Prospective Evaluation of AI Generated History of Present Illness Drafts in the Emergency Department

NCT07804407 | Recruiting

• 2 other identifiers: 4-2026-0708C-2026-0010
• Study Type: observational
• Target: 100
• Locations: 1 country
• Sites: 1

Brief Summary

This single-center prospective observational study evaluates whether an on-premise, large language model-based tool (EDnote) that generates a draft history of present illness (HPI) from real-time transcription of the patient-physician encounter produces documentation of higher quality than conventional physician-written HPI in the emergency department. The treating physician writes the conventional HPI blinded to the AI draft. Unedited AI-generated HPI drafts and conventional HPI are compared through blinded review by independent emergency physicians.

Conditions

All Emergency Department Visits Meeting Eligibility Criteria

Keywords

emergency department; ai scribe; ambient scribe; history of present illness

Outcome Measures

Primary Outcomes (1)

• Encounter-level clinical fact capture rate

  At blinded review, within 6 months after each index ED visit

Secondary Outcomes (3)

• Modified PDQI-9 score

  At blinded review, within 6 months after each index ED visit

• AI HPI draft generation time

  At index ED visit

• Rater identification of document authorship (AI vs. physician-written)

  At blinded review, within 6 months after each index ED visit

Study Arms (1)

ED patients with EDnote-assisted initial encounter

Adult emergency department patients whose initial encounter was recorded by EDnote. While the treating physician conducts usual care and writes the conventional HPI, EDnote generates an AI HPI draft in the background; the physician remains blinded to the draft. Both HPIs are evaluated by independent raters blinded to document source.

Other: EDnote HPI (AI-generated HPI draft); Other: Conventional physician HPI

Interventions

EDnote HPI (AI-generated HPI draft) OTHER

On-premise large language model-based tool that transcribes the patient-physician encounter in real time and generates an unedited HPI draft. The treating physician is blinded to the AI HPI draft.

ED patients with EDnote-assisted initial encounter

Conventional physician HPI OTHER

HPI written by the treating emergency physician according to usual practice, blinded to the EDnote-generated HPI draft.

ED patients with EDnote-assisted initial encounter

Eligibility Criteria

• Age: 19 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Adult (18-64), Older Adult (65+)
• Sampling Method: Non-Probability Sample

Study Population

Adult patients presenting to the emergency department of a single tertiary teaching hospital in Korea.

You may qualify if:

• Age ≥19 years; presenting to the emergency department; initial encounter documented with EDnote; verbal consent to study participation.

You may not qualify if:

• Cardiac arrest; limited ability of both patient and guardian to communicate verbally; refusal of participation.

Sponsors & Collaborators

• Yonsei University: lead

Study Sites (1)

Yonsei University Severance Hospital

Seoul, Seoul City, 03722, South Korea

RECRUITING

MeSH Terms

Conditions

Emergencies

Condition Hierarchy (Ancestors)

Disease Attributes; Pathologic Processes; Pathological Conditions, Signs and Symptoms

Central Study Contacts

Stephen GW LEE

CONTACT

82-2-2228-2678; leestephengyungwon@yuhs.ac

Study Design

• Study Type: observational
• Observational Model: COHORT
• Time Perspective: PROSPECTIVE
• Sponsor Type: OTHER
• Responsible Party: PRINCIPAL INVESTIGATOR
• PI Title: Clinical assistant professor

Study Record Dates

• First Submitted: September 1, 2026
• First Posted: September 4, 2026
• Study Start: September 1, 2026
• Primary Completion (Estimated): December 31, 2026
• Study Completion (Estimated): December 31, 2026
• Last Updated: September 22, 2026

Record last verified: 2026-09

Data Sharing

• IPD Sharing: Will not share

Locations

KR (1)