📄 Abstract
In many of the northern states, Nigeria has less than one dentist to serve 50,000 people and there are no dental officers of any cadre in Primary Healthcare Centres (PHCs) in rural areas. Community Health Extension Workers (CHEWs), who are the main first contact providers for rural patients, are poorly trained in formal dental care, and have to judge if a patient has an emergency that may result in life threatening complications like Ludwig's angina, a fast spreading infection of the airway which if not managed promptly has a reported mortality rate of 8-40%. In this paper, an Android application called DTICS (Dental Triage and Intelligent Classification System) is presented which categorizes the patient's reported dental complaints into Emergency, Urgent and Routine for English, Hausa and Nigerian Pidgin language. DTICS uses a three-layer safety architecture: deterministic, fully offline rules layer, which encodes 8 emergency rules based on evidences, and has absolute priority over AI inference; an LLM classification layer, which classifies free-text and multilingual symptom description that are not captured by the rules; and a human escalation layer, which routes emergency and low-confidence cases to a dental officer with the help of an auditable Confirm/Override interface. The Android client is built using Clean Architecture, Hilt dependency injection, Room persistence (encrypted using SQLCipher) and WorkManager-based background synchronisation, and has been informally evaluated by practicing dentists for usability. The system architecture is presented, as well as the rationale behind its conservative and rule-first safety design and a shadow-mode evaluation protocol measuring the agreement with dental officer triage decisions (Cohen's κ) and emergency sensitivity. DTICS proves that a clinically conservative, offline-resilient triage tool can be created in the connectivity and language landscape of Nigeria, and provides a replicable framework for AI-based triage in resource-limited environments.
🏷️ Keywords
📚 How to Cite:
Grace Ojochenemi Emmanuel-Anorue, Sadiya Mohammed Dala , DTICS: AN OFFLINE-FIRST MULTILINGUAL DENTAL TRIAGE SYSTEM FOR NIGERIAN PRIMARY HEALTHCARE , Volume 11 , Issue 9, September 2026, EPRA International Journal of Research & Development (IJRD) , Pages: 81 - 84 ,