This map displays several neighborhood-level measures of populations with greater risk associated with COVID-19. To develop the indices, we used demographic and health characteristics thought to increase a person's risk for severe symptoms or death, as identified by national and world health organizations.
Centers for Disease Control and Prevention says that "some people are at higher risk of getting very sick from this illness, [including] older adults [and] people who have serious chronic medical conditions like heart disease, diabetes, [and] lung disease." The National Cancer Institute says that people with certain types of cancer or treatments can have weaker immune systems, making them more susceptible to COVID-19. The National Institute on Drug Abuse says those who smoke may also be more susceptible to the disease. The World Health Organization suggests that people with asthma may be more vulnerable to a serious respiratory infection from Coronavirus.
Using demographic data from the U.S. Census Bureau and estimates of neighborhood health data from the CDC's 500 Cities Project, we can estimate areas with high-risk populations.
Because it's hard to understand eight different maps, each with a different risk factor, we used statistics (specifically principle component analysis) to simplify and synthesize the information. We found that there are two distinct groups of indicators that relate to risk populations. These roughly split our eight indicators into two groups of four, with one group related to socioeconomic status and one group related to age.
The socioeconomic risk group contains behaviors and outcomes that tend to relate to income and economic opportunity. These are the uninsured population, asthma, smoking rates, and diabetes.
The age risk group contains demographics (age itself) and health outcomes that are linked to older age populations. This includes cancer and heart disease.
Neighborhoods just outside of downtown Indianapolis have the highest estimated risk when it comes to factors that are related to income. The neighborhood just north of 30th Street and east of Fall Creek (tract 3508.00) has one of the highest socioeconomic risk index scores. In this neighborhood, an estimate one-fifth of the population lacks health insurance, one-sixth of adults have asthma, one-third smoke, and a quarter have diabetes. The area near New York Street and Sherman Drive also rates as very high risk on this index due to a high concentration of people who lack health insurance, have asthma, and smoke.
Age-related risk is more geographically diverse. Areas in northern Marion County have high estimated rates of cancer and have older populations. The Martindale-Brightwood and Meadows neighborhoods (northeast of downtown) have high estimated rates of heart disease and some pockets of older populations. The Far Eastside and the far southside (near U.S. 31 and County Line Road) also have high risk populations according to the age-related risk factors.
To use this map, click on a neighborhood area (census tract) and explore its risk factors in the legend to the left. Expand each index to see the individual variables that have the strongest impact on that measure. To change the data displayed on the map, click “map this” located next to the index name or variable name in the legend. If you zoom in, additional neighborhood names will appear on the map.
This was built by SAVI, a program of the Polis Center at IUPUI. If you have any questions, please contact email@example.com.
Geographies shown are census tracts. Demographic data (age groups) are from the 2017 American Community Survey 5-Year Averages (via SAVI). Health data are from the CDC's 500 Cities Project (via SAVI). These indicators are only available for Fishers, Carmel, and Indianapolis. This data does not include other counties or the excluded cities in Marion County, like Speedway and Lawrence.
We have limited information about the risk factors related to COVID-19, how they relate, and their relative importance. This information changes quickly, but we will work to update this map as new information becomes available.
The combined risk index was created by standardizing each indicator by calculating z-scores for each tract (in other words, how far above or below average is each tract for each indicators). Principal component analysis was used to develop a measure highlighting tracts with many overlapping risk factors. This revealed two components that had the strongest explanatory power. Component one, what we call socioeconomic-related risk factors, explained 50% of the variation in these eight indicators. Component two, what we call age-related risk factors, explained 33% of variation.