Using GIS to focus your sampling efforts: A community-based example

Accurately identifying your target population for a community-based survey can be time consuming and expensive. Geographic Information Systems (GIS) can be used, along with other data, to assist in locating populations of interest. Our study utilized census information and data from a local health district and the state health department to create a set of variables which were overlaid onto the county's census tracts. The variables of interest in this study were factors that predict utilization of services provided by community providers who serve un- and underinsured people. A scoring system was developed where each variable was ranked from 1-50 (for the 50 census tracts in the county). The 14 variables of interest were then summed, giving each census tract a composite score from 74 to 680 out of 700 possible points. The top five census tracts were selected for follow-up. Once the census tracts were selected, specific neighborhoods within each census tract were identified. This session will discuss how GIS was used to focus the surveying efforts for our study with the local health district. We will discuss geocoding health data, protecting the confidentiality of client/patient information, obtaining current census tract estimates, and obtaining data from state programs (e.g. the Women, Infants, and Children Nutritional Program) and birth certificates. We will discuss how these data sources were used to develop a ranking strategy that resulted in the differentiation of census tracts, which helped in the selection of census tracts most likely to contain the highest concentration of our target population. We will also discuss how we compared these data to the health district's client information to differentiate between census tracts that had high potential need for service versus those with high utilization.