COVID-19 Report Series: Medical Risk

This report was created to help guide strategic planning in response to the COVID-19 outbreak in Pennsylvania by identifying counties in Pennsylvania with populations that have medical risk factors relating to COVID-19. Rather than mapping current hot spots of COVID-19, this report's purpose is to highlight areas where the outcomes from COVID-19 may be more severe.

The Centers for Disease Control and Prevention (CDC) list age and chronic illnesses like heart disease and diabetes medical risk factors for COVID-19. Pennsylvania ranked among the 10 states with the highest population over 60 years of age in 2018 [U.S. Census Bureau, Population Estimates Division] and among the top 10 states for hospitalizations related to cardiovascular disease in the population age 65 and over during 2014-2016 [CDC, Atlas of Heart Disease and Stroke].

In order to highlight which counties have the most vulnerable populations, medically, we collected data for indicators relating to medical risk (age, cardiovascular disease, diabetes), healthcare services risk (available hospitals, staff, and beds), and proximity risk (population density, nursing home occupancy) from the U.S. Census Bureau, the Centers for Disease Control and Prevention (CDC), and the Pennsylvania Department of Health.

We then ranked each indicator on a scale of 1 to 67, with 1 representing the lowest amount of vulnerability (e.g., a population with less older adults or more beds and doctors per person) and 67 representing the highest amount of vulnerability (e.g., a population with a large percentage of seniors in nursing homes or a large diabetes prevalence rate).

Ranks were averaged across indicators in each risk category (i.e., medical risk, proximity risk, and health services risk) and then each risk category index was averaged into a final Medical Vulnerability Index (MVI). For clarity the index values are expressed as percentages.

The following maps present each indicator used to develop the index and the final MVI values for each of Pennsylvania’s counties. Hover over the icon in the bottom left of the map frame to view the legend.

Older age is a significant risk factor that can worsen health outcomes for COVID-19 sufferers. Figure 1 shows the percentage of the population that is 60 years or older by county according to the U.S. Census Bureau’s Population Estimates Division.

Sullivan County had the highest percentage of population age 60 and over (39.0%). Sullivan was followed by Cameron (36.6%), Susquehanna (32.3%), Wayne (32.3%), and Potter (32.2%). Philadelphia County had the lowest percentage of population age 60 and over (19.3%). Philadelphia was followed by Centre (19.9%), Chester (23.0%), Lehigh (23.0%), and Delaware (23.3%).

Outcomes from COVID-19 worsen with increased age. Figure 2 shows the percentage of the population that is 85 years or older by county according to the U.S. Census Bureau’s Population Estimates Division.

Cameron County had the highest percentage of residents age 85 and over (4.1%). Cameron was followed by Montour (3.6%) and a four-way tie between Cambria, Lawrence, Mercer, and Sullivan (3.4%). Monroe County had the lowest percentage of residents age 85 and over (1.7%). Monroe was followed by Perry (1.8%), Centre (1.9%), Philadelphia (1.9%), and Forest (2.0%).

Cardiovascular issues are linked to increased mortality from COVID-19. Figure 3 shows the number of hospitalizations from 2014 to 2016 due to cardiovascular disease for the population 65 and older by county as reported by CDC’s Interactive Atlas of Heart Disease and Stroke. 

Venango County (260.4 per 1,000) had the highest rate of cardiovascular disease related hospitalizations. Carbon (252.7), Indiana (249.2), Mercer (246.0), and Lawrence (244.6) also had some of the highest rates of hospitalization for cardiovascular disease for those over 65 years of age. Potter County had the lowest cardiovascular hospitalization rate in Pennsylvania (125.1). Susquehanna (144.2) and Fulton (145.6), Lancaster (154.5) and Lebanon (160.6) rounded out the five counties with the lowest hospitalization rates. 

Complications related to diabetes are also negatively associated with COVID-19 outcomes. Figure 4 shows the percentage of the population age 20 and over with diabetes in 2016 by county according to the CDC’s United States Diabetes Surveillance System.

The county with the highest rate of diabetes was Fulton County (15.4%). Mifflin (14.7%), Huntingdon (14.2%), and Venango (13.0%) followed in the ranking of the highest rates, while Blair and Clarion (12.6%) tied for the fifth highest percentage. Perry County had the lowest percentage of individuals over 20 years of age with diabetes at 5.8%. Following Perry were Carbon (6.3%), Bucks (6.4%), Adams (6.6%), and Montgomery (7.0%). 

Pennsylvania is a large state, and hospitals, along with the emergency services they provide, are unevenly distributed across the Commonwealth. Figure 5 shows the density of General Acute Care (GAC) hospitals (number of hospitals per 1,000 square miles) as of December 31, 2018 by county using hospitals listed in the Pennsylvania Department of Health – Division of Informatics’ 2018 Hospital Report.

Philadelphia had the greatest density of hospitals at 111.7 hospitals per 1,000 square miles in 2018. Philadelphia was followed by Delaware (21.8), Allegheny (19.2), Montgomery (18.6), and Lehigh (11.6). Seven counties had no GAC hospitals included in the in 2018 Hospital Report, including: Cameron, Forest, Juniata, Perry, Pike, Snyder, and Sullivan.

As the number of COVID-19 cases in Pennsylvania continues to grow, the number of hospital beds available is critical for managing the increase. Figure 6 shows the number of beds per 100,000 residents in GAC hospitals by county in 2018 according to the Pennsylvania Department of Health – Division of Informatics’ 2018 Hospital Report.

Montour County had the highest rate of beds per resident (3,065 per 100,000 residents) due to a combination of a single but well-staffed GAC hospital and a smaller resident population. Montour was followed by Bradford (538), Lehigh (507), Allegheny (441), and Dauphin (434). As seven counties had no GAC hospitals, they had rates of 0 beds per resident (see Hospital Facilities).

A  2019 study from the Association of American Medical Colleges  highlights the physician shortage in the United States. That report showed that rural areas are the most under-served. Figure 7 shows the rate of physicians (including physician assistants) hired or contracted full or part time at GAC hospitals per 100,000 population by county according to the Pennsylvania Department of Health – Division of Informatics’ 2018 Hospital Report. 

Montour County had the highest rate of physicians per resident (4,304.0 per 100,000 residents). Montour was followed by Dauphin (546.0), Potter (295.0), Union (279.0), and Clearfield (263.0). Since seven counties did not have GAC hospitals, they also had rates of 0.0 physicians per resident (see Hospital Facilities).

Nurses are a critical part of healthcare infrastructure and the capacity to treat those affected by COVID-19. Figure 8 shows the rate of nurses (excluding nurse anesthetists) per 100,000 persons hired or contracted full or part time at GAC hospitals by county according to the Pennsylvania Department of Health – Division of Informatics’ 2018 Hospital Report. 

Montour County, again, had the highest rate of nurses (10,082.2 per 100,000 residents). Montour was followed by Dauphin (545.6), Lehigh (1305.4), Clearfield (1,039.2), and Bradford (1,009.3). Again, since seven counties did not have GAC hospitals, they also had rates of 0.0 nurses per resident (see Hospital Facilities).

Population density can play an important role in the transmission of disease. Figure 9 shows the population density per square mile by county according to the U.S. Census Bureau’s 2018 Population Estimates. 

Philadelphia had the highest population density at 11,801.4 persons per square mile. It’s neighboring county, Delaware, which ranked second, was nearly a quarter as dense (3,072.3 persons per square mile). Montgomery (1,715.7), Allegheny (1,668.8), and Lehigh (1,066.3) rounded out the top five densest counties in Pennsylvania. Cameron County was the least dense with 11.3 persons per square mile. Sullivan (13.5), Potter (15.4), Forest (17.0), and Fulton (33.2) were the other counties with the lowest population densities. 

Nursing homes have been hotspots for COVID-19 and often contain a large portion of vulnerable population in a single location. Figure 10 shows the percentage of the population 60 years of age and over that are living in nursing homes by county as reported by the Pennsylvania Department of Health – Division of Informatics’ Nursing Home Report.

Sullivan County had the highest percentage of individuals 60 years of age or over that were living in nursing homes (6.4%). Sullivan was followed by McKean (4.5%), Montour (4.5%), Forest (3.8%), and Blair (3.5%). Pike County had the lowest percentage of individuals 60 years of age or over that were living in nursing homes (0.6%). Pike was followed by Monroe (1.0%), Bedford (1.2%), Snyder (1.3%), and Armstrong (1.3%).

Nursing homes that are fuller have an increased likelihood of spreading communicable diseases quickly and overwhelming staff. Figure 11 shows the average occupancy rate of nursing homes by county as reported by the Pennsylvania Department of Health – Division of Informatics’ Nursing Home Report.

Fulton County had the highest average occupancy rate among its nursing homes (96.5%). Fulton was followed by Mifflin (94.9%), Montour (94.0%), Juniata (93.5%), and Sullivan (93.5%). Warren County had the lowest average occupancy rate among its nursing homes (63.8%). Warren was followed by Clarion (74.9%), Wayne (76.3%), Armstrong (76.4%), and Potter (81.4%). 

The Medical Vulnerability Index (MVI) is the combination of all the factors discussed above to provide an indication of how high the medical risk is for each county. The percentage value for MVI represents to what percentage a county ranked across each indicator, where 100% would indicate that a county ranked highest among each indicator.

Values for MVI ranged from 19.4% (Dauphin) to 52.3% (Sullivan). The three counties with the highest MVI values included Sullivan (52.3%), Northumberland (48.2%), and Juniata (45.9%). Figure 12 shows the MVI values for each county in Pennsylvania.

Sullivan County had the highest MVI value across Pennsylvania counties (52.3%). It’s high MVI value was a result of ranking highest for the percentage of residents in a nursing home, and ranking highly for the percent of the population age 85 and over, diabetes prevalence, unfavorable healthcare capacity (specifically, facilities per square mile, nurses, and beds), and occupancy rates of nursing homes.

Northumberland had the second highest MVI value (48.2%). Northumberland’s high MVI value was driven in part by its overall population density and a lack of healthcare infrastructure to support the county’s population. It also has a large percentage of its population age 60 and over in nursing homes.

Juniata County had the third highest MVI value (45.9%). Juniata County’s high MVI value was driven largely its lack of healthcare services, having no GAC hospitals. It also had a high average occupancy rate among its nursing homes.

Dauphin County, notably, had the lowest MVI value (19.4%). The most unfavorable risk indicator was for diabetes prevalence, for which it ranked moderately high (27th in the state). Otherwise, it had a low percentage of population over 60, comparatively, and a robust healthcare infrastructure. 

1. Centers for Disease Control and Prevention. Age-adjusted Diabetes prevalence rates for population 20 and over, 2016. United States Diabetes Surveillance System.
2. Centers for Disease Control and Prevention. Smoothed Cardiovascular Disease death rates for the total population, 2014-2016. Interactive Atlas of Heart Disease and Stroke.
3. Centers for Disease Control and Prevention. Smoothed Cardiovascular Disease death rates for population 65 and over, 2014-2016. Interactive Atlas of Heart Disease and Stroke.
4. Centers for Disease Control and Prevention. Smoothed Cardiovascular Disease hospitalization rates for population 65 and over, 2014-2016. Interactive Atlas of Heart Disease and Stroke.
5. Pennsylvania Department of Health. Hospital Reports, 2018.
6. Pennsylvania Department of Health. Nursing Home Reports, 2018.
7. U.S. Census Bureau. County Population Estimates by Characteristics. Vintage 2018 Population Estimates Program.

All rates developed for the report (e.g., staff per 100,000, beds per 100,000, percent of population 60 and over in nursing homes, etc.) used the Vintage 2018 Population Estimates, except in cases where the data were provided as a rate (e.g., cardiovascular hospitalizations, diabetes prevalence, etc.).

All maps and analyses were produced by the Pennsylvania State Data Center (PASDC) at the Institute of State and Regional Affairs of Penn State Harrisburg. We would like to thank our Research Assistant, Jennifer Straub, for all of her excellent and expeditious work in getting this report published.

Questions? Get in touch with us at pasdc@psu.edu or on Twitter  @pasdc_psu !