Urgent care centers and retail clinics have been touted as a solution to the national problem of non- emergent emergency department (ED) use. Specifically, they may improve access to care in vulnerable areas where the ED offers the only safety-net or after-hours care option. Since these clinics offer non-emergent care that is less costly but of similar or better quality to that delivered in the ED, rerouting non-emergent patients to these clinics may yield efficiency gains. Despite these purported benefits, these alternative care sites remain largely unexplored in the health services literature. Little is known about where these centers locate, whether they are geographically accessible and open when needed, and the impact they have on rates of non-emergent ED use. This knowledge gap is partially due to a lack of data on the clinics and their patient visits, but also because of endogeneity involved when a patient chooses to visit either the ED or an alternative site for care. The proposed project will address these gaps in the literature, providing important foundational information about these clinics. In particular, the aims of this project are to: 1) use a national-level dataset of urgent care centers and retail clinics to examine the market characteristics associated with the location of these care sites; 2) estimate by payer source (Medicaid, uninsured, private insurance) the percentage of non-emergent ED visits that could feasibly be treated in alternative care sites, given real-world access barriers like geographic distance and hours of operation; 3) examine the effect of alternative care site presence on the rate of non- emergent ED use in local hospitals, assessing whether the relationship varies by hospital payer mix. To meet these aims, each clinic location and its hours of operation has been geocoded with GIS software, using data from an exclusive, national-level database obtained from a private vendor. Characteristics of the surrounding census tract-level markets have been linked to each location, using data from a variety of sources, including the U.S. Census American Community Survey and Business Patterns surveys. Multivariate regression models will be estimated to determine whether clinics locate in areas with vulnerable populations. Using the HCUP State Emergency Department Databases (SEDD), linked with the American Hospital's Annual Survey data, non-emergent ED visits will be analyzed by patient zip code, hospital location, time of day, and insurance status to determine the number of visits that could feasibly have been treated in a local alternative care site. Finally, a regression discontinuity design will be used to estimate the effect of alternative care site presence on the rate of local non-emergent ED visits, by taking advantage of daily closing times of clinics.
/Relevance to Public Health Urgent care centers and retail clinics represent a large, rapidly growing, and gravely unexamined sector of acute healthcare delivery in the U.S. The proposed project will provide important, foundational knowledge about the potential and actual impact of these centers on access to acute care for minority populations, and non- emergent ED use. Its findings will help inform both policy and research agendas aimed at improving healthcare access and efficiency, especially for vulnerable populations and Medicaid recipients.