Using geographic information systems to evaluate cardiac arrest survival.

OBJECTIVE: To evaluate cardiac arrest survival using geographical information systems (GIS) methodology.
METHODS: Patient data were obtained from a fire district Utstein-style adult cardiac arrest registry that also included address data. All incident locations were geocoded and fire station first-due areas were mapped by using the new computer-aided dispatch geographic data. Retrospective assignment of first-due versus second-due fire response unit was done by using a GIS "point-in-polygon" algorithm Survival to hospital admission was the primary outcome measure for incidents responded to by first-due versus second-due apparatus controlling for other potential predictors of survival using logistic regression. Cluster analysis was also performed to evaluate potential areas of high or low rates of survival.
RESULTS: There were 461 eligible patients with an average age of 67+/-17 years, 63% were male, 53% had a witnessed arrest, bystander cardiopulmonary resuscitation was performed in 38%, bystander automatic external defibrillator (AED) Page: 1 was used in 0.01%, ventricular fibrillation or ventricular tachycardia were the presenting rhythms in 44%, the average response time was 5.5+/-2.1 minutes, and survival to hospital admission was 17%. There was no significant difference in response time between survivors (4.97 minutes) and non-survivors (5.52 minutes), (difference 0.55 minutes, 95%CI -0.08 to 1.18 min). The number of cardiac arrest calls varied from 1 to 49 for each station and the rate of second-due response varied from 0 to 19%. There was a nonsignificant association of survival to hospital admission for the first-due area cohort: odds ratio 0.70, 95% CI 0.38-1.29.
CONCLUSION: GIS is a new methodology for analyzing EMS incident data. It adds a spatial component of analysis to traditional statistical techniques. No spatial difference was found on patient survival in this analysis.