The Application of Large Language Model in Emergency Chest Pain Triage (ALERT)

This study will evaluate the accuracy and efficiency of large language model in emergency triage.

Study Overview

• Status: Recruiting
• Conditions: Chest Pain
• Intervention / Treatment: Diagnostic test: Application of large language model in emergency chest pain triage.; Diagnostic test: According to the normal procedures to receive medical treatment

Detailed Description

The study is to evaluate the value of large language model in emergency triage, their accuracy and efficiency were evaluated and compared with traditional triage. To explore whether the model can effectively reduce the workload of medical staff, while improving the speed and quality of triage. In addition, the ability of the model to predict serious medical events such as acute heart events and strokes was evaluated. It also included surveys of patients; acceptance and satisfaction with the use of the artificial intelligence-assisted triage system. Analyze the economic benefits of adopting this technology, including cost saving and optimal allocation of resources.

• Study Type: Interventional
• Enrollment (Estimated): 2000
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Xiangbin Meng; Phone Number: 17600220171; Email: puthxnk@126.com

Study Locations

• China
  • Beijing
    • Beijing, Beijing, China
      • Recruiting
      • Peking University Third Hospital

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

Description

Inclusion Criteria:

1. All patients with chest pain entered the emergency triage procedure.
2. patients aged 18 and above.

Exclusion Criteria:

1. Patients with severe cognitive impairment or inability to communicate.
2. There are patients who have been explicitly referred to specific departments (for example, some of the 120 transfer patients, who may go directly to the green channel) .
3. Patients with unstable vital signs .
4. Patients with potential medical problems.
5. Is participating in other clinical trials.
6. Failure to follow test procedures.
7. Those who refuse to sign the informed consent form.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Diagnostic
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: Large Language Model Diagnostic; Patients interacted with the large-language model triage system MedGuide-V5 during the waiting period before or after routine triage in the emergency department. During this phase, MedGuide-V5 will automatically record data and metrics during communication with patients.	Diagnostic test: Application of large language model in emergency chest pain triage.; The large language model MedGuide-V5 is able to quickly extract key information from a patients description, and by analyzing these descriptions, it provides physicians with a possible initial diagnosis to help them quickly prioritize the treatment of patients.
Active Comparator: Routine diagnostic and therapeutic procedure; After the artificial intelligence system evaluation, the patients will receive the diagnosis and treatment according to the normal procedure. The overall time of artificial triage, the triage of patients, and other data will be recorded. Patient visits should not be delayed by the use of artificial intelligence systems for evaluation.	Diagnostic test: According to the normal procedures to receive medical treatment; After the artificial intelligence system evaluation, the patients will receive the diagnosis and treatment according to the normal procedure. The overall time of artificial triage, the triage of patients, and other data will be recorded. Patient visits should not be delayed by the use of artificial intelligence systems for evaluation.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
The Diagnostic Accuracy Rate of MedGuide-V5	To assess the consistency of the diagnosis of chest pain made by physicians with the assistance of large language models with the actual diagnosis made by patients after all examinations were completed.	through study completion, an average of 10 months

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
The Satisfaction of Medical Personnel	To evaluate the satisfaction and acceptance of medical personnel with the use of large language models in assisting triage systems through methods such as questionnaire surveys. The name of this questionnaire is: Researcher Evaluation Form, with scores ranging from 1 to 10. The higher the score, the more helpful the large language model is to researchers.	during evaluation
Medical Personnel Treatment Plan Adjustment Rate	The number of times medical personnel adjust treatment plans after receiving feedback from MedGuide V5's results and referring to the suggestions provided by the large language model.	during evaluation
Emergency Department Revisit Rate within 30 Days	Evaluate the occurrence of patients revisiting the emergency department or being readmitted within 30 days after large language model-assisted triage and traditional triage.	during evaluation

Collaborators and Investigators

• Sponsor: Peking University Third Hospital
• Collaborators: Qingdao Municipal Hospital; Tianjin Medical University General Hospital; Jinan Central Hospital; The First Hospital of Hebei Medical University
• Investigators: Principal Investigator: Yi-Da Tang, MD, PhD, Peking University Third Hospital

Study record dates

Study Major Dates

• Study Start (Actual): December 20, 2023
• Primary Completion (Estimated): December 20, 2024
• Study Completion (Estimated): December 20, 2024

Study Registration Dates

• First Submitted: December 22, 2023
• First Submitted That Met QC Criteria: July 8, 2024
• First Posted (Actual): July 9, 2024

Study Record Updates

• Last Update Posted (Actual): July 9, 2024
• Last Update Submitted That Met QC Criteria: July 8, 2024
• Last Verified: July 1, 2024

More Information

Terms related to this study

• Keywords: Chest Pain; large language model
• Additional Relevant MeSH Terms: Pathologic Processes; Pain; Neurologic Manifestations; Disease Attributes; Emergencies; Chest Pain
• Other Study ID Numbers: M2023828

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No