Identifying Unique Neighborhood Characteristics to Guide Health Planning for Stroke and Heart Attack: Fuzzy Cluster and Discriminant Analyses Approaches

Abstract

BackgroundSocioeconomic, demographic, and geographic factors are known determinants of stroke and myocardial infarction (MI) risk. Clustering of these factors in neighborhoods needs to be taken into consideration during planning, prioritization and implementation of health programs intended to reduce disparities. Given the complex and multidimensional nature of these factors, multivariate methods are needed to identify neighborhood clusters of these determinants so as to better understand the unique neighborhood profiles. This information is critical for evidence-based health planning and service provision. Therefore, this study used a robust multivariate approach to classify neighborhoods and identify their socio-demographic characteristics so as to provide information for evidence-based neighborhood health planning for stroke and MI.Methods and FindingsThe study was performed in East Tennessee Appalachia, an area with one of the highest stroke and MI risks in USA. Robust principal component analysis was performed on neighborhood (census tract) socioeconomic and demographic characteristics, obtained from the US Census, to reduce the dimensionality and influence of outliers in the data. Fuzzy cluster analysis was used to classify neighborhoods into Peer Neighborhoods (PNs) based on their socioeconomic and demographic characteristics. Nearest neighbor discriminant analysis and decision trees were used to validate PNs and determine the characteristics important for discrimination. Stroke and MI mortality risks were compared across PNs. Four distinct PNs were identified and their unique characteristics and potential health needs described. The highest risk of stroke and MI mortality tended to occur in less affluent PNs located in urban areas, while the suburban most affluent PNs had the lowest risk.ConclusionsImplementation of this multivariate strategy provides health planners useful information to better understand and effectively plan for the unique neighborhood health needs and is important in guiding resource allocation, service provision, and policy decisions to address neighborhood health disparities and improve population health.

Association of a higher density of specialist neuroscience providers with fewer deaths from stroke in the United States population

“…Although data on other specific risk factors were not available in the ARF, our multivariate regression analysis demonstrated a sustained effect after controlling for socioeconomic status and urbanicity of the county. In the literature, 6,12,16,18,25,30,33,36 there is an established association of lower socioeconomic status with increased cardiovascular risk factors in the US, allowing us to hypothesize that the latter should not be affecting our analysis after controlling for the former. In addition, the present analysis was performed on the county level, recognizing the importance of small-scale variations as underlined by previous studies.…”

“…36 Various groups have investigated the association of various factors with this phenomenon. 16,25,30,33,36 Most studies, however, have focused on factors associated with increased stroke incidence rather than higher case fatality. Lower socioeconomic status, hypertension, and diabetes have most commonly been associated with higher stroke prevalence.…”

“…In addition, the present analysis was performed on the county level, recognizing the importance of small-scale variations as underlined by previous studies. 30 It has been shown 1,29 that comprehensive stroke centers have a beneficial impact on patient outcomes. A health system can be described as comprising 3 domains: 7 1) structure (buildings, equipment, human capital); 2) process (triage, operational algorithm); and 3) outcomes.…”

A typology of neighborhoods and blood pressure in the RECORD Cohort Study

“…Hierarchical cluster analysis using Ward's method starts with each multidimensional observation (neighborhood) as a single cluster and then repeatedly merges the next two closest clusters in terms of Euclidian distances between observations until a single, all-encompassing cluster remains [30]. Application of this method results in a neighborhood typology wherein neighborhoods that are substantively comparable on selected characteristics are grouped together even though they are not necessarily geographically adjacent [31][32][33][34][35]. Following assessment of corresponding dendograms, we examined results for n ¼ 4 to n ¼ 7 clusters, attempting to identify substantively distinct neighborhood types appearing at each separation point.…”

Small-area Variation in Hypertension Prevalence among Black and White Medicaid Enrollees

“…6,32 Moreover, small-area analyses of hypertension prevalence can guide local place-based planning, health programming, and community resource development efforts. 35,36 The observed geographic overlap of hypertension prevalence hotspots for Black and White Medicaid enrollees in South Carolina provides an opportunity to target interventions to improve health outcomes in low-income populations. Our findings highlight the need to better understand and address racial residential segregation and its health consequences, and indicate the value of ACA provisions establishing provider network adequacy standards for Medicaid managed care plans in ensuring health care accessibility for urban and rural residents with hypertension and related chronic conditions.…”