The Road to a Rural Health Atlas Runs Through Community Hospitals

October 2018
Topics: Health, Public Health, Data Analytics, Decision Support (General), Health Innovation
Rural hospitals are closing all over the nation, limiting residents’ access to health resources. MITRE is using data analytics to help one state see how such closures could affect local communities and consider possible interventions.
Roan Mountain Medical Center

An estimated 83 hospitals have closed since 2010 in the United States and another 673 are considered vulnerable or extremely vulnerable. As a result, in many rural areas, people are now further away from hospitals and other healthcare resources. This is a growing problem as people living in rural areas have high prevalence of diabetes, hypertension, obesity, and other chronic diseases, including substance use disorders.

MITRE saw states struggling with decisions about providing healthcare resources and knew data analytics could help them make critical decisions that would improve rural health outcomes.  We’re now working with South Carolina to construct a Rural Health Atlas that illustrates how hospital closings affect the health of rural populations.

Our researchers are using public health-related data, including claims records, to understand the causes and impacts of rural hospital closures. The data will also help states develop interventions to minimize any negative effects on their communities’ overall health. MITRE’s eight-member team is led by principal investigators Angelica Cristello, a senior public health/healthcare domain specialist, and Grace Moon, a lead scientist, both in our Public Health and Innovation Department.

The Rural Health Atlas is the product of a partnership between MITRE’s independent research and development program and the South Carolina Office of Rural Health. According to Business Insider, South Carolina ranks among the top 10 unhealthiest states in the nation and has the highest risk of hospital closure: 50 percent of hospitals are at risk, according to a study published in 2017.  Meanwhile, the national average risk of closure is 41 percent.

“Because rural healthcare is facing significant challenges, MITRE has conducted numerous research projects with various states on how to assess and improve access to care,” explains Dr. Sybil Klaus, who leads MITRE’s health research program. “We’re combining our public health and healthcare knowledge with our strong technical capabilities to research and develop new ways of using health related data to improve health outcomes.”

Countering Disparities Between Rural and Urban Health

"We saw that local program and policy makers don’t typically have access to comprehensive data sets that allow them to fully understand all the factors that contribute to poor health outcomes," Moon says. "Our project aims to fill that gap. We're looking at different sets of data about the factors that shape communities’ health. That includes data related to healthcare resources and hospitals in danger of closing. It also includes public health data, which encompasses demographics such as education level and income information. Pulling all that together will allow policy makers to make evidenced-based decisions on how to achieve better rural health outcomes."

"The disparities between rural health areas and their urban counterparts are large and growing," Cristello adds. "We’re creating a tool that enables decision makers to visualize the impact of health system changes. Initially, we’re focusing on hospital closures. To know where and how to intervene, we need to visualize the impact of more than just healthcare-related changes, but we’re starting with hospital closures and shifts in provider availability.”

The Rural Health Atlas combines layers of numerical data about the potential closing of a hospital and the severity of its impacts. The tool will let users visualize interrelated data, including hospital closures and other healthcare data, together with data about health policies and factors such as employment, education, transportation, broadband network availability, and disease burden.

That combination of data is converted into an interactive “big picture” that shows the potential effects of the closure. For example: Imagine a computer screen that shows you all the hospitals and all the outpatient facilities in your area. As you hover your cursor over a hospital, you can see which counties that facility serves, based on the number of patients who use it. You can also see things like health rates in the service area. If that hospital were to close, you could see the increase in transportation, bed, and provider requirements for moving patients to other facilities, and the types of patients—based on demographics or health condition—who would be more likely to struggle to find resources close by.

To Close or Not to Close—It's a Complex Question

States and their communities need to weigh the pros and cons of many factors when making decisions about closing hospitals. These include: What is the actual impact of this closure? Can the surrounding area and infrastructure absorb the demand that this hospital can no longer meet? If not, what are the types of interventions that could mitigate the negative effect?

In some cases, a hospital doesn’t have enough demand to financially support it. Then decision makers would look at surrounding outpatient facilities, or another hospital not too far away. Perhaps subsidizing transportation to another hospital would be possible instead of keeping an underused location open.

Then there are resources not in public databases. "Thousands of churches in South Carolina play a crucial role in the social determinants of health such as food and transportation, and so we want to make sure relevant context-specific data sources are included where available" Cristello says.  

As Cristello and Moon continue developing the Rural Health Atlas for South Carolina, they see it as a tool that can be applied to other states. "We're sowing the ground to replicate the Atlas nationally by using data sets like the American Family Survey and the national Behavioral Risk Factor Surveillance System," Cristello explains.

Moon also sees the Atlas methodology applied to broader settings. "In addition to state and federal views, the tool can be customized to examine lower levels, down to the county or other designations relevant to decision makers. This could be especially useful for large states such as California."

Those interested in this data analysis tool, or other MITRE innovations, can contact us at

by David A. Van Cleave

Explore more at MITRE Focal Point: Health.


Interested in MITRE's Work?

MITRE provides affordable, effective solutions that help the government meet its most complex challenges.
Explore Job Openings

Publication Search