Artificial Intelligence in Health Systems: What’s Missing in Latin America and the Caribbean?

Latin America and the Caribbean faces a serious quality problem in their health systems—something unacceptable for a middle-income region. The truth is, there are deaths that could be avoided, services that could be improved, and care that could be delivered with greater dignity.

In the region, investment in health is significantly lower than in OECD countries. Average public health spending amounts to USD 775 per capita, less than a fifth of the OECD average (USD 4,075). The number of nurses per 1,000 people is also three times lower than in OECD countries. And these are just a few indicators of the gap that must be closed.

Limited and low-quality access to primary care has direct consequences on the health and well-being of populations. About 40% of deaths from noncommunicable diseases in the region occur among people under the age of 70, compared to 29% in Europe and Central Asia. In addition, stunting rates remain a persistent challenge, and there is a considerable obesity issue.

Challenges in healthcare access and quality affect people in unequal ways. In Guatemala and Honduras, stunting in children under 5 years old exceeds 40% and 15% respectively in some areas, while in Chile and Brazil it is under 10%. Inequality within countries is also high: In Guatemala, stunting affects significantly more children under five in rural areas than in urban settings.

Improving primary health care is a key solution to ensure access to quality services for all. In facing the region’s challenges, artificial intelligence (AI) offers an opportunity to close gaps more quickly and opens a space for innovation in three key areas:

• Prevention: AI can help identify risk patterns and promote preventive actions. For example, it can contribute to anticipate disease outbreaks and support informed decisions to mitigate their impact.
• Diagnosis: AI can also improve the accuracy and speed of disease diagnosis. This includes contributing to the analyses of medical images such as X-rays, CT scans, and MRIs—which is beneficial in rural areas where access to specialists is more often limited.
• Management: AI can support resource allocation by predicting staffing needs and optimizing the distribution of medications, supplies, and vaccines.

Most importantly, these technologies can improve people’s experience within the healthcare system. For example, AI-driven processes using photos can detect vision problems in patients with diabetes. If there’s any risk, the patient is quickly referred to a specialist—reducing wait times and the risk of blindness.

The road ahead

Although some health facilities in Latin America and the Caribbean are already adopting new technologies, public policy needs to focus on improving the experience for everyone—not just a few. And for that, there are several pending tasks:

First, the digital divide remains unacceptably high. More than half of rural households still lack reliable internet access, which limits the use of digital services—including for booking appointments or renewing prescriptions online.

Second, we’re still relying on pen and paper. Many hospitals and health centers use paper-based medical records, complicating the integration of digital solutions. In Costa Rica, Chile, and Mexico, on average, only 65% of primary health care centers use electronic medical records—compared to 93% in OECD countries. And that figure reflects a few countries; the reality could be even worse. Digital information systems tend to be isolated, fragmented, and underused.

Third, the digital skills gap needs to be addressed. Only one-third of workers in the region use digital tools in their jobs, compared to more than half in OECD countries.

What’s needed to make AI work in health care?

Digital in health care is not new—it’s already driving significant changes. However, establishing conditions to ensure effective use of AI is crucial:

• Data must reflect the reality of Latin America and the Caribbean. AI models trained on data from other regions won’t always reflect the local context, which can limit the ability to identify appropriate solutions.
• Privacy. Having access to lots of data about an individual can help deliver the right services more efficiently—but irresponsible use of this data can have serious consequences.
• Interoperability. Interoperable information systems can exchange data, making it easier to apply AI solutions in the health sector.
• The right balance between technology and the human factor. Digital tools should enhance the work of healthcare professionals. Human expertise and technology must work hand-in-hand to deliver high-quality care at the lowest possible cost.

As health systems continue to evolve, digital tools and AI will become integrated into how care is delivered. But reaching this requires a firm commitment to improving access and quality—starting with primary care. The Alliance for Primary Health Care in the Americas, a partnership between the World Bank, PAHO, and the IDB, aims to drive this digital transformation across the region. If we overcome today’s challenges together, millions of people will be able to access a health system that is effective, dignified, and high quality.

Source: blogs.worldbank.org