AB117. The role of AI and digital technologies in access to diagnostics in crisis-prone areas: a scoping review

Background: Health systems across the globe face mounting challenges due to compounded crises such as poverty, infectious disease outbreaks, climate disasters, and reductions in external aid. Fragile and displacement-affected border regions across low- and middle-income countries (LMICs) exemplify these pressures, where large-scale population movements and constrained access to affordable healthcare place sustained stress on diagnostic and treatment services, particularly for communicable diseases. Artificial intelligence (AI) and digital technologies are increasingly recognized as transformative tools for improving access to health services in LMICs, particularly in diagnostics, telemedicine, supply chain management, and epidemic forecasting. These technologies, when adapted to local contexts, can help overcome systemic barriers and improve health outcomes in resource-constrained settings.

Methods: This scoping review maps the landscape of AI and digital technologies used for diagnostics, with a focus on how these technologies engage with challenges faced by displaced and migrant populations. It synthesizes evidence on technology characteristics, implementation contexts, stages of deployment, effectiveness, and integration with health systems to inform the development of sustainable and decentralized diagnostic systems. The review delves deeper into countries that comprise the Association of Southeast Asian Nations (ASEAN) as they face complex health crises driven by poverty, climate change, infectious diseases, natural disasters, and the withdrawal of external aid. The Thailand-Myanmar border region, in particular, is challenged by a large influx of displaced people, the termination of external assistance, and reduced free medical services. These factors strain health systems and highlight the urgent need for innovative, resilient approaches.

Results: The results show that human-centered design approaches—actively involving end-users such as patients and frontline health workers—are critical for effective and ethical deployment, especially among marginalized populations. However, challenges remain, such as limited high-quality data for AI model training, weak infrastructure (electricity, connectivity), low AI literacy, and the deployment of “black-box” AI systems with low explainability and community trust.

Conclusions: This scoping review will inform a project that will (I) identify issues surrounding access to diagnostics on the National Essential Diagnostics List (NEDL) for populations in crisis-prone areas, particularly at the Myanmar border in Thailand; and (II) explore how AI and other digital technologies could support the development of self-reliant systems for health in this context. This study underlines the importance of collaboratively designing AI and digital solutions with local stakeholders that are informed by cultural contexts, and grounded in principles of equity, resilience, and sustainability.

Keywords: Refugee health; digital health; artificial intelligence (AI); National Essential Diagnostics List (NEDL); health systems