The three-phase model of cardiac arrest as applied to ventricular fibrillation in a large, urban emergency medical services system.

Background: Cardiac arrest is responsible for significant morbidity and mortality, with consistently poor outcomes despite the rapid availability of prehospital personnel for defibrillation attempts in patients with ventricular fibrillation (VF). Recent evidence suggests a period of cardiopulmonary resuscitation (CPR) prior to defibrillation attempts may improve outcomes in patients with moderate time since collapse (4-10 min).

Objectives: To determine cardiac arrest outcomes in our community and explore the relationship between time since collapse, performance of bystander CPR, and survival.

Methods: Non-traumatic cardiac arrest data were collected prospectively over an 18-month period. Patients were excluded for: age <18 years, a "Do Not Attempt Resuscitation" (DNAR) directive, determination of a non-cardiac etiology for arrest, and an initially recorded rhythm other than VF. Patients were stratified by time since collapse (<4, 4-10, > 10 min, and unknown) and compared with regard to survival and neurological outcome. In addition, patients with and without bystander CPR were compared with regard to survival.

Results: : A total of 1141 adult non-traumatic cardiac arrest victims were identified over the 18-month study period. This included 272 patients with VF as the initially recorded rhythm. Of these, 185 had a suspected cardiac etiology for the arrest; survival to hospital discharge was 15% in this group, with 82% of these having a good outcome or only moderate disability. Survival was highest among patients with time since collapse of less than 4 min and decreased with increasing time since collapse. There were no survivors among patients with time since collapse greater than 10 min. Among patients with time since collapse of 4 min or longer, survival was significantly higher with the performance of bystander CPR; there was no survival advantage to bystander CPR among patients with time since collapse less than 4 min.

Conclusions: The performance of bystander CPR prior to defibrillation by EMS personnel is associated with improved survival among patients with time since collapse longer than 4 min but not less than 4 min. These data are consistent with the three-phase model of cardiac arrest.