Previously submitted to: JMIR Formative Research (no longer under consideration since May 27, 2026)
Date Submitted: Dec 14, 2025
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Bridging the Literacy Gap in Digital Health: A Prospective Study of a Dialect-Native AI Triage Chatbot for Urgent Care Optimization
ABSTRACT
Background:
Emergency Department (ED) overcrowding is a critical public health challenge in low-to-middle-income countries (LMICs), exacerbated by limited triage resources and linguistic barriers. Traditional AI symptom checkers often fail in these settings due to high literacy requirements, lack of dialect support, and the friction of app downloads.
Objective:
This study evaluates the clinical validity and operational impact of a novel 'zero-friction' AI pre-triage system integrated into WhatsApp. The system utilizes Large Language Models (LLMs) to process multimodal inputs (voice, text, images) in local dialects (Moroccan Darija), English, or French for automated risk stratification.
Methods:
A prospective, single-center diagnostic accuracy study was conducted at the Avicenne Military Hospital in Marrakech. A cohort of 150 consecutive adult patients underwent automated pre-triage via the AI system, followed immediately by standard physician triage (Gold Standard). The AI architecture utilized a secure middleware orchestrator to integrate OpenAI’s GPT-4 and Whisper models. Primary outcomes were concordance (Cohen’s Kappa) and diagnostic accuracy (Sensitivity/Specificity). Secondary outcomes included time-to-triage and user satisfaction.
Results:
The AI system demonstrated substantial agreement with physician judgment (Kappa=0.70; concordance 80.7%). Sensitivity for 'Urgent' cases was 92.3%, Specificity 93.7%, and NPV 97.2%. Mean triage time was reduced by 23.8 minutes (p < 0.001).
Conclusions:
This is the first known validation of a multimodal, dialect-native AI triage agent deployed on a ubiquitous messaging platform in North Africa. The system proved to be a clinically safe and highly efficient adjunct to human triage, offering a scalable solution to ED bottlenecks in resource-constrained environments.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.