Related Experiment Video
Updated: Jul 3, 2026
05:36
Standardized Model of Ventricular Fibrillation and Advanced Cardiac Life Support in Swine
Published on: January 30, 2020
The Interrater Reliability of Cardiac Arrest Video Review: A Pilot Study
Moshe Ryan Ostreicher1,2, Joseph Landers1,2, Molly McCann Pineo1,2
1Northwell, New Hyde Park, NY.
Summary
Cardiac arrest (CA) video review shows excellent interrater reliability (IRR) for time-based data but only moderate IRR for categorical variables. This finding is crucial for quality improvement and research in emergency medicine.
Area of Science:
- Emergency Medicine
- Cardiology
- Medical Quality Improvement
Background:
- Audio and video recording of cardiac arrest (CA) events are valuable tools for quality improvement, education, and research.
- However, the interrater reliability (IRR) of data abstracted from these recordings has not been extensively studied.
- Accurate data collection is essential for meaningful analysis and improvement initiatives.
Purpose of the Study:
- To evaluate the interrater reliability (IRR) of data abstracted from video recordings of cardiac arrest (CA) events in the Emergency Department (ED).
- To assess the consistency of data collection for both categorical and noncategorical variables.
- To inform the use of CA video review as a reliable quality improvement tool.
Main Methods:
- A retrospective study involving 2 blinded, expert physicians reviewing 21 randomly selected CA videos from an academic ED.
- Abstraction of predefined categorical (e.g., initial rhythm) and noncategorical (e.g., time to defibrillation) variables.
- Interrater reliability (IRR) assessed using Cohen's kappa for categorical data and interclass correlation coefficient (ICC) for noncategorical data.
Main Results:
- Excellent IRR was observed for noncategorical and time-to-event variables, with ICC values ranging from 0.79 to 1.00.
- Moderate IRR was found for categorical variables, including initial ED rhythm (κ = 0.69) and endotracheal tube presence (κ = 0.70).
- High agreement was noted for specific categorical variables like mechanical compression and vascular access presence.
Conclusions:
- Cardiac arrest (CA) video review demonstrates excellent IRR for collecting noncategorical, time-based data, supporting its use in quality improvement.
- The IRR for categorical variables is only moderate, suggesting potential limitations or areas for standardization.
- Further refinement of data abstraction protocols may enhance reliability for all variable types in CA video review.
