Investigations

How the coronavirus exposed health disparities in communities of color

The signs were there all along.

As states across the United States began publishing preliminary coronavirus case and death data, a pattern emerged. In cities like New York, Chicago and the District, coronavirus deaths were disproportionately affecting black and brown communities. Additional data suggested that certain chronic health conditions, such as hypertension, diabetes and others, were associated with complications from covid-19, the disease caused by the coronavirus. What once was perceived as a disease that affected only the elderly in a few areas was now wreaking havoc in the bodies of 20- and 30-year-old people of color across the nation.

Percentage of nonwhite population compared to deaths

Percentage

of non-white

population

Percentage

of non-white

deaths

83%

82

68%

67

New York

Chicago

Major city

Most densely populated

major city

89

72

64

63

D.C.

Shreveport, La.

Medium city

Small city

Data as of May 18

Percentage

of non-white

population

Percentage

of non-white

deaths

82

83%

68%

67

New York

Chicago

Major city

Most densely populated

major city

89

72

64

63

D.C.

Shreveport, La.

Medium city

Small city

Data as of May 18

Percentage

of non-white

population

Percentage

of non-white

deaths

83%

82

68%

67

New York

Chicago

Most densely populated

major city

Major city

89

72

64

63

Shreveport, La.

D.C.

Medium city

Small city

Data as of May 18

Percentage

of non-white

population

Percentage

of non-white

deaths

89

82

83%

72

68%

67

64

63

New York

Chicago

D.C.

Shreveport, La.

Most densely populated

major city

Major city

Medium city

Small city

Data as of May 18

Percentage

of non-white

population

Percentage

of non-white

deaths

89

83%

82

72

68%

67

64

63

New York

Chicago

D.C.

Shreveport, La.

Most densely populated

major city

Major city

Medium city

Small city

Data as of May 18

To assess which communities could be more vulnerable to covid-19, The Washington Post analyzed chronic health and social vulnerability estimates. The analysis looked at census tracts with above-average rates of six chronic health indicators: heart disease, diabetes, high cholesterol, high blood pressure, chronic obstructive pulmonary disease (COPD) and obesity.

The analysis uses data provided by PolicyMap, a data analytics company, who used statistical methods to estimate health risk across the nation, and data from the CDC’s social vulnerability index. Social vulnerability is defined as a community with a risk of being severely affected by a hazardous event like a natural disaster or disease outbreak.

National average of chronic disease indicators

Heart disease: 4% Diabetes: 10.8%
High cholesterol: 33.4% High blood pressure: 32.4%
COPD: 6.6% Obesity: 30.1%
Source: The Behavioral Risk Factor Surveillance System, 2017 estimates

Examining underlying conditions is helpful for understanding the health implications an outbreak like the coronavirus, but health conditions tell only part of the story. Risk is increased when factoring in living conditions and access to health care. Communities of color may be more likely to live in densely populated areas in cities because of the history of racial segregation in the United States. And black and Latinx Americans are also two to three times more likely than white Americans to be uninsured, according to a report on covid-19 and race from the CDC.

High-risk health conditions

above national rates

More risk

Less risk

High-risk health conditions

above national rates

Less risk

More risk

High-risk health conditions

above national rates

Less risk

More risk

High-risk health conditions

above national rates

Less risk

More risk

High-risk health conditions

above national rates

Less risk

More risk

In the early days of the pandemic, few states published detailed demographic data on who was being affected by the virus. Some of this was because of the scale of the virus and the unexpected nature of its transmissions. Another hurdle was a standardized CDC form that takes an estimated 30 minutes to fill out, which may have put a strain on local agencies, CDC spokesman Scott Pauley told The Post.

Without key demographic data, it could be difficult for local and state governments to marshal necessary resources for the hardest-hit communities. Furthermore, without this data, these majority-minority communities — which often employ essential workers in places like hospitals, grocery stores and shipping facilities — will probably see more hardship as the virus spreads among vulnerable population

“This basic demographic data is in everyone’s health records. It’s routinely collected all the time,” said Nancy Krieger, a professor of social epidemiology at Harvard University’s T.H. Chan School of Public Health.

“It truly makes no sense in anyone’s pandemic planning or any just regular health record that these data would not be there,” Krieger said.

She also noted that the lack of an online form that is easy to fill out increases the burden on the CDC because all forms need to be digitized afterward, prolonging up-to-date information on the demographics hardest hit by the pandemic.

Measuring the virus’s impact

As of May 22, there are over 7,700 reported cases of the coronavirus in the District, claiming at least 412 lives according a Post tracker of deaths and cases. The non-white population, who represent 64 percent of the city population, based on recent census estimates, account for 89 percent of all virus-related deaths.

High-risk health conditions

above national rates

More risk

Less risk

Majority-minority

neighborhoods

Over 60% of the population

in these tracts are minority

Fort Lincoln

Anacostia

2 miles

Other

Majority-white

neighborhoods

Over 60% of the population

in these tracts is white

Cleveland

Park

Georgetown

Capitol

Hill

High-risk health conditions

above national rates

More risk

Less risk

Majority-minority

neighborhoods

Over 60% of the population

in these tracts are minority

Fort Lincoln

Anacostia

2 miles

Other

Majority-white

neighborhoods

Over 60% of the population

in these tracts is white

Cleveland

Park

Georgetown

Capitol

Hill

High-risk health conditions

above national rates

Less risk

More risk

Majority-minority

neighborhoods

Majority-white

neighborhoods

Over 60% of the population

in these tracts are minority

Over 60% of the population

in these tracts is white

Cleveland

Park

Fort Lincoln

Georgetown

Capitol

Hill

Anacostia

2 miles

Other

High-risk health conditions

above national rates

Less risk

More risk

Majority-minority

neighborhoods

Other

Majority-white

neighborhoods

Over 60% of the population

in these tracts are minority

Over 60% of the population

in these tracts is white

Cleveland

Park

Fort Lincoln

Georgetown

Capitol

Hill

Anacostia

2 miles

High-risk health conditions

above national rates

Less risk

More risk

Majority-minority neighborhoods

Other

Majority-white neighborhoods

Over 60% of the population in

these tracts are minority

Over 60% of the population in

these tracts is white

Cleveland

Park

Fort Lincoln

Georgetown

Capitol

Hill

Anacostia

2 miles

A majority of confirmed cases in the District are in some of the city’s densest neighborhoods, which have large majority-minority populations as well as high rates of chronic health conditions. Wards 5, 7 and 8, which flank the eastern edge of the city, are majority-black areas and have above national rates of chronic health conditions such as diabetes, obesity and heart disease. These are also the wards with some of the highest-known case counts of the virus, according to city health data.

The effect of the coronavirus extends beyond dense urban areas like the District and into smaller cities like Shreveport, La., located in northwestern Louisiana near the Texas and Arkansas borders. In Caddo Parish, where Shreveport is located, black Americans are dying at 2.4 times the rate of whites, according to a Post analysis of state health data.

As of May 22, the state reported at least 36,500 cases and at least 2,506 deaths, according to The Post’s national coronavirus database.

The Louisiana Health Department estimates that black Americans account for 55 percent of all deaths in the state, according to the most recent figures. Almost 60 percent of residents who died had hypertension and 37 percent had diabetes.

High-risk health conditions

above national rates

Less risk

More risk

Majority-minority

neighborhoods

Over 60% of the population

in these tracts are minority

Martin Luther

King Jr. area

Mooretown

3 miles

Other

Majority-white

neighborhoods

Over 60% of the population

in these tracts is white

Hyde Park

Brookwood

High-risk health conditions

above national rates

Less risk

More risk

Majority-minority

neighborhoods

Over 60% of the population

in these tracts are minority

Martin Luther

King Jr. area

Mooretown

3 miles

Other

Majority-white

neighborhoods

Over 60% of the population

in these tracts is white

Hyde Park

Brookwood

High-risk health conditions

above national rates

Less risk

More risk

Majority-minority

neighborhoods

Majority-white

neighborhoods

Over 60% of the population

in these tracts are minority

Over 60% of the population

in these tracts is white

Martin Luther

King Jr. area

Mooretown

3 miles

Hyde Park

Brookwood

Other

High-risk health conditions

above national rates

Less risk

More risk

Majority-minority

neighborhoods

Other

Majority-white

neighborhoods

Over 60% of the population

in these tracts are minority

Over 60% of the population

in these tracts is white

Martin Luther

King Jr. area

Mooretown

Hyde Park

Brookwood

3 miles

High-risk health conditions

above national rates

Less risk

More risk

Majority-minority neighborhoods

Other

Majority-white neighborhoods

Over 60% of the population in

these tracts are minority

Over 60% of the population in

these tracts is white

Martin Luther

King Jr. area

Mooretown

Hyde Park

Brookwood

3 miles

In recent months, states like Louisiana have gotten better at reporting the racial and ethnic demographics of covid-19 victims, and more recently, the National Center for Health Statistics has started to release more detailed provisional state-level estimates. But even the agency acknowledges in its release notes that much of the data is delayed because of the many steps are required to collect data from coroners offices across the nation.

The pandemic has laid bare the staggering costs of decades of disinvestment and inadequate care of black and brown communities across the country and the desperate need to collect and measure the impact of the disease as it pertains to race.

Explore chronic health rates in your community

Select an area or condition to highlight:

Areas with a minority population greater than 50 percent. Compared with whites, some minority groups have higher rates of chronic conditions.

High-risk health conditions above national rates
  • Less risk
  • More risk
Use two fingers to pan around the map.
Pinch to zoom.

Tap a Census tract to see details

Methodology

To understand the link between chronic health outcomes and race in communities of color, The Washington Post analyzed tract-level chronic health estimates provided by PolicyMap.

PolicyMap's model relies on chronic health data from the CDC’s Behavioral Risk Factor Surveillance System (BRFSS) and estimates prevalence in census tracts using a statistical model.

The Post created an index of communities most at risk by analyzing which census tracts had above national rates of six chronic conditions and then scored each tract based on its distance from the mean. The following six conditions were chosen based on CDC guidelines of populations most at risk to have severe complications from the virus: heart disease, diabetes, high cholesterol, high blood pressure, chronic obstructive pulmonary disease and obesity.

Social vulnerability, crowded housing and uninsured population data comes from the Agency for Toxic Substances and Disease Registry's Social Vulnerability Index. Vulnerability rank is based on several factors including socioeconomic status, household composition, disability, race and ethnicity, language, housing type and transportation access. High vulnerability is considered anything in 90th percentile of all tracts based on these metrics.

City death rates were pulled from the health departments of New York, Chicago, D.C. and the state of Louisiana.

Race and ethnicity population data comes from the 2014-2018 American Community Survey estimates from data.census.gov.

Maps use census tract geometry from IPUMS NHGIS at the University of Minnesota, standardized to the 2010 census boundaries. Census tracts for the Shreveport map are from the 500 Cities project.

Originally published May 22, 2020. Story was updated to clarify how PolicyMap developed its estimates.

Story and graphics editing by Armand Emamdjomeh.

Aaron Williams

Aaron Williams is an investigative data reporter who specializes in data analysis and visualization for The Washington Post. Before joining the investigative team he was a reporter for the Post graphics desk. He previously covered housing, campaign finance and police for the San Francisco Chronicle and the Center for Investigative Reporting.

Adrian Blanco

Adrián Blanco Ramos is a graphic reporter in the graphics department at The Washington Post. He previously worked at Spanish newspaper El Confidencial focusing on data visualization, data analysis and investigative journalism. He participated in the International Consortium of Investigative Journalist’s Paradise Papers investigation.

Share