Automated external defibrillators: to what extent does the algorithm delay CPR?

STUDY
OBJECTIVE: Maximizing cardiopulmonary resuscitation (CPR) during resuscitation may improve survival. Resuscitation protocols stack up to 3 shocks to achieve defibrillation, followed by an immediate postdefibrillation pulse check. The purpose of this study is to evaluate outcomes of rhythm reanalyses immediately after shock, stacked shocks, and initial postshock pulse checks in relation to achieving a pulse and initiating CPR.
METHODS: We conducted an observational study of patients with ventricular fibrillation treated by first-tier emergency medical services (EMS). We collected data from EMS, dispatch, and hospital records. Additionally, we analyzed automatic external defibrillator recordings to determine the proportion of cardiac arrest victims who were defibrillated and achieved a pulse according to shock number (single versus stacked shock), proportion of victims with a pulse during the initial postdefibrillation pulse check, and interval from initial shock to CPR.
RESULTS: The study included 481 cardiac arrest subjects. Automatic external defibrillators terminated ventricular fibrillation with the initial shock in 83.6% (n=402) of cases. A second shock terminated ventricular fibrillation in an additional 7.5% (n=36) of cases, and a third shock terminated ventricular fibrillation in 4.8% (n=23) of cases. The initial sequence of 3 shocks failed to terminate ventricular fibrillation in 4.1% (n=20) of cases. In total, automatic external defibrillators performed 560 rhythm reanalyses during the initial shock sequence and delivered 122 "stacked" shocks. Termination of ventricular fibrillation was not synonymous with return of a pulse. The initial shock produced a pulse that was eventually detected in 21.8% (105/481) of cases. Stacked shocks produced a pulse in 10.7% (13/122) of cases. For the 24.5 % (n=118) of cases in which a pulse returned, the pulse was detected during the initial postshock pulse check only 12 times, or 2.5% of all cases. The median interval from initial shock until CPR was 29 (23,41) seconds.
CONCLUSION: Rhythm reanalyses, stacked shocks, and postshock pulse checks had low yield for achieving or detecting return of a pulse. CPR was not initiated until 29 seconds after the initial shock.