Opinion: Digital tool to help countries leapfrog via AI-driven health solutions

Many low- and middle-income countries have the potential to leapfrog high-income countries in their adoption of artificial intelligence-enabled technologies in health.

Digital technologies and ICT solutions such as e-banking, e-commerce, and even blockchain have often been adopted faster and more comprehensively in LMICs than in high-income countries. With the accelerated digital transformation following COVID-19, adoption of health technologies is likely to follow this trend.

Today, we face a convergence between the pandemic and pre-existing conditions such as noncommunicable diseases, all driven by political, economic, and social factors. Indeed, the pandemic has created a void in which other diseases are left potentially undiagnosed and untreated. This makes the need for a digital and data-driven transformation in health even more critical.

In addition, the systemic challenges that LMICs are grappling with — such as shortages of health workers and infrastructure, and rapid urbanization — emphasize how such countries have the most to gain from new technologies. COVID-19 has made it clear that managing the health of a population is now data-dependent.

Yet, most countries still need to build adequate and usable health datasets. A lack of sufficient skills in digital technology and data science among workforces, leaders, and populations at large is compounding this challenge, as is the absence of strong governance and regulatory systems to safely deploy AI solutions in health.

It’s now imperative for policymakers to make the investments needed to create environments that enable successful deployment of AI-driven solutions in their health systems. If not, existing health inequities risk widening even further.

However, it’s difficult to know where to invest resources in a way that creates impacts on the largest number of people. This was the genesis for the ITU/UNESCO Broadband Commission Working Group on Digital and AI in Health's “Reimagining Global Health through Artificial Intelligence: The Roadmap to AI Maturity” report.

In it, a road map defining six areas that countries should invest in to advance their readiness to deploy AI in health is proposed. These six areas are:

1. People and workforce: Countries should build data science and digital skills in their people and workforce by prioritizing these disciplines in formal education and on-the-job training.

2. Data and technology infrastructure: Countries need to prioritize the enablers behind AI in health, including broadband connectivity for all, robust technology architecture, access to quality and representative data, data privacy and security, data stewardship, interoperability, and fair, transparent, and explainable algorithms and AI models.

3. Governance and regulatory systems: Strong leadership is critical to establish the robust governance structures and regulations necessary to ensure AI innovation targets national health priorities. The foundations of good AI governance include a national strategy and budget development, clear costing and implementation plans, regulations that preserve privacy and security and put people first while balancing innovation, and integrating human rights.

4. Design and processes: Existing national health systems and health workflows are not always ready to integrate AI solutions. Stakeholders may want to perform gap analyses for technical and user requirements and collaborate broadly to ensure successful integration. All design of AI-driven solutions should be needs-driven and human-centric.

5. Partnerships and stakeholders: Countries must drive effective, goal-oriented public-private partnerships and data collaboratives, and nurture relationships with local partners, innovators, and patient organizations. Participation in international working groups and task forces enables access to best practices. Perhaps most important of all, progress toward AI maturity requires high-level political support from different ministries and the head of state.

6. Business models: Innovative and sustainable business models are essential. Countries can foster diverse funding mechanisms and develop incentive mechanisms, experiment with novel pricing models and monetization strategies, and advance the use of innovative financing mechanisms for social impact.

To make the road map toward AI maturity in health actionable, the Novartis Foundation has translated it into a freely-available and user-friendly online AI in Health Maturity assessment tool. The tool helps countries assess their readiness to deploy AI in health, and pinpoints areas that need further strengthening to realize the full potential of these new technologies.

The tool will support more countries in being able to tap into the successes of AI in health, many of which are already being seen. For example, in rural areas of Rwanda, one doctor may serve as many as 60,000 people.

But through a government partnership with Babylon Health, called Babyl Rwanda, every person aged over 12 is now able to access digital health consultations, which are supported by an AI-powered triage and symptom checker platform. More than 30% of Rwanda’s adult population has signed up.

In India, hospitals are using AI to predict a patient’s heart attack risk seven years before it might happen. This risk score, combined with local patient data, largely outperforms the traditionally used cardiovascular risk scores that were based on data on Caucasian populations only.

Other countries will be able to see similar gains in the health care space once they can ready themselves for innovative AI deployment. With the launch of this AI in Health maturity assessment tool, our ultimate goal is to ensure no country is left behind and to create an environment that allows every country to realize the potential of AI technology in health.

Visit the HealthTech Dialogue Hub series for more coverage on how to catalyze AI-driven solutions for health and care delivery. You can join the conversation using the hashtag #HealthTech.