Formative Modeling of Multi-Agent AI-Based AR Decision Support for Rural Emergency Department Triage

Emergency Department (ED) triage requires clinicians to make rapid prioritization decisions under uncertainty, fragmented information, and limited time. This formative study examines how AI-supported representation structures may reorganize triage reasoning workflows rather than simply reduce workload. We propose a multi-agent AI-based Augmented Reality (AR) representation designed to externalize multiple AI-generated clinical perspectives and supporting rationales within a shared situational context. To investigate the cognitive implications of this approach, Hierarchical Task Analysis (HTA)-informed cognitive workflow analysis and Cogulator-based operator-level modeling were conducted using an ambiguous chest pain triage scenario. Preliminary modeling suggested a qualitative shift in workflow structure, from repeated information search and contextual reconstruction toward externally supported comparison and verification. These findings suggest that representation structure itself may shape how cognitive activities are organized during triage reasoning, particularly in rural ED environments where collaborative support resources are limited.