CARERC Pilot: Geocoding and Environmental Exposure Assessment in Kentucky

Abstract Researchers are increasingly employing geocoding to identify precise locations of study participants'' previous residences to identify historical and long term trends in disease, and assess retrospective environmental exposures, particularly in case-control studies. However, there is a substantial risk of misclassification of exposure when using older addresses records, as past redevelopment, and readdressing can alter or remove addresses, rendering them unable to be located (or located precisely) in current datasets. Furthermore, evidence suggests that demographic characteristics, notably minority race/ethnicity and low income, are associated with poor geocoding precision. It is possible that such socioeconomic disparities in geocoding are magnified or otherwise influenced by changes in addressing reference data over time. These communities are also at higher risk for environmental exposures, as minority and low-income communities have historically been forced into areas adjacent to industrial activity or near nuisance sites, such as quarries, landfills, or major roadways and related infrastructure. To address the existing limitations and assess disparities with historical geocoding, we propose a pilot study that will identify geographic areas where standard geocoding protocols are less likely to locate individuals precisely, assess demographic characteristics associated with geocoding precision, and improve geocoding by building and testing a geocoding protocol to evaluate improved locational precision and environment exposure assessment.