Artificial Intelligence and Machine Learning for Asthma Care in Nigeria and Sub-Saharan Africa: Evidence, Opportunities, and Adoption Barriers

Asthma affects an estimated 363 million people worldwide and caused approximately 442,000 deaths in 2023. The burden of this disease falls disproportionately on low-income and middle-income countries where the disease is mostly underdiagnosed and under-treated. Artificial intelligence (AI) and machine learning (ML) have been presented as tools for earlier diagnosis, personalised treatment, exacerbation prediction, and remote monitoring. However, much of the supporting research originates from high-income settings. Our review evaluates what the supporting evidence offers for asthma care in Nigeria and the African region. We conducted a literature search across PubMed, IEEE Xplore, Google Scholar, and similar databases, with the resulting review library consisting of 170 screened records. 116 records were reviewed in full, while 31 were retained for in-depth review of their techniques, data sources, algorithms, conclusions, and implications. Four themes emerged: an expanding but unevenly equipped African AI ecosystem; AI/ML tools for diagnosis, exacerbation prediction, and digital monitoring developed and validated almost entirely outside Africa; persistent non-technical barriers to asthma care in sub-Saharan Africa, including under-diagnosis, restricted access to medications, and insufficient awareness among carers and educators; and unresolved ethical, legal, and data governance issues. The reviewed studies demonstrate technical feasibility in high-income settings but reveal a lack of African validation. We conclude that AI and ML can enhance asthma care in Nigeria only if models are validated on African data, integrated into existing care frameworks, and governed by explicit regulations. We therefore propose seven recommendations for policymakers, clinicians, and researchers.