BACKGROUND: External chest compression is considered to play a significant role in cardiopulmonary resuscitation (CPR), but during a rhythm check, chest compressions must be discontinued to avoid artifacts. A new multifunctional electrocardiograph (ECG; Radarcirc) has been developed for use in clinical settings. METHODS AND RESULTS: The performance of the Radarcirc and conventional ECG (CoECG) during CPR was compared in a single-center, non-randomized, sequential self-controlled study. CPR was performed on 41 out-of-hospital cardiac arrest patients. Cardiac rhythm with and without chest compressions during a rhythm check was measured using leads I and II. When the rhythm changed during CPR, it was measured as another waveform. Fifty ECG recordings were obtained, of which 27 were asystole, 18 pulseless electrical activity, and 5 ventricular fibrillation (VF). The area under the receiver-operating characteristic curve (AUC) for VF was 0.448 (95% confidence interval (CI) 0.274-0.622) for lead II of the CoECG, and 0.797 (95%CI 0.684-0.910) for lead II of the Radarcirc. The AUC for VF was 0.422 (95%CI 0.219-0.626) for lead I of the CoECG, and 0.987 (95%CI 0.975-1.00) for lead I of the Radarcirc. CONCLUSIONS: Diagnoses based on the data from Radarcirc were more accurate in predicting rhythm during chest compressions than those based on data from the CoECG.
BACKGROUND: External chest compression is considered to play a significant role in cardiopulmonary resuscitation (CPR), but during a rhythm check, chest compressions must be discontinued to avoid artifacts. A new multifunctional electrocardiograph (ECG; Radarcirc) has been developed for use in clinical settings. METHODS AND RESULTS: The performance of the Radarcirc and conventional ECG (CoECG) during CPR was compared in a single-center, non-randomized, sequential self-controlled study. CPR was performed on 41 out-of-hospital cardiac arrestpatients. Cardiac rhythm with and without chest compressions during a rhythm check was measured using leads I and II. When the rhythm changed during CPR, it was measured as another waveform. Fifty ECG recordings were obtained, of which 27 were asystole, 18 pulseless electrical activity, and 5 ventricular fibrillation (VF). The area under the receiver-operating characteristic curve (AUC) for VF was 0.448 (95% confidence interval (CI) 0.274-0.622) for lead II of the CoECG, and 0.797 (95%CI 0.684-0.910) for lead II of the Radarcirc. The AUC for VF was 0.422 (95%CI 0.219-0.626) for lead I of the CoECG, and 0.987 (95%CI 0.975-1.00) for lead I of the Radarcirc. CONCLUSIONS: Diagnoses based on the data from Radarcirc were more accurate in predicting rhythm during chest compressions than those based on data from the CoECG.