Predicting drops in cardiac output during general anesthesia using breathing-monitor data
Development of an Artificial Intelligence Model for Predicting Intraoperative Changes in Cardiac Output Using Capnography During General Anesthesia
This will test whether an AI algorithm can use capnography (breathing monitor) data to predict significant drops in cardiac output in adults having elective surgery with general anesthesia.
Quick facts
| Study type | Observational |
|---|---|
| Enrollment | 2005 (estimated) |
| Ages | 19 Years to 75 Years |
| Sex | All |
| Sponsor | Samsung Medical Center Academic / other |
| Locations | 1 site (Seoul) |
| Trial ID | NCT07061548 on ClinicalTrials.gov |
What this trial studies
This is an observational, single-center project enrolling adults undergoing elective surgery with endotracheal general anesthesia who have continuous capnography and invasive arterial waveform monitoring. Investigators train an AI model on 5-minute segments of capnography data labeled by whether cardiac output fell by 20% or more during that period. The resulting algorithm would generate alarms for predicted decreases in cardiac output without adding invasive devices. Patients with emergency, cardiac/thoracic surgery or moderate-or-worse preexisting lung disease are excluded, and only cases with at least 30 minutes of intraoperative monitoring are included.
Who should consider this trial
Good fit: Adults (roughly 18–75 years) having elective surgery under general anesthesia with endotracheal intubation who are monitored with continuous capnography and invasive arterial pressure for at least 30 minutes are ideal candidates.
Not a fit: Patients having emergency operations, cardiac or thoracic surgery, those with moderate-or-worse preexisting pulmonary disease (asthma/COPD) or procedures shorter than 30 minutes are not eligible and are unlikely to benefit from this capnography-based algorithm.
Why it matters
Potential benefit: If successful, the algorithm could provide a noninvasive early warning of falling cardiac output during anesthesia, potentially allowing timelier interventions and reducing need for invasive monitoring.
How similar studies have performed: Some prior research has explored using capnography waveform features to infer hemodynamic changes, but using AI to predict intraoperative cardiac output drops from capnography is relatively novel with limited clinical validation so far.
Eligibility criteria
Show full inclusion / exclusion criteria
Inclusion Criteria: * Elective surgery under general anesthesia * Adult patients (18 \< age \< 76) * Patients who were monitored invasive arterial blood pressure (waveform) and capnography (numeric) Exclusion Criteria: * Emergency surgery * Cardiovascular and thoracic surgery * Known Asthma and Chronic obstructive pulmonary disease (COPD) * Preoperative pulmonary function test (PFT) abnormality over moderate grade * Intraoperative monitoring duration less than 30 minutes
Where this trial is running
Seoul
- Samsung Medical Center — Seoul, South Korea (Recruiting)
Study contacts
- Study coordinator: Heejoon Jeong, MD
- Email: heejoonjeong@skku.edu
- Phone: +82-2-3410-0841
How to participate
- Review the eligibility criteria above with your treating physician.
- Visit the official trial page on ClinicalTrials.gov for the most current contact information and recruitment status.
- Contact the listed study coordinator or principal investigator to request pre-screening. Pre-screening is free and never obligates you to enroll.