Silent Deployment and Prospective Evaluation of Zhunxin Agent for Emergency Chest Pain Assessment
Trial statusNot yet recruiting
Trial phaseNot listed
Trial typeObservational
Biological sexAll
Age18+
SponsorPeking University Health Science Center
About this trial
The goal of this prospective, multicenter observational study is to evaluate how well Zhunxin Agent, an artificial intelligence system for chest pain assessment, performs in real-world emergency department settings. The study will include people who receive routine care for acute chest pain at participating hospitals.
The main questions this study aims to answer are:
1. How accurately can Zhunxin Agent identify serious causes of acute chest pain during routine emergency care? 2. How well can Zhunxin Agent distinguish people at different levels of clinical risk?
Zhunxin Agent will run in silent mode using clinical information collected during routine care. Its predictions will not be shown to treating clinicians and will not affect clinical decisions, diagnostic tests, or treatment.
Researchers will compare the predictions made by Zhunxin Agent with the final clinical diagnoses established through routine medical care.
Participants will:
1. Receive usual emergency care according to local clinical practice 2. Have routinely collected clinical data analyzed by Zhunxin Agent in the background 3. Receive no additional treatment or clinical intervention based on the Agent's predictions
Eligibility criteria
This trial does not accept healthy volunteersQualifiers
Participants aged 18 years or older.
Participants presenting to a participating emergency department with acute chest pain, chest discomfort, or other symptoms that prompt clinical evaluation for an acute chest pain-related condition.
Participants undergoing an electrocardiogram as part of routine clinical care during the index emergency department encounter.
Participants with at least one electrocardiogram that can be processed by Zhunxin Agent during the index emergency department encounter.
Disqualifiers
Participants whose emergency department presentation does not involve evaluation for acute chest pain.
Participants without an electrocardiogram obtained during the index emergency department encounter.
Participants for whom the electrocardiogram or other required input data are missing, corrupted, technically inaccessible, or otherwise unusable such that Zhunxin Agent cannot generate a valid prediction.
Duplicate records corresponding to the same emergency department encounter.
Trial population
The study population will consist of consecutive participants presenting with acute chest pain or symptoms suggestive of an acute chest pain condition to the emergency departments of participating hospitals during the prospective silent-deployment period. Participants will receive routine clinical care, while Zhunxin Agent will operate in the background without providing its predictions to treating clinicians.
Trial design
Design model
Cohort
Time perspective
Prospective
Treatments tested in this trial
Not listed
Trial groups
No trial groups listed
Trial outcomes
Primary outcomes
1
Percentage of Participants Correctly Classified by Zhunxin Agent Across the Prespecified Acute Chest Pain Diagnostic Categories
Diagnostic accuracy will be reported as the percentage of evaluable participants for whom the final diagnostic classification generated by Zhunxin Agent matches the final reference diagnosis. The prespecified diagnostic categories are unstable angina, non-ST-segment elevation myocardial infarction, ST-segment elevation myocardial infarction, aortic dissection, pulmonary embolism, and other causes of chest pain. The numerator will be the number of participants with a concordant Zhunxin Agent classification and reference diagnosis, and the denominator will be the total number of evaluable participants with both a valid Zhunxin Agent prediction and an available final reference diagnosis.
Time frame
Up to 30 days after emergency department presentationSecondary outcomes
1
Diagnostic Performance of Zhunxin Agent for High-Risk Acute Chest Pain Conditions Assessed by Sensitivity, Specificity, Positive Predictive Value, Negative Predictive Value, and AUROC
Diagnostic performance of Zhunxin Agent will be evaluated using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and the area under the receiver operating characteristic curve (AUROC), using the final reference diagnosis as the reference standard. Prespecified high-risk acute chest pain conditions include ST-segment elevation myocardial infarction, aortic dissection, and pulmonary embolism. Sensitivity will be calculated as the proportion of participants with a target condition who are correctly identified by Zhunxin Agent. Specificity will be calculated as the proportion of participants without the target condition who are correctly identified as negative. PPV will be calculated as the proportion of participants with a positive Zhunxin Agent prediction who have the target condition, and NPV as the proportion of participants with a negative prediction who do not have the target condition. AUROC will be calculated using the predicted probabili
Time frame
Up to 30 days after emergency department presentation2
Percentage of Participants With a Confirmed High-Risk Acute Chest Pain Condition Within Each Zhunxin Agent Risk Category
The clinical risk stratification performance of Zhunxin Agent will be evaluated by calculating the percentage of participants with a confirmed high-risk acute chest pain condition within each prespecified risk category. Zhunxin Agent classifies participants into low-, intermediate-, and high-risk categories for ST-segment elevation myocardial infarction, aortic dissection, and pulmonary embolism using prespecified thresholds. For each target condition and risk category, the numerator will be the number of participants with the corresponding final reference diagnosis, and the denominator will be the total number of participants assigned to that risk category. The proportions will be reported separately for the low-, intermediate-, and high-risk categories.
Time frame
Up to 30 days after emergency department presentationLocations
Sponsors and contacts
Click on the lead sponsor to view all of their trials.
Source ClinicalTrials.gov
This trial is not recruiting at the moment. You can still explore other options: