Mapping the uninsured using secondary data: an environmental justice application in Dallas

Over the last 5 years, environmental justice (EJ) researchers have been calling for incorporation of health outcomes more directly into spatial studies of socio-demographics and environmental hazards. To date, researchers have not incorporated insurance status (an access to health care variable) in their models although access to care likely has an important association with the probability of health effects due to environmental exposures. As such, insurance status represents an important variable within spatial EJ studies focused on health, and the lack of spatially explicit access to care data is a critical limitation in the field. As a solution, we offer a method of using uninsured appendicitis cases, acquired secondarily from state hospital admissions data, to estimate rates of uninsurance at the zip-code level. We apply the technique to explore relationships between cancer risk from hazardous air pollutants and estimated rates of uninsurance, a previously unexplored phenomenon. Then, we compare the uninsurance findings to those related to poverty to illustrate how uninsurance, as a variable, compares to a more traditional socio-economic predictor used in EJ studies. The relationship between cancer risk from hazardous air pollutants and uninsurance is weaker than the relationship between risk and poverty, but both are statistically significant. As such, we conclude with a discussion of the importance of considering insurance status in spatial studies of EJ focused on health.