In 1981, pioneering healthcare geographers Roy Pechansky and Thomas William developed a framework to understand access to healthcare both spatially and spatially. This framework, now referred to as “The Five A’s of Access” by health geographers is still used today to understand health behavior. One of these As is availability; a measure of the magnitude of services available to the population (Pechansky and Thomas, 1981.) This measure is a container-based and typically aggregated over a geographic area and normalized to a population. In addition to obstacles such as insurance, transportation, and time, research has also shown the provider supply is an issue for accessing care in North Carolina, as a shortage has been ongoing (Lichtstein et al.,2009.) The majority of North Carolina residents live in urban centers, which are clustered in the central region of the state. The eastern and western regions are much more rural, and residents in these regions tend to have worse provider/person ratios and are likely to be further away from hospitals that treat ambulatory conditions (Health Resources and Administration, 2023.) It is important that we understand the extent of these disparities to best serve populations in all regions of North Carolina, which this analysis aims to do by studying the volume of hospital beds available per 1000 county residents.

This analysis focuses on the question: Do the different regions of North Carolina differ in the number of hospitals per 1000 residents? Using county-level data from the 2020 census, an ANOVA (analysis of variance) statistical test will be used to answer this question. The variable of hospital beds per 1000 residents is an important measure used in accessibility studies in healthcare geography (Levesque et al., 2013.) Understanding the volume of services available to people can help us understand how people interact with the healthcare system, and what changes can be made to ensure people have access to efficient, local care.

ANOVA (analysis of variance) is a statistical test used to determine if 2 or more population means are different. It works by comparing the ratios of variance within a group and between groups. If there is more variance between groups than within groups, than the means are statistically different (Soetewey, 2025.) For this project, I ran ANOVA to determine if there are differences in means between the eastern, central, and western regions of NC.

The ANOVA results indicated there is no significant statistical difference in beds per 1000 residents across western, central, and eastern North Carolina. This differs from what I expected, since research has shown that rural areas have a lower volume of resources per resident, and eastern and western NC are much more rural than central NC. However, eastern NC does have only 2.03 hospital beds per 1000 residents, which is less than the national average of 2.35. Even though there was no statistically significant difference by region, there is are still disparities between rural and urban populations, and NC is still below the national average with roughly 2.1 beds per 1000 residents (Trilliant Health, 2024.) Analysis was done at the county-level, but further research could be done at the census tract-level to see if there are statistically significant differences at this scale, which may not be evident at the county level due to the MAUP (modifiable areal-unit problem.) Additionally, a container-based approach was used for this analysis. It did not account for the distribution of populations within a county, or the fact that county boundaries are entirely arbitrary, so people often travel across county lines for care. Future analysis could be done at varying scales to observe possible trends, and other measures of availability can be used such as physicians per 1000 residents to better understand access to care in NC.