Spatial Epidemiology and Temporal Trends of Heart Attack and Stroke in Middle Tennessee

Despite declines in mortality risks of myocardial infarction (MI) and stroke in the US since the 1960’s, the burdens of these conditions remain high. These conditions require emergency and specialized care and therefore quick transportation of patients to appropriate hospitals is critical. Geographic disparities in MI and stroke burdens have been consistently reported in the US with the south-east having the highest risks. Most studies of geographic disparities have been performed at county or higher geographic units. Therefore, spatial patterns at neighborhood levels are unclear. Moreover, it’s important to investigate disparities at neighborhood levels to better understand neighborhood health needs. Therefore, the goal of this study was to investigate neighborhood disparities associated with MI and stroke in Middle Tennessee. Specific objectives were to investigate: (a) geographic disparities in timely access to emergency care; and (b) geographic disparities in MI and stroke mortality risks.

Street network, hospital, population, and mortality data (1999-2007) were obtained from Streetmap USA, the Joint Commission on Accreditation of Health Organizations, US Census Bureau, and the Tennessee Department of Health, respectively. Network analysis was used to investigate and identify neighborhoods lacking timely access to emergency MI and stroke care. Moran’s I and Kulldorff’s spatial scan statistic were used to investigate geographic hot-spots of MI and stroke mortality risks at both the county and neighborhood levels. Poisson and negative binomial models were used to investigate predictors of identified geographic patterns.

A temporal increase in the percentage of the population with timely geographic access to stroke and cardiac centers was observed. In 2010, about 5% of the population, located mainly in rural neighborhoods, lacked timely access to a stroke center. Significant (p<0.05) hot-spots of MI and stroke mortality risks were identified at both neighborhood and county levels. However, clusters identified at the neighborhood level were more refined. Neighborhoods with higher proportions of older populations and those with lower education had significantly (p<0.05) higher mortality risks.

These findings are vital for guiding health planning, resource allocation and service provision in an effort to provide needs-based services to the population. This is important in reducing/eliminating neighborhood disparities and improving population health.