Abstract 69: A Tale of Two Cities: The Role of Neighborhood Socioeconomic Status in Spatial Clustering of Bystander CPR in Austin and Houston

Abstract

Objective: To determine 1) the location of bystander CPR clusters and 2) the effect of neighborhood-level risk factors on clusters. Methods: Secondary analysis of the Austin/Travis County and Houston/Harris County Cardiac Arrest Registry to Enhance Survival (CARES) dataset. Population: Consecutive arrests from October 1, 2006 to December 31, 2009 (Austin) and from January 1, 2007 to December 31, 2009 (Houston). Data Analysis: Spatial scan statistic (SaTScan v9.1) was used to test spatial clusters of bystander CPR stratified by private and public arrest location. A Bernoulli model was used to determine significant clustering of the case (all OHCA with a non-EMS bystander initiated CPR) location distribution as compared to the controls (all other OHCA events) location distribution. “High” clusters indicated areas where the number of bystander CPR events were greater than expected while “low” clusters indicated areas where the number of bystander CPR events were lower than expected. Results: Two “high” private and two public (one “low” and one “high”) clusters were identified in Austin. Two “high” private clusters had approximately 1.8-2.6 times more bystander CPR cases than expected, while the “high” public cluster, located in the same area as the private “high” cluster, had 2.0 times more bystander CPR cases than expected. The “low” public cluster was in the Austin city center and contained nearly 80% fewer than expected. Three private (two “high” and one “low) and one “high” public clusters, were identified in Houston. Two “high” private clusters in the south had nearly 4.4 times more than expected, while the “low” private cluster included only 34 cases, 57% fewer than expected. The “high” public cluster was, just west of city center, included 14 cases, 2.5 times more than expected. However, when adjusted for neighborhood median household income, all clusters disappeared in Austin, though only the public arrest cluster disappeared in Houston. Conclusion: Our findings suggest that high and low clusters of bystander CPR can be partially explained by median household income. Further research will be required to better understand why such vast disparities in the provision of bystander CPR exist.