Authors: Samuel Rutunda, Gwydion Williams, Kleber Kabanda, Francis Nkurunziza, Solange Uwiduhaye, Eulade Rugegamanzi, Cyprien Nshimiyimana, Vaishnavi Menon, Mira Emmanuel-Fabula, Alastair K. Denniston, Xiaoxuan Liu, Emery Hezagira & Bilal A. Mateen
This study tested how well artificial intelligence tools could answer clinical questions asked by community health workers in Rwanda. Researchers collected more than 5,000 real questions and compared answers from several AI systems with those from local clinicians. The AI tools generally gave higher-quality responses and were far cheaper to use. Although performance dropped slightly when using the local language, the results suggest that AI tools could help support community health workers and improve care in low-resource health systems.
Download Resource
Resource Topic: Healthcare, Healthcare accessibility, Large Language Models, Low-income
Resource Type: Qualitative
Year: 2026
Region: Sub-Saharan Africa (SSA)
Country: Rwanda
Publisher May Restrict Access: No
