AI-guided Prediction and Treatment of Cardiac Arrest

Improving Cardiac Arrest Outcomes Using Artificial Intelligence Guided Precision Treatments

Sudden cardiac arrest is a major health problem, and most people don't survive. One big reason is that even if resuscitation is successful, people commonly have recurrent cardiac arrests (rearrest). Right now, it is not possible to accurately predict a rearrest or prevent it. The investigators have developed a machine learning device that uses the heart tracing (ECG) to predict when and why a rearrest occurs. The investigators plan to test if it will accurately and effectively help EMS providers predict rearrest and provide timely treatment to increase survival after cardiac arrest. To determine if this machine learning device will work in the real world, the investigators need to find out if there are barriers to using it, and whether EMS providers will think it is useful and will help them improve the care of patients who have a cardiac arrest. The investigators will first test the device in live simulated cardiac arrest scenarios to see if the providers can use it and if they find the device potentially valuable in taking care of patients. In a second study, the investigators will test how accurate the device is in predicting if a cardiac arrest will happen again in patients who have just been brought back to life after a cardiac arrest. EMS providers will attach the device, but it will only work in the background. EMS will take care of patients as they normally would, without using or knowing what the device says. To see if the device is accurate at predicting another cardiac arrest, the investigators will analyze the results offline, and compare what the device says to what actually happens to the patient. By comparing what the device predicts to what actually happens, the investigators can see how well it predicts another cardiac arrest and estimate how it might improve treatment of patients.

Study Overview

• Status: Not yet recruiting
• Conditions: Sudden Cardiac Arrest
• Intervention / Treatment: Device: Machine learning-guided cardiac arrest prediction device

Detailed Description

Sudden cardiac arrest (SCA) remains a major cause of mortality in the United States. Despite significant efforts to improve resuscitation outcomes, survival remains poor. Moreover, SCA is often the first manifestation of underlying heart disease, after which survivors typically suffer secondary chronic cardiovascular and neurological disease due to the primary insult. Resuscitation from SCA is initiated in patients suffering from life-threatening arrhythmias, generally ventricular tachycardia/fibrillation (VT/VF) or pulseless electrical activity (PEA), that if successful, is followed by a return of spontaneous circulation (ROSC). This, however, is commonly followed by repeated rearrest due to either VT/VF, PEA, or asystole. Importantly, the mechanisms of VT/VF and PEA vary significantly, and protocolized, non-directed treatments can worsen outcomes. Currently, treatments are typically initiated after rearrest occurs and without regard to an individual patient's unique personal or arrest characteristics, factors which are known predictors of cardiac arrest outcomes. Recent studies have shown that a quicker and more targeted treatment response (i.e., shorter time to treat with rearrest type-specific treatment) for cardiac arrest can significantly improve survival and downstream chronic sequelae after SCA. Therefore, the ability to predict cardiac arrest or rearrest occurrence and cause (i.e. VT/VF and PEA) could guide early personalized therapy and ultimately improve resuscitation outcomes, which is the long-term translational goal of this effort. Additionally, early intervention could be critical in rural communities, where transport times to definitive hospital care are long, and resources are limited.

In ongoing work the investigators have recently observed in emergency medical services (EMS) patients that adding clinical parameters (e.g., age, primary arrest type) to features derived from the ECG T-wave in a machine learning (ML) model improves prediction accuracy of VT/VF and PEA (>90%). The investigators propose a feasibility study of ML-guided prediction of rearrest. The overall hypothesis is that an ECG biomarker, combined with clinical features, can predict rearrest and its cause (VT/VF or PEA), which will significantly improve time to treatment and cardiac arrest acute and chronic outcomes. The investigators further hypothesize that such technology can be readily adopted and successfully implemented by emergency responders. The investigators plan two interrelated clinical trials, one an observational trial in simulated cardiac arrest, and a second observational trial in EMS patients. The following specific aims will test these hypotheses.

Aim 1. Determine end-user performance and satisfaction with a fully automated ML-guided rearrest prediction device using simulated cardiac arrest scenarios. Emergency providers, in simulated real-time resuscitation scenarios using real rearrest ECG recordings will be studied to identify barriers and facilitators to technology implementation and adoption, as well as provider satisfaction and the perceived value of the intervention. This will be achieved by determining provider performance metrics (e.g., initiating ML-guided therapies) and using qualitative tools to assess end-user engagement, trust and perceived utility. These studies will provide foundational data for the implementation and scalability of the technology for both a clinical trial and guidance for future FDA and regulatory approval.

Aim 2. Test the accuracy of a multiclass ML model for real-time prediction of rearrest occurrence and its type (VT/VF vs. PEA) and impact on time to treatment, in an observational clinical trial of cardiac arrest patients. Leveraging an ongoing collaboration with Cleveland and community EMS, the investigators will enroll EMS cardiac arrest patients and observe for the occurrence of rearrest in a proof-of-concept observational validation study. The accuracy of the fully automated ML model to predict rearrest and its type will be determined, but treatment guidance will not be tested at this time. Providers will be blinded to the ML output, and ML will not direct treatment. Offline, prediction accuracy and estimated time to ML-guided treatment decision will be compared to actual rearrest type and time of treatment. These results will provide preliminary safety and accuracy results, sample size estimates, and recruitment and informed consent processes for a future randomized controlled clinical trial to test ML-guided rearrest treatments.

• Study Type: Interventional
• Enrollment (Estimated): 68
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Lance Wilson, MD; Phone Number: 12169786274; Email: lwilson@metrohealth.org
• Study Contact Backup: Name: Julie Nichols Research Coordinator, RN; Phone Number: (216) 957-6488; Email: jnichols@metrohealth.org

Study Locations

• United States
  • Ohio
    • Cleveland, Ohio, United States, 44109
      • The MetroHealth System
      • Contact: Julie Nichols Research Coordinator, RN; Phone Number: (216) 957-6488; Email: jnichols@metrohealth.org
      • Sub-Investigator: Jeremiah Escajeda, MD
      • Sub-Investigator: Thomas Noeller, MD
      • Sub-Investigator: Joseph Piktel, MD
      • Contact: Lance Wilson Attending Physician and Professor, Emergency Medicine, MD; Phone Number: 216-978-6274; Email: lwilson@metrohealth.org

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: Yes

Study Population

For Phase 1: EMS provider volunteers from an Urban and Rural EMS system For Phase 2: Consecutive sample of all cardiac arrest patients treated for out of hospital cardiac arrest.

Description

Inclusion Criteria:

• Adult (18 years of age or older) EMS providers (Simulation trial)
• Adult (18 years of age or older) patients have attempted resuscitation from out-of-hospital SCA of any etiology (Clinical trail)

Exclusion Criteria:

• Non-English-speaking providers
• Providers who do not care for cardiac arrest patients
• Prisoners
• Pediatric patients under age of 18
• DNR/DNI
• No resuscitation attempted (declared deceased in field by EMS)

Study Plan

How is the study designed?

Design Details

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

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: Emergency Medical Service Providers; Emergency Medical Service Providers will experience high fidelity cardiac arrest simulations and test the barriers and facilitators to using a machine learning guided prediction device in simulated cardiac arrest patients.	Device: Machine learning-guided cardiac arrest prediction device; A machine learning-guided cardiac arrest prediction device will be used to predict recurrence of cardiac arrest after initially successful resuscitation. It will also predict if the recurrent cardiac arrest is caused by ventricular fibrillation/tachycardia or pulseless electrical activity.
Other: Patients who experience cardiac arrest cared for by EMS; Patients who experience cardiac arrest will receive normal standard of care treatments. A machine learning guided prediction device will run in the background and also receive the normally acquired ECG data. Offline, the accuracy of the device to predict recurrent cardiac arrest and the type of rearrest which occurs after successful return of spontaneous circulation will be determined.	Device: Machine learning-guided cardiac arrest prediction device; A machine learning-guided cardiac arrest prediction device will be used to predict recurrence of cardiac arrest after initially successful resuscitation. It will also predict if the recurrent cardiac arrest is caused by ventricular fibrillation/tachycardia or pulseless electrical activity.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Mean Implementation Acceptability Score	Mean score on a 20-item post-simulation survey adapted from the Consolidated Framework for Implementation Research (CFIR). Each item is rated on a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree). The composite score is calculated as the mean of all items (range 1-5), with higher scores indicating greater perceived implementation acceptability.	Assessed once immediately after completion of the simulation session (within 5 hours of enrollment).
Calculated time to treatment benefit	Determination of estimated change in time to treatments for cardiac arrest patients from the observational clinical trial of the ML-guided prediction device. Time to treatment will be measured (in seconds) from time to EMS arrival at scene to treatment time for the first rearrest is rendered. This will be compared to calculated time to treatment, measured from EMS arrival to machine learning prediction (in seconds).	From subject enrollment up to 2 hours

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Accuracy of ML-guided rearrest predication	Standard accuracy measures of device predication will performance in the phase 2 observational clinical trial. Machine learning prediction of rearrest occurrence and rhythm type will be compared to the actual observed rearrest occurrence and rhythm type. Accuracy will be defined as the percentage of correct predictions (True Positives + True Negatives) over the total dataset and shown in a confusion matrix.	From subject enrollment up to 2 hours
Time to machine learning guided prediction	Time to machine learning prediction of rearrest will be performed in the observational clinical trial of the ML-guided prediction device. Time to prediction will be measured (in seconds) from the time of EMS arrival to time the machine learning prediction device predicts the first rearrest with confidence greater than or equal to 70%.	From subject enrollment up to 2 hours
Time to device deployment in simulated cardiac arrest	Time to machine learning device deployment will be measured in cardiac arrest simulations. Time to deploy the device in simulated cardiac arrest patients will be measured (in seconds) from the time of simulated EMS arrival to time the machine learning prediction device is deployed and begins to collect data.	Assessed once immediately after completion of the simulation session (within 5 hours of enrollment).

Collaborators and Investigators

• Sponsor: MetroHealth Medical Center
• Collaborators: National Center for Advancing Translational Sciences (NCATS)

Study record dates

Study Major Dates

• Study Start (Estimated): August 1, 2026
• Primary Completion (Estimated): June 1, 2029
• Study Completion (Estimated): June 1, 2029

Study Registration Dates

• First Submitted: February 17, 2026
• First Submitted That Met QC Criteria: February 27, 2026
• First Posted (Actual): March 5, 2026

Study Record Updates

• Last Update Posted (Actual): March 16, 2026
• Last Update Submitted That Met QC Criteria: March 12, 2026
• Last Verified: March 1, 2026

More Information

Terms related to this study

• Keywords: machine learning; cardiac arrest; sudden cardiac death; ventricular tachycardia; ventricular fibrillation; pulseless electrical activity; rearrest
• Additional Relevant MeSH Terms: Cardiac Conduction System Disease; Cardiovascular Diseases; Pathologic Processes; Heart Diseases; Arrhythmias, Cardiac; Death; Death, Sudden; Tachycardia; Pathological Conditions, Signs and Symptoms; Heart Arrest; Death, Sudden, Cardiac; Tachycardia, Ventricular; Ventricular Fibrillation
• Other Study ID Numbers: STUDY00001031; UM1TR004528 (U.S. NIH Grant/Contract)

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: YES
• IPD Plan Description: Deidentified demographic and outcome data. Model outcome data.
• IPD Sharing Time Frame: After completion of study and study reporting, and an additional 2 years.
• IPD Sharing Supporting Information Type: STUDY_PROTOCOL; SAP

Drug and device information, study documents

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