In Nigeria, Community Health Extension Workers (CHEWs) are essential to providing frontline health care, particularly in rural and underserved areas where physician shortages are most pronounced. However, they face a significant challenge in their limited access to real-time, comprehensive information that could support timely, accurate decision-making for a wide range of health concerns. Studies have shown that poor utilization of primary health care facilities in Nigeria is often attributed to perceived poor quality care, with CHEWs sometimes struggling to access the necessary information to manage patients effectively (Luka-Lawal et al. 2020). Existing research indicates that CHEWs can deliver safe and effective care when given adequate support. A study conducted in South-South Nigeria found that CHEWs had good results when given access to physician consultation via telephone (Ordinioha and Onyenaporo 2010). These findings suggest that providing decision-support resources enhances the ability of CHEWs to manage cases at the community level.

Given the demonstrated benefits of telephonic physician support, this study explores whether a system driven by artificial intelligence (AI) could serve a similar role in enhancing CHEW decision-making capabilities. Unlike physician-supervised telephone consultations, which require ongoing human resources, a Large Language Model (LLM)-based system could provide scalable, on-demand access to health care knowledge without relying on direct physician availability. By integrating LLMs into the CHEW workflow, this study seeks to evaluate the feasibility of AI-driven decision support in improving patient outcomes, reducing unnecessary referrals, and strengthening primary health care service delivery in Nigeria.