Abstract 187: Cardiac Arrest Incidence, Bystander Cardiopulmonary Resuscitation Rates, and Community Characteristics Differ by High and Low Risk Areas

Abstract

Introduction: Survival from out-of-hospital cardiac arrest (OHCA) has significant variation which may be due to differing rates of bystander cardiopulmonary resuscitation (BCPR). Defining and understanding the community characteristics of high-risk areas (census tracts with low BCPR rates and high OHCA incidence) can help inform novel interventions to improve outcomes. Our objectives were to identify high and low risk census tracts in Franklin County, Ohio and to compare the OHCA incidence, BCPR rates, and community characteristics. Methods: This was a cross-sectional analysis of OHCA events treated by Columbus Division of Fire in Franklin County, Ohio from the Cardiac Arrest Registry to Enhance Survival between 1/1/2010-12/31/2017. Included cases were 18 and older, with a cardiac etiology OHCA in a non-healthcare setting, with EMS resuscitation attempted. After geocoding to census tracts, Local Moran’s I and quartiles were used to determine clustering in high risk areas based on spatial Empirical Bayes smoothed rates. Community characteristics, from the 2014 American Community Survey, were compared between high and low risk areas. Results: From the 3,841 included OHCA cases, the mean adjusted OHCA incidence per census tract was 0.67 per 1,000 with a mean adjusted BCPR rate of 31% and mean adjusted survival to discharge of 9.4%. In the 25 census tracts identified as high-risk areas, there were significant differences in characteristics compared to low-risk areas, including a higher proportion of African Americans (64% vs. 21%, p<0.001), lower median household income ($30,948 vs. $54,388, p<0.001), and a higher proportion living below the poverty level (36% vs. 20%, p<0.001). There was a 3-fold increase in the adjusted OHCA incidence between high and low risk areas (1.68 vs. 0.57 per 1,000, p<0.001) with BCPR rates of 27% and 31% (p=0.31), respectively. Compared to a previous analysis, 9 (36%) census tracts persisted as high-risk but an additional 16 were newly identified. Conclusions: Neighborhood-level variations in OHCA incidence are dramatic with marked disparities in characteristics between high and low risk areas. It is possible that improving OHCA outcomes requires multifaceted interventions to address social determinants of health.