Software annotation of defibrillator files: Ready for prime time?

Area of Science:

• Emergency Medicine
• Cardiology
• Biomedical Engineering

Background:

• High-quality chest compressions are crucial for improving outcomes in cardiac arrest patients.
• Defibrillators capture vital chest compression data during cardiopulmonary resuscitation (CPR), useful for quality improvement.
• Automated software can analyze defibrillator data and measure chest compression metrics, but accuracy evidence is limited.

Purpose of the Study:

• To compare chest compression fraction (CCF) and rate measurements derived from automated software annotation versus manual and limited manual annotation of defibrillator files.
• To evaluate the accuracy of automated software in measuring chest compression metrics during out-of-hospital cardiac arrest (OHCA) CPR.

Main Methods:

• A retrospective observational study involving 100 OHCA patients undergoing CPR.
• Chest compression bioimpedance waveforms were analyzed from CPR initiation to defibrillator removal.
• Software annotations were compared against two manual annotation methods (complete and limited) using intraclass correlation coefficient (ICC) analysis.

Main Results:

• Chest compression rate measurements showed excellent agreement across all methods (ICC > 0.90), with mean rates between 108.1-108.6 compressions/minute.
• Chest compression fraction (CCF) measurements varied significantly: software (0.64 ± 0.19), manual (0.86 ± 0.07), and limited manual (0.81 ± 0.10).
• ICC for CCF between manual and limited manual annotation was 0.69, and 0.83 for individual minute epochs.

Conclusions:

• Automated software demonstrates high accuracy for measuring chest compression rate during OHCA CPR.
• Significant discrepancies exist between automated software and manual methods for CCF measurements, indicating a clinical importance.
• Limited manual annotation provides results similar to complete manual annotation for CCF, suggesting a viable alternative for quality assessment.