UNLOCKING AI'S POTENTIAL IN AFRICAN HEALTHCARE: THE ESSENTIALITY OF A CLEAR LEGAL FRAMEWORK

UNLOCKING AI'S POTENTIAL IN AFRICAN HEALTHCARE: THE ESSENTIALITY OF A CLEAR LEGAL FRAMEWORK

INTRODUCTION

There is no doubt that the use of Artificial Intelligence (AI) in recent years is transforming the healthcare system and opening up a whole new world of medical possibilities. This is true notjust in developed and higher-income countries but around the world generally, particularly in Africa, which faces significant challenges in its healthcare system. AI is being used across the continent to improve accuracy in diagnosis, streamline clinical workflows and enhance patient care, from dataset management in Morocco and genomic analysis in South Africa, to medical image analysis in Ghana and COVID-19 tracking in Ethiopia.

 

Accordingly, the impact of AI in transforming the continent's healthcare systems is undeniable, but this possibility comes with some challenges. However, realizing this full prospect requires more than just technological advancements; it necessitates a robust legal framework that fosters innovation whilst safeguarding patient rights and safety, data privacy and public health. Several countries have begun developing a governance policy and/or legal framework for the implementation of AI in different sectors particularly in healthcare while some are yet to take action.

 

This article will examine how AI is already transforming the African healthcare system, the current state of AI regulations and the crucial role of clear legal frameworks in harnessing AI's potential in the continent's healthcare.

 

HOW AI IS TRANSFORMING AFRICAN HEALTHCARE

AI has recently revolutionized various facets of healthcare in Africa, driving innovation and progress in multiple areas. The World Health Organization's Regional Office for Africaacknowledges AI's vast potential to transform healthcare outcomes on the continent. They stress the need for Africa to prepare for this transformation and highlight various AI innovations strengthening health systems. Similarly, the Africa Center for Disease Control and Prevention emphasizes this.

 

In Kenya, for instance, a mobile-optimized AI system has been developed to assess gestational age and fetal malpresentation, with results comparable to clinical standards and high accuracy across operators and devices. AI-supported echocardiography has enhanced heart function diagnosis and monitoring, addressing limited access to expert care in Lesotho.

 

In South Africa, deep learning algorithms are improving health workers' abilities to diagnose HIV using lateral flow tests. Additionally, a predictive system is being used to enhance human resource planning, forecasting health worker retention in public service. Meanwhile, a modern review of AI in disease diagnosis showed promising results, with high accuracy and reliability across various metrics, including sensitivity, specificity and precision.

 

To assess emergency obstetric care and enhance maternal health outcomes, Nigeria's OnTime Consortium in partnership with Google created an AI-driven tool, harnessing the Directions API. Similarly, a digital tool, Mobile Application for Tuberculosis Screening (MATS) is making a significant impact in the country by streamlining tuberculosis screening and diagnosis, and connecting patients to essential care. Additionally, AI has been leveraged in Nigeria's pharmaceutical industry, with a notable example being an app developed by five high school girls using MIT's open-source software to detect counterfeit drugs.

 

In Zambia, a study by Bellemo et al demonstrated AI's potential in diagnosing diabetic retinopathy, showing promising results comparable to human assessments. Likewise, the Delft Institute's CAD4TB software has shown encouraging results in pilot studies in Tanzania and Zambia, accurately diagnosing pulmonary tuberculosis from chest radiographs and performing on par with human experts

 

Rwanda Innovation Fund and Viebeg Technologies are also using AI to enhance healthcare accessibility in Rwanda. The initiative streamlines medical supply procurement, enabling healthcare facilities to maintain precise inventory levels and connect directly with manufacturers, thereby reducing costs by up to 40% and eliminating intermediaries.

 

All these examples demonstrate AI's transformative potential in African healthcare. However, it should be noted that this prospect comes with some challenges ranging from limited access to quality data to infrastructure constraints, sustainability, cultural and language barriers, skills gaps and funding. Therefore, to fully realize this potential, a comprehensive legal framework and/or governance policy is crucial in the continent. 

 

CURRENT STATE OF AI REGULATION

Globally, the regulation of AI generally is still in its fragmented stages, most countries have begun developing a governance policy and regulations for the implementation of AI in different sectors including the health sector. While some still depend on ancillary regulatory frameworks such as data protection, digital health, consumer protection and intellectual property, there have been significant improvements in recent years.

 

The European Union has taken a pioneering step by enacting the EU AI Act, the world's first comprehensive AI law. This legislation regulates AI systems across various sectors based on their risk level. To ensure effective implementation, the European Parliament has formed a joint working group to oversee the process, iron out operational details, and clarify key provisions before the Act takes effect on the 1st of August, 2024.

 

In contrast, China has adopted a sectoral approach, introducing regulations for specific AI applications such as internet recommendation algorithms, deep synthesis technology and generative AI. This targeted approach enables China to address particular challenges, develop tailored rules and gradually build its regulatory expertise with each new policy.

 

In the United States, a range of federal laws, bills, executive orders and frameworks address various aspects of AI, from administrative initiatives to regulating specific applications like deepfakes and foundation models. A significant milestone was Executive Order 14110 in 2023, which introduced a comprehensive suite of AI initiatives. The FDA oversees AI in healthcare, focusing on software as a medical device.

 

While there is no specific AI regulation in Australia, an AI ethics framework exists. Australian regulators are actively enforcing relevant laws in the AI space through the Digital Platform Regulators Forum (DP-REG), comprising ACCC, ACMA, eSafety and OAIC. The DP-REG focuses on digital platform technologies, including AI issues.

 

In Africa, the AI regulatory landscape is still evolving like in other parts of the world. A study covering 12 countries, including Nigeria, South Africa, and Kenya, found no specific AI legislation, with AI adoption guided by related frameworks like data protection and intellectual property. Nevertheless, there is a growing need for AI-focused regulations. 

 

For instance, in Nigeria, existing laws like the National Health Act, Cybercrimes Act and National Data Protection Act provide some oversight, but a report recently revealed that a draft national AI policy is forthcoming. While South Africa is considered to be lagging in AI development and regulation, Kenya has introduced a bill to establish the Kenya Robotics and Artificial Intelligence Society, requiring registration and licensing for robotics and AI businesses.

 

Egypt's National Council for Artificial Intelligence (NCAI) developed a phased AI adoption strategy, starting with regional cooperation. The initial phase focuses on establishing Egypt as a key player in regional AI collaboration focusing on investment, public awareness and AI capability development. Later phases will prioritize harnessing AI's growth potential domestically.

 

Regardless, just recently, the African Union adopted the Continental Artificial Intelligence Strategy to accelerate digital transformation, harness AI's potential, and address challenges in education, health, agriculture and governance. The strategy focuses on developing human capital, fostering innovation and creating a conducive regulatory environment in the continent.

 

Additionally, the African Union and China have agreed to cooperate on AI research and development, including policy dialogue, technology investment, industrial cooperation, talent exchange and data security. The African Commission on Human and Peoples' Rights has also adopted Resolution 473, calling for a human-rights-centered approach to AI governance, regional regulatory frameworks and guidelines to address the impact of AI on human rights in Africa. 

 

THE NEED FOR CLEAR LEGAL FRAMEWORK

Recognizing the importance of regulating AI in healthcare, the World Health Organization (WHO) has outlined six key considerations to mitigate the risks posed by the use of AI in healthcare. Although the considerations are not guidance or policy but are intended as a resource for relevant stakeholders in medical device ecosystems, including manufacturers, regulatorsand healthcare practitioners, to navigate the safe development and deployment of AI-powered medical devices.

 

WHO stresses the value of transparency and documentation to build trust in AI systems among developers, manufacturers and end-users. To achieve this, WHO recommends comprehensive documentation of the product life cycle and development processes. This approach helps prevent data biases and manipulation. Further, WHO advises a holistic risk-based approach throughout the product life cycle, encompassing pre- and post-market deployment. This involves addressing critical issues such as intended use, continuous learning, human intervention, training models and cybersecurity threats.

 

WHO also emphasizes the need to clearly define an AI system's intended use, considering its dependence on code, training data, setting and user interaction. This ensures safe and effective performance. Likewise, WHO recommends validating AI systems through external data validation, ensuring transparency in data usage and quality. This minimizes risks to patients by preventing biased or poor-quality data from compromising AI models.

 

Moreover, WHO stresses the essence of rigorous evaluation systems to ensure data quality before AI systems are released. This prevents amplification of biases and ensures compliance with data protection and privacy laws. Yet, WHO encourages collaboration among stakeholders, including developers, manufacturers, healthcare practitioners, patients and policymakers, to enhance the safety and quality of AI technologies.

 

Based on WHO's key regulatory considerations, it's clear that a robust legal framework for AI is essential in Africa, where healthcare systems face unique challenges. The use of AI algorithms in healthcare, which rely on large amounts of patient data for training and decision-making, highlights the need for proper regulations. Without such regulations, there's a risk of sensitive medical information being misused or accessed without authorization.

 

A well-established legal framework can mitigate this risk by ensuring patient privacy is protected through measures like data anonymization, purpose limitation and transparency in data usage. Additionally, it ensures informed consent, allowing patients to understand how their data will be used so they can opt in or out. A framework will clarify responsibilities among healthcare providers, AI developers and regulators thus promoting accountability and transparency throughout the AI deployment process.

 

Also, in Africa with a diverse healthcare landscape, AI algorithms can perpetuate biases in historical data, making fairness a critical concern. Legal guidelines can address this by mandating fairness assessments, bias mitigation and regular audits. Certification processes can ensure AI-driven healthcare solutions meet rigorous safety standards. A legal framework will guarantee quality benchmarks are met, which is vital in the continent where healthcare resources are unevenly distributed.

 

Additionally, striking a balance between promoting AI innovation and safeguarding intellectual property rights is crucial. Effective legal provisions can foster research and development while preventing monopolies that limit access to AI-driven healthcare solutions. To address the complexities of Africa's varied legal landscapes, collaboration among countries is necessary to standardize AI regulations. Regional organizations can play a key role in facilitating knowledge exchange, capacity building and collaborative efforts to tackle cross-border challenges.

 

CONCLUSION

AI holds immense potential to transform healthcare in Africa but this potential can only be fully realized with a clear legal framework that promotes innovation while protecting patient rights, data privacy and public health. Drawing on the World Health Organization's regulatory guidelines, African nations must collaborate, foster public-private partnerships and develop certification processes to ensure AI innovation is safe and effective. Establishing a continental AI governance body and implementing robust data protection regulations will further ensure transparency, accountability, and fairness in AI deployment. By taking these steps, Africa can develop human capital and drive economic growth to improve health outcomes and enhance the quality of life across the continent. Thus, collective action is crucial for Africa to create a brighter and healthier future.

REFERENCES

Image Source: https://www.law.com/newyorklawjournal/2024/02/02/generative-ai-in-health-care-diagnosing-the-legal-landscape-for-dr-genai/

1.  National Center for Biotechnology Information. (2021). Artificial Intelligence in Healthcare: Transforming the Practice of Medicine. PubMed Central. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8285156/
2. World Bank Group. (2024). World Bank country classifications by income level for 2024-2025. Retrieved from https://blogs.worldbank.org/en/opendata/world-bank-country-classifications-by-income-level-for-2024-2025
3.  Kirigia, J. M., & Barry, S. P. (2008). Health challenges in Africa and the way forward. International Archives of Medicine, 1(27)
4.  Schwaller, F. (2023, December 1). Could AI transform health care in Africa? Deutsche Welle. Retrieved from https://amp.dw.com/en/could-ai-transform-health-care-in-africa/a-67597556
5.  Owoyemi, A., Owoyemi, J., Osiyemi, A., & Boyd, A. (2020). Artificial Intelligence for Healthcare in Africa. Frontiers in Digital Health, 2. https://doi.org/10.3389/fdgth.2020.00006
6.  Solaiman, B., & Cohen, I. G. (Eds.). (2024). Research Handbook on Health, AI and the Law. Edward Elgar Publishing. Available at: Elgaronline
7.  World Health Organization, Regional Office for Africa. (n.d.). AI-driven Innovations. Retrieved from https://innov.afro.who.int/emerging-technological-innovations/12-ai-driven-innovations
8.  Africa CDC. (2023). Digital Transformation Strategy. Retrieved from https://africacdc.org/download/digital-transformation-strategy/
9.  Gomes, R.G., Vwalika, B., Lee, C., et al. (2022). A mobile-optimized artificial intelligence system for gestational age and fetal malpresentation assessment. Nature Communications, 13, Article 194. https://doi.org/10.1038/s43856-022-00194-5.
10.  Emmanuel Firima, Lucia Gonzalez, Molulela Manthabiseng, Matumaole Bane, Blaise Lukau, Bailah Leigh, Beat A. Kaufmann, Maja Weisser, Alain Amstutz, Jasper Tromp, Niklaus Daniel Labhardt & Thilo Burkard (2024) Implementing focused echocardiography and AI-supported analysis in a population-based survey in Lesotho: implications for community-based cardiovascular disease care models.  https://www.nature.com/articles/s41440-023-01559-6
11.   Ngo, H. T., Akarapipad, P., Lee, P. W., Park, J. S., Chen, F. E., Trick, A. Y., Hsieh, K., & Wang, T. H. (2023). Rapid and Portable Quantification of HIV RNA via a Smartphone-enabled Digital CRISPR Device and Deep Learning. medRxiv. https://doi.org/10.1101/2023.05.12.23289911
12.   Kumar, Y., Koul, A., Singla, R., & Ijaz, M. F. (2022). Artificial intelligence in disease diagnosis: A systematic literature review, synthesizing framework and future research agenda. Journal of Ambient Intelligence and Humanized Computing, 14, 8459-8486 https://link.springer.com/article/10.1007/s12652-021-03612-z

 

DISCLAIMER NOTICE
This is a free educational site for your general information and enlightenment ONLY. This site, by itself, does not create a Client/Attorney relationship. Recipients are therefore advised to seek professional legal counselling to their specific situations when they do arise. Questions, comments, criticisms, suggestions, new ideas, contributions, etc are welcome. The articles here, are protected by Intellectual Property Law and Regulations and is reproduced by Medical Law Hub under permission for the Medical Law Hub community.