by Jay Colbert
Many sources, both nationally and locally, have noted the disparities that exist in who is contracting COVID-19. All of the charts below use only three numbers: total count of COVID cases; percent of COVID positive tests by race; population counts by race. While these charts are based on the same data, they reveal different aspects of the problem. We examine three different ways to highlight these disparities.
A range plot compares distribution of cases to the distribution of population
The first chart is called a range plot, which is a relatively new way of visualizing data. Orange dots represent percent of population for each race. You can see from the chart that there are about twice as many white residents in Marion County as black residents. The blue dots show what percent of all COVID-19 positive tests are associated with each race: about 39% for blacks and 42% for whites.
The distance between the blue and orange dot shows how over- or under-represented a population is in terms of positive test scores. For example, while black residents make up 28% of Marion County’s population, they make up 39% of cases where race is known. This tells us that blacks are over-represented in positive test results.
Percent of positive tests differs widely from population distribution
Bar charts are a simple way to view the same comparison
The next chart is a more traditional bar chart. It shows the exact numbers as the range plot just as simple bars. We see the same message, that blacks represent about half the total population of whites but have a relatively outsized share of positive tests.
Black patients more likely to test positive than other groups
Rates allow for a more direct comparison
Our final chart is another simple bar chart, but with the arithmetic done a little differently. Instead of showing percent of population and a percent of tests by race, we divide the number of positive tests by population to get a rate of positive test within each race group. We can see from this that blacks are about twice as likely to have tested positive than whites in Marion County.
Blacks 2.4 Times more Likely than Whites to Test Positive in Indiana
In Indiana overall, blacks are 2.4 times as likely to test positive than whites. In both Marion County and Indiana, Hispanics and Latinos are only slightly more likely than whites to have tested positive and females are more likely to have tested positive than males.
In Marion County, 482 out of every 100,000 black residents tested positive for COVID-19 by April 21. That rate is 230 for white residents. Men of any race test positive at a rate of 395 per 100,000, slightly less than the rate of 477 for women.
But there are big limitations in this data. It’s a bit of a stretch to apply the rates from just those tested to the population in general. The tests are not random samples from the entire population, not even a random sample of those with symptoms. There are certain criteria for people to even be tested, such as severe symptoms, other health conditions, and working in a front-line occupation. Those criteria could skew the numbers in one direction or another.
However, we can accurately say that a certain percent of each race has tested positive for the virus, but from there we need to exercise caution on what conclusions we draw and what actions we take based on the numbers.
In the coming weeks, we will be examining these differences deeper. What underlying factors could cause black individuals to test positive at more than twice the rate of whites? Are there factors that make blacks more likely to contract COVID-19 or develop more severe cases? And how does this combined public health and economic crisis exacerbate existing inequalities? Look for this story in our Coronavirus Data Hub.
Jay Colbert is data manager at The Polis Center at IUPUI. This article was originally published at The SAVI Community Information System website.
