Study objective: To assess whether bystander CPR (BCPR) on collapse affects initial rhythm and outcome in patients with witnessed, unmonitored out-of-hospital cardiac arrest (OHCA).
Design: Prospective cohort study. Student's t test, the chi 2 test, and logistic regression were used for analysis.
Setting: Suburban emergency medical service (EMS) system.
Participants: Patients 19 years or older with witnessed OHCA of presumed cardiac origin who experienced cardiac arrest before EMS arrival between July 1989 and July 1993.
Results: Of 722 patients who met the entry criteria, 153 received BCPR. Patients who received BCPR were younger than those who did not: 62.5 +/- 15.4 years versus 66.8 +/- 15.1 years (P < .01). We found no differences in basic or advanced life support response intervals or in frequency of AED use. More patients initially had ventricular fibrillation (VF) in the BCPR group: 80.9% versus 61.4% (P < .01). The interval to definitive care for ventricular tachycardia (VT)/VF was longer for the BCPR group (8.59 +/- 5.3 versus 7.45 +/- 4.7 minutes; P < .05). The percentage of patients discharged alive who were initially in VT/VF was higher in the BCPR group: 18.3% versus 8.4% (P < .001). In a multivariate model, BCPR is a significant predictor for VT/VF and live discharge with adjusted ORs of 2.7 (95% CI, 1.7 to 4.4) and 2.4 (95% CI, 1.5 to 4.0), respectively. For those patients in VT/VF, BCPR predicted live discharge from hospital with an adjusted OR of 2.1 (95% CI, 1.2 to 3.6).
Conclusion: Patients who receive BCPR are more often found in VT/VF and have an increased rate of live discharge, with controls for age and response and definitive care intervals. For VT/VF patients, BCPR is associated with an increased rate of live discharge.