Munachiso Elekwa, Sydani Institute for Research and Innovation

Healthcare systems in Africa face a wide range of challenges. The region carries about 25% of the global disease burden but is served by only 3% of the world’s health workforce, leaving many countries with critical shortages of doctors, nurses, and other health professionals (World Health Organization, 2023). These shortages are compounded by preventable deaths linked to poor logistics and supply chain planning, limited access to care in rural and underserved communities, and weak or fragmented health data systems.

Digital health and artificial intelligence are increasingly being explored as ways to address these gaps. Digital health refers to the use of digital technologies, such as electronic health records, mobile health tools, and data platforms, to support healthcare delivery. Artificial intelligence refers to computer systems that can analyse data, recognize patterns, and support decision-making. When applied thoughtfully, these tools can support health workers, improve early detection of disease, strengthen supply chains, and improve the accessibility of healthcare. In Southern and western Africa, early implementations are already demonstrating measurable impact. The pressing question now is not whether digital health and artificial intelligence can improve healthcare in Africa, but how these solutions can be scaled and sustained to build intelligent health systems that work at the population level.

Practical Use Cases from African Health Systems

In South Africa, tuberculosis screening has been constrained by limited radiology capacity and delays in interpreting chest X-rays, particularly in high-burden settings. To address this gap, public health programmes have deployed CAD4TB, an artificial intelligence–based tool that analyses chest X-ray images and estimates the likelihood of active tuberculosis. In real-world screening studies, CAD4TB demonstrated sensitivity of up to 96 percent and specificity of approximately 85 percent, enabling clinicians to prioritise high-risk patients for confirmatory testing and treatment while reducing unnecessary investigations (Van Ginneken et al., 2017; World Health Organization, 2023). By automating initial screening, the tool has helped improve efficiency in large-scale tuberculosis programmes, although its impact depends on access to quality imaging equipment and integration into existing workflows.

In Rwanda, delays in accessing blood products contributed significantly to maternal deaths from postpartum haemorrhage, especially in rural facilities. In response, the national health system introduced Zipline’s drone-based blood delivery service in 2016. A retrospective study analysing 12,733 blood product orders over 32 months reported a 51 percent reduction in in-hospital maternal mortality from postpartum haemorrhage after implementation, alongside a 67 percent reduction in blood product expirations within one year (The Lancet Global Health, 2023). Median delivery time fell to 41 minutes compared with 139 minutes by road, demonstrating how digitally enabled logistics can strengthen emergency care when aligned with national health system needs.

What it will Take to Scale and Sustain Impact

Across Africa, digital health and artificial intelligence solutions are still largely operating at pilot or sub-national level. While pockets of progress exist, most health systems are not yet structurally ready to support scale. This matters because intelligent tools only perform well when the systems around them are strong. Scaling is therefore not a technical problem alone, but a health systems challenge. To move from isolated successes to population-level impact, a few foundational issues must be addressed first.

• Digitization of health data: Across most African health systems, particularly in low- and lower-middle-income countries, patient records are still paper-based or spread across disconnected systems. This leads to incomplete and inconsistent data, limiting service planning and clinical decision-making. Artificial intelligence tools rely on large volumes of structured, high-quality data to perform well. Without digitized records, interoperable systems, and standardized data capture, these tools cannot learn reliable patterns or be safely scaled.
• Data governance and privacy: Data governance frameworks in many African settings are underdeveloped. Some countries have introduced data protection laws and digital health strategies, yet enforcement, clarity on data ownership, and guidance on secondary data use for public health and research are often limited. This creates uncertainty for patients, implementers, and funders alike. Governments must lead by establishing clear, enforceable rules that protect privacy while enabling responsible data use for system planning and innovation.
• Capacity Building and Infrastructure: Reliable power, internet connectivity, and basic digital tools remain unevenly distributed, particularly in rural areas. Beyond infrastructure, health workers need targeted training to use digital systems confidently and integrate them into everyday workflows. There is also a need for more university courses and funded research focused specifically on AI for health, to build a workforce that understands both health systems and data-driven technologies. AI tools are most effective when they support, rather than complicate, clinical and operational decision-making.
• Alignment with national health priorities: Digital health and AI initiatives that are not embedded within national strategies, financing mechanisms, and service delivery models rarely last beyond donor funding cycles. Policymakers must ensure that digital solutions address clearly defined system needs, funders should support integration rather than isolated innovation, and implementers must design with scale, sustainability, and public-sector realities in mind.

The future of intelligent health systems in Africa will not be determined by technology alone, but by the strength of the systems that surround it. Digital health and artificial intelligence will only deliver lasting impact in Africa if countries invest first in the systems that allow these tools to work at scale.

References

1. World Health Organization. (2023). Global health workforce statistics. World Health Organization. https://www.who.int/data/gho/data/themes/health-workforce
2. Nzimande, N., Dunbar, R., Naidoo, P., et al. (2025). Performance of CAD4TB artificial intelligence technology in tuberculosis screening programmes among the adult population in South Africa and Lesotho. Journal of Clinical Tuberculosis and Other Mycobacterial Diseases, 40, 100540. https://doi.org/10.1016/j.jctube.2025.100540
3. Nisingizwe, M. P., Niragire, F., Ndahindwa, V., et al. (2022). Effect of unmanned aerial vehicle (drone) delivery on blood product delivery time and wastage in Rwanda: A retrospective, cross-sectional study and time series analysis. The Lancet Global Health, 10(4), e564–e569. https://doi.org/10.1016/S2214-109X(22)00047-5