AI-POWERED COMMUNITY DIGITAL HEALTH COMPANION (CDHC) “VIRTUAL DOCTOR” CASE STUDY: BUKEDI SUB-REGION
No Thumbnail Available
Date
2026-09-03
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
ABSTRACT
This study focused on the design and development of an AI-powered Community Digital Health Companion (CDHC), referred to as the “Virtual Doctor,” to improve access to health information and basic symptom guidance among rural communities in the Bukedi Sub-Region of Eastern Uganda. The study was motivated by challenges including delayed access to health information, shortage of healthcare personnel, long distances to health facilities, inconsistent informal health advice, language and literacy barriers, and limited availability of digital health tools adapted to local contexts.
The main objective of the study was to develop an AI-powered CDHC “Virtual Doctor” that provides localized and accessible healthcare information to rural communities in the Bukedi Sub-Region. The study adopted a Design Science Research approach combined with Rapid Application Development (RAD). Data were collected using questionnaires, interviews, observation and document analysis involving community members, healthcare professionals and IT experts. The system was implemented using React for the frontend, Flask and Python for the backend, PostgreSQL for database management, and the Anthropic Claude API for conversational intelligence.
The developed system provided user registration and authentication, conversational symptom consultation, AI-powered health guidance, triage classification, referral recommendations, consultation history, feedback collection, administrative content management and emergency keyword detection. Unit testing, integration testing and user acceptance testing demonstrated that the system was functional, usable, responsive and generally acceptable to users and healthcare stakeholders.
The study concluded that conversational AI can provide useful supplementary support for improving access to timely health information in underserved rural communities. However, limitations included internet dependency, limited localized health datasets, variability in AI responses, resource constraints and limited geographical coverage. The study recommends multilingual support, offline or low-bandwidth functionality, integration with formal health systems, expansion of the health knowledge base, enhanced AI safety monitoring and voice-based interaction.