The aim of this project was to broaden the RUPRI Center's longtime focus on closures of independent pharmacies in rural places by identifying rural communities that lost all retail pharmacies from 2003 through 2019. This work expanded previous Center work beyond a focus on local independent pharmacies by incorporating chain pharmacies, including pharmacies located in larger retail outlets such as grocery stores and merchandising stories (e.g., Shopko, Walmart). First, the Center identified the number of pharmacy closures over time by the type of pharmacy (independent, chain) and by rurality.

1. Hypothesis 1: The Center hypothesized that it would see recent increases in closures of chain pharmacies or pharmacies that are part of a larger retail outlet in rural areas.
   
   Second, the Center identified pharmacy deserts (i.e., zero pharmacies) at the Primary Care Service Area (PCSA) level by rurality.
2. Hypothesis 2: The Center hypothesized that the number of pharmacy deserts in rural areas was increasing.
   
   Aim 2 of this project used multivariate regression techniques to compare the characteristics of communities with no pharmacy (pharmacy deserts) to those with at least one pharmacy. More specifically, the Center tested the following hypotheses using a multivariate regression framework.
3. Hypothesis 3: Pharmacy deserts are associated with higher rates of poverty, higher percentages of persons insured through Medicare and Medicaid, and higher percentages of uninsured.
4. Hypothesis 4: Pharmacy deserts are associated with low population density and/or population declines.
5. Hypothesis 5: Pharmacy deserts are associated with declines in the number of primary care providers in the community.

Design: For Aim 1, Hypothesis 1, the Center applied an approach previously used by the RUPRI Center to classify and track rural pharmacies. Monthly data was evaluated to determine if a pharmacy is closed or open. The Center considered a pharmacy closed if the provider identification number ceases to be included in the monthly data set or when a closure date was specified. Care was taken to assess other pharmacy activity in the geography to account for facility sales or transfers.

For Aim 1, Hypothesis 2, and Aim 2, Hypotheses 1-3, the Center created a longitudinal data set of pharmacy deserts at the PCSA level from 2003 to 2019 by linking multiple data sources. For each year, the Center classified a given PCSA as a pharmacy desert if it did not have any pharmacy operating within its boundaries. The Center merged data on community characteristics, such as poverty level, population count and age distribution, population gain or loss, and insurance status from the American Community Survey using PCSA component ZIP code tabulation areas. Community primary care providers were tabulated from National Provider Identifier data.

Publications

• Rural and Urban Pharmacy Presence – Pharmacy Deserts
  Policy Brief
  RUPRI Center for Rural Health Policy Analysis
  Date: 08/2022
  This policy brief continues the RUPRI Center's ongoing examination of the availability of community pharmacies and their provided services in rural areas of the U.S. The brief also provides a deeper analysis of counties with no retail pharmacies (i.e. pharmacy deserts) based on metropolitan/nonmetropolitan locations.