ELICITED: EHR-grounded Longitudinal Interactive Conversations for Information-seeking Triage Evaluation and Decision-making

2026-08-09

AI Summary

This work addresses a critical limitation in existing emergency triage benchmarks, which rely solely on static clinical snapshots and fail to capture the dynamic, interactive nature of real-world physician assessment through iterative questioning. To bridge this gap, we propose EHR2Dial-Triage—the first dynamic triage dialogue generation framework based on MIMIC-IV-ED. By incorporating role constraints and temporal boundaries, our approach precisely aligns electronic health record (EHR) events with dialogue turns and introduces diverse patient personas. Framing triage as an interactive dialogue process, EHR2Dial-Triage enables structured, multi-dimensional evaluation of information-gathering strategies, evidence integration, Emergency Severity Index (ESI) prediction accuracy, and communication quality, thereby establishing a novel benchmark for quantitatively assessing the comprehensive performance of intelligent triage systems.

Abstract

Emergency-department (ED) triage requires clinicians to rapidly identify patients who need immediate attention, determine who can safely wait, and prioritize limited clinical resources. At presentation, however, information may be limited to a chief complaint and initial vital signs. Clinically important details, including symptom onset and progression, associated symptoms, medical history, and medication use, are often obtained through focused conversation. Effective triage therefore requires clinicians to identify information gaps, ask appropriate follow-up questions, and update their assessment as new evidence becomes available. Most existing ED benchmarks evaluate acuity prediction from a fixed clinical snapshot. Although this formulation measures predictive performance after patient information has been assembled, it does not capture the interactive process through which triage-relevant evidence is elicited and interpreted. Existing medical dialogue datasets support the study of clinical communication, but dialogue statements are not always linked to temporally ordered events in the electronic health record (EHR). We introduce EHR2Dial-Triage, an agentic conversation-generation framework and benchmark grounded in MIMIC-IV-ED. The framework constructs triage conversations under explicit role-based and temporal information boundaries. Each accepted patient disclosure is linked to its supporting EHR event and the first dialogue turn at which it becomes available. EHR2Dial-Triage enables controlled evaluation of information elicitation, evidence use, five-level Emergency Severity Index prediction, and patient-facing communication across models and patient personas. It provides a structured setting for studying conversational triage as a dynamic process of clinical information acquisition, reasoning, and communication.

Problem

• emergency triage
• interactive conversation
• information elicitation
• electronic health record (EHR)
• clinical decision-making

Innovation

• EHR-grounded dialogue
• interactive triage
• temporal information elicitation
• clinical reasoning
• conversational AI