Performance of chest compressions by laypersons during the Public Access Defibrillation Trial

Increasing evidence indicates that health professionals often may not achieve guideline standards for cardiopulmonary resuscitation (CPR). Little is known about layperson CPR performance.

The investigation was a retrospective cohort study of cardiac arrest patients treated by layperson CPR and one model of automated external defibrillator (AED) as part of the Public Access Defibrillation Trial (n=26). CPR was measured using software that integrates the event log, ECG signal, and thoracic impedance signal. We assessed chest compression fraction (proportion of attempted resuscitation spent performing chest compressions), prompted compression fraction (proportion of attempted resuscitation spent performing compressions during AED-prompted periods), compression rate, and compressions per minute.

Of the 26 cases, 13 presented with ventricular fibrillation and 13 with nonshockable rhythms. Overall, during the period when patients did not have spontaneous circulation, the median chest compression fraction was 34% (IQR 17-48%), median prompted chest compression fraction was 49% (IQR 30-66%), and the median chest compression rate was 96/min (IQR 90-110/min). Taken together, the median chest compression delivered per minute among all arrests was 29 (IQR 20-42). CPR characteristics differed according to initial rhythm: median chest compression per minute was 20 (IQR 13-29) among ventricular fibrillation and 42 (IQR 28-47) among nonshockable rhythms (p=0.003).

In this study of trained laypersons, CPR varied substantially and often did not achieve guideline parameters. The findings suggest a need to improve CPR training, consider changes to CPR protocols, and/or improve the AED-rescuer interface.