AI-generated HPI drafts in the ED: comparing to physician-written history
Official title: Prospective Evaluation of Artificial Intelligence-generated History of Present Illness Drafts in the Emergency Department
Run by Yonsei University · NCT07804407 · Data updated 2026-11-12
In the emergency department, researchers test whether an on-site AI tool can draft a history of present illness that is higher quality than the doctor’s usual note.
Quick facts
| Study type | Observational |
|---|---|
| Ages | 19 Years and up |
| Sex | All |
| Enrollment | 100 (estimated) |
| Sponsor | Yonsei University Academic / other |
| Locations | 1 site (Seoul, Seoul City) |
| Status | Recruiting |
| Started | September 1, 2026 |
| Main results expected | December 31, 2026 |
| Study ends | December 31, 2026 |
| Registry record updated | September 22, 2026 |
| Trial ID | NCT07804407 on ClinicalTrials.gov |
Is this trial a fit for you?
✓Likely a good fit
Adults (19 years or older) presenting to the Yonsei University Severance Hospital ED who are able to consent and have an EDnote-documented initial encounter.
✕Probably not a fit
Patients with communication barriers or those whose data quality is limited by real-time transcription may not experience benefit.
A plain-English reading of the study's rules. Only the study team can confirm whether you qualify.
What this trial studies
This single-center, prospective observational study looks at an on-premise large language model tool (EDnote) that creates a draft history of present illness from real-time transcripts. The treating physician composes the conventional HPI without seeing the AI draft, and both are compared by blinded emergency physicians. Quality of documentation is assessed to see if the AI draft improves the completeness and clarity of the HPI. Eligible participants are adults presenting to Yonsei University Severance Hospital ED, with consent obtained, excluding those unable to communicate or who are in cardiac arrest. The study aims to understand whether AI assistance can enhance ED documentation without altering physician judgment.
Why it matters
Potential benefit: If successful, patients could receive more accurate and comprehensive ED documentation, potentially improving care coordination and outcomes.
How similar studies have gone: Prior work in AI-assisted clinical documentation has shown mixed results; this study adds prospective, blinded comparison in an ED setting.
Who can join: the full rules
Show full inclusion / exclusion criteria
Inclusion Criteria: * Age ≥19 years; presenting to the emergency department; initial encounter documented with EDnote; verbal consent to study participation. Exclusion Criteria: * Cardiac arrest; limited ability of both patient and guardian to communicate verbally; refusal of participation.
Where it's running
-
Yonsei University Severance Hospital
Seoul, Seoul City, South Korea Recruiting
Study contacts
- Study coordinator: Stephen GW LEE
- Email: leestephengyungwon@yuhs.ac
- Phone: 82-2-2228-2678
Similar studies
Other studies for the same condition or treatment. Recruiting studies are listed first.
-
Patient-specific MRD liquid biopsy for resected NSCLC
This study tests a personalized MRD blood and bronchial washing test after surgery to see if it can signal recurrence earlier in people with resected NSCLC.
-
Autologous adipose-derived stem cell injections for knee osteoarthritis
We’re testing whether injecting a patient’s own fat-derived stem cells into the knee can help people with knee OA feel less pain and move better, compared with a standard hyaluronic acid treatment.
-
SVF cellular mix, secretome, and knee response in osteoarthritis
This study tests the cellular composition and secretome of research-only adipose-derived SVF in adults with knee OA who are already planned to receive an autologous SVF injection as part of their care.
-
2-mm subchondral drilling vs sequential 4-mm and 2-mm drilling during CARTISTEM implantation for knee cartilage defects
This study tests whether drilling the knee with a 2-mm drill alone works as well as sequential 4-mm then 2-mm drilling when combined with CARTISTEM treatment for medial knee cartilage defects.
-
Blood biomarker cohort for memory disorders in older adults
This observational cohort will collect blood samples and brain imaging data from adults 50+ with memory disorders and a small group of healthy young adults to help find biomarkers.
-
Difference in finger, earlobe, and forehead oxygen levels to predict outcomes in dyspnea
In adults with breathing trouble, this study tests whether differences in SpO2 readings from different body sites can predict who might have worse outcomes.
How to take part
- Review the eligibility criteria above with your treating physician.
- Check the official listing on ClinicalTrials.gov for the latest contact details and recruitment status.
- Contact the study coordinator or principal investigator to ask for pre-screening. Pre-screening is free and never obligates you to enroll.