New study examines how generative AI can assist community health workers in Rwanda
PATH and partners are conducting a research trial to assess how an AI knowledge assistant can help community health workers better triage and treat their patients.
Kigali, Rwanda, July 31, 2025—A pioneering non-interventional (silent) trial has been launched in Rwanda to assess the potential for generative artificial intelligence (GenAI) to support community health workers (CHWs) to more effectively diagnose and treat common health concerns and prevent unnecessary referrals to a health facility.
This trial is led by PATH in collaboration with a consortium of research [University of Global Health Equity, the Centre for the Fourth Industrial Revolution (C4IR) Rwanda, and the University of Birmingham], governmental [Rwanda Biomedical Centre (RBC)], and technology (Digital Umuganda) partners. CHWs began to enroll in the study in July 2025 and results will be available in September 2025.
CHWs in Rwanda are critical frontline health care workers who address basic health needs in their communities. Their standard of care for triaging, diagnosing, and treating common health complaints is guided by decision protocols that use a series of structured, preset questions. This limits the details they can collect during a patient consultation and, as a result, some health concerns are not escalated quickly enough, whereas others that could be treated at the community level are unnecessarily referred.
Aligned with Rwanda’s digital health transformation strategy, the RBC and the Ministry of Health developed a smartphone-based tool for CHWs, the Community Electronic Medical Record system (C-EMR), which guides CHWs through these structured diagnostic protocols. In parallel to the roll-out of the C-EMR, the Rwandan Government began exploring the value of AI-enabled tools for improving community health care quality.
"This study will offer crucial insights on how integrating an LLM into community health workers’ workflow could improve their decision-making capabilities by providing real-time, context-specific, and evidence-based guidance. The generated evidence will guide us in safely and effectively bringing our health workforce into the AI revolution,” said Prof. Claude Mambo Muvunyi, Director General of the Rwanda Biomedical Centre.
“This study will offer crucial insights on how integrating an LLM into community health workers’ workflow could improve their decision-making capabilities by providing real-time, context-specific, and evidence-based guidance.”— Prof Claude Mambo Muvunyi, Director General, Rwanda Biomedical Centre
In this trial, selected CHWs are using an enhanced version of the C-EMR that leverages smartphone technology to record patient interactions via the built-in microphone. A text-to-speech function creates a transcript of the interaction as the patient describes their symptoms and answers the CHW’s questions. A large language model (LLM) then reviews the transcript and produces guidance around whether to refer the patient, as well as an appropriate differential diagnosis and management plan. An expert panel will evaluate the accuracy and appropriateness of the LLM outputs.
“Through this research partnership, we are supporting Rwanda’s innovation ecosystem to build contextually appropriate AI-enabled solutions. We are laying the foundation for ensuring AI can be harnessed in a way that advances health outcomes, equitably and sustainably,” said Crystal Rugege, Managing Director, Rwanda Centre for the Fourth Industrial Revolution.
As a non-interventional trial, CHWs will not interact with the LLM or see the outputs, and, therefore, it will have no influence on their clinical decisions. However, the trial serves a critical role in generating urgently needed evidence on how popular LLMs (e.g., GPT-4, Gemini, and Claude) perform when prompted in regional African languages, like Kinyarwanda, and when asked health questions in an African context.
Prof Bilal Mateen, Chief AI Officer at PATH, said:
“Integration of GenAI into high-risk, high-reward scenarios like this one in Rwanda requires deliberate progress. This ‘silent’ trial will generate the critical data assets needed to deliver a best-in-class product and demonstrate to our government partners that the tool we’re proposing to deploy will deliver value. If all goes to plan, by the end of the year, we’ll have progressed to implementing a phase 3 trial of an LLM-based clinical decision support tool in Africa.”
This research trial is one of three being conducted by PATH and its partners to produce evidence and advance knowledge on the safe and effective use of LLM-enabled clinical decision support tools in African primary health care settings.
“We are laying the foundation for ensuring AI can be harnessed in a way that advances health outcomes, equitably and sustainably.”— Crystal Rugege, Managing Director, C4IR Rwanda
About PATH
PATH is a global nonprofit dedicated to achieving health equity. With more than 40 years of experience forging multi-sector partnerships, and with expertise in science, economics, technology, advocacy, and dozens of other specialties, PATH develops and scales up innovative solutions to the world’s most pressing health challenges. The first-ever phase 3 randomized controlled trial of an LLM-based clinical decision support tool in Africa is currently being led by PATH and its partners in Kenya.
For media inquiries, please contact: Lauren Grella, Senior Director of Marketing & Communications - media@path.org
About the University of Global Health Equity
The University of Global Health Equity (UGHE) is a university based in Rwanda that is building the next generation of global health professionals—doctors, nurses, researchers, and public health and policy experts—into leaders and changemakers who strive to deliver more equitable, quality health services for all. An initiative of Partners in Health (PIH) and led by internationally recognized faculty and staff from around the world, UGHE is an independent university that builds on PIH’s three decades of experience in delivering health services to some of the world’s poorest communities.
About the C4IR Rwanda
The Rwanda Centre for the Fourth Industrial Revolution (C4IR) was established in 2020, during the World Economic Forum Annual Meeting in Davos. Part of a network of 22 centers globally, C4IR is shaping the trajectory of the Fourth Industrial Revolution with local knowledge that can empower global change. Informed by national development priorities, Rwanda has decided to focus its Centre’s work on artificial intelligence and machine learning, data governance, and the data economy.
About the Rwanda Biomedical Centre
The Rwanda Biomedical Centre (RBC) is Rwanda's central health implementation agency. RBC was established in 2011 through a merger of fourteen key health institutions. RBC strives to improve the health of the Rwandan population by providing high-quality, affordable, and sustainable health care services. This is accomplished through the implementation of preventative, rehabilitative, and curative health interventions. RBC conducts scientific research, provides diagnostics services, and implements innovative health interventions to protect the nation against diseases and other health threats.
About Digital Umuganda
Digital Umuganda is an AI and open data company based in Kigali, Rwanda. Its mission is to enable access to information in local African languages, which are underrepresented in the digital world. Digital Umuganda is dedicated to changing this narrative by leveraging AI and open data, creating high-quality language datasets, and developing innovative AI solutions to ensure that every voice is heard and no language is left behind in the digital age.