LLMs: A Promising New Tool for Improving Healthcare in Low-Resource Nations

Abstract

This paper explores the potential of large language models (LLMs) in addressing healthcare inequalities, particularly in underserved nations with provider shortages, limited resources, and funding constraints. The UN’s Sustainable Development Goal 3 aims to achieve health and well-being for all, but disparities persist. Recent advancements in LLMs, exemplified by OpenAI’s GPT-4, demonstrate their ability to surpass human performance on certain medical exams, offering an opportunity to augment the limited number of physicians and meet unmet healthcare needs affordably. By customizing LLMs for healthcare, various applications become possible, including automating patient screening, assisting with diagnosis and treatment, enabling virtual health assistants to track health indicators and educate communities, supporting frontline health workers in addressing basic healthcare needs, translating medical knowledge for accessibility, and aiding multilingual healthcare providers. However, challenges such as inadequate data and infrastructure, risks of bias and privacy breaches, cost and access barriers, and security threats must be addressed to ensure that the benefits of LLMs reach communities facing the greatest challenges and threats. This paper highlights the potential benefits, challenges, and the need for collaboration to harness the power of LLMs effectively in healthcare and contribute to achieving the UN’s Sustainable Development Goal 3.