Predicting the long-term uninsured population and analyzing their gaps in physical access to healthcare in South Carolina

Objective: South Carolina (SC) opted not to expand Medicaid under the Patient Protection and Affordable Care Act. The long-term uninsured (LTU) is a core portion of the uninsured population. We aim to predict areas with high concentration of the LTU and analyze LTU’s physical access to healthcare in terms of providers’ quantity, types, and location in SC.

Methods: The geographical unit of analysis is Zip Code Tabulation Area (ZCTA). Socioeconomic and demographic data are extracted from the American Community Survey. We build a statistical model to predict the LTU with 16 independent aggregate variables (age, gender, race, education, family status, employment, and poverty). Propensity scores are computed to categorize 424 ZCTAs into the lowest (Q1) and highest (Q5) quintiles of the LTU concentration. Socioeconomic characteristics of each quintile are then estimated and compared across the quintiles. Care provider types, often utilized by uninsured individuals including free health clinics, federally qualified health centers, rural health clinics, and Welvista clinics, are geocoded. Types of clinics and their availabilities are computed and compared: ZCTAs without any kind of clinics, average number of specific types, and number of all clinics per 100,000 population. Using ArcGIS, a set of color-coded maps are created to provide geovisualization of analytical results.

Results: First, ZCTAs with highest LTU concentration, on average have higher rates of minorities, lower socioeconomic status, lower education level, and higher unemployment than ZCTAs with less LTU concentration. Second, there is a significant number of ZCTAs that have high LTU concentration but do not have high Medicaid coverage. Third, about half of ZCTAs in each of quintiles Q3-Q5 do not have any health clinic within their boundaries. Fourth, free health clinics - the type of health care providers that is most likely to be used by the uninsured population is least available in Q5. Last but not least, the average number of clinics per 100,000 population is lowest in Q5.

Conclusion: LTU is a hard-to-reach group with structural socioeconomic disadvantages. The uninsured, especially LTU lack not only health insurance but also physical access to care. While free clinics are effective in reaching out to these populations, their scope is limited because professional services are provided almost entirely by volunteers and local hospitals for needed technical services. Expansion of health insurance does not automatically solve the access gap for the uninsured unless health care resources are more available in the surrounding areas.