacterium tuberculosis/rifampicin) testing on the GeneXpert platform, while non-TB radiographic abnormalities were referred for further evaluation. Descriptive analyses summarised screening yield, diagnostic outcomes, and linkage to care. In total, 9,585 individuals were screened through 93 outreach activities, and 3,166 (33%) chest radiographs were flagged as abnormal by AI. Overall, 1,336 were classified as having presumptive TB, of whom 1,123 (84%) produced sputum for Xpert MTB/RIF testing. 204 were diagnosed with bacteriologically confirmed TB, and 194 (95%) were initiated on treatment. A further 199 people were clinically diagnosed with TB following radiologist and/or clinical review. Among abnormal radiographs, 2,367 (75%) showed features suggestive of CVDs or CRDs. All such individuals were referred to tertiary facilities; however, only 12% completed the referral. Programmatic adaptations supported TB linkage but had limited impact on non-TB referral completion. AI-enabled community chest X-ray screening is feasible for TB case finding in rural Nigeria and achieves high linkage to TB treatment. Chest radiography also identified many abnormalities requiring further evaluation for CVD and CRD, most not confirmed within the study. Limited decentralisation of non-communicable disease services constrains care continuity for CVDs and CRDs. Integrated screening programmes should be paired with strengthened primary healthcare capacity, complementary tools such as blood pressure measurement, and context-specific community engagement strategies.