As COVID-19 continues to spread across the country, state and local health officials rush to try to detect and contain outbreaks before they get out of control. A key to that is testing, and despite a slow start, testing has increased around the country.
But it’s still not always easy to get a test. While many things can affect access to testing, location is an important starting point.
NPR investigated the location of public testing sites in Texas, one of the first states to reopen, to see how they were distributed between predominantly white and predominantly minority areas. The investigation found that in four out of six of the largest cities in Texas, testing sites are disproportionately located in whiter neighborhoods.
With evidence growing that black and Latino communities are harder hit by this deadly disease, community leaders say that testing disparities are problematic. Many experts warn that if communities don’t test the most vulnerable, they could miss pockets of infection and have new large outbreaks.
“If you’re casting a very small net, and you’re shining a flashlight on a small portion of infections that are out there, you might think you’re doing OK,” says Dr. Jennifer Nuzzo, lead epidemiologist for the Johns Hopkins COVID Testing Insights Initiative. “Whereas there’s this whole pool of infections that you haven’t seen.”
Nationally, it’s hard to determine where there may be testing disparities because data is scarce. Most states and cities across the country either do not track or do not report the racial breakdown of tests that are conducted. But there are media reports of racial disparities that suggest that the patterns identified in Texas are happening in other parts of the country.
“I was acutely concerned from the very beginning,” says Dr. Wayne Frederick, president of Howard University, a historically black college in Washington, D.C., who is leading efforts by the university to bring more testing to black neighborhoods. The lack of attention paid to minority health care, he says, “is not a good strategy for our health care system and for the vulnerable part of our community.”
Texas: Whiter neighborhoods have more testing sites
In Dallas, a stark divide exists between the north and south ends of town. Interstate 30 bisects the city and largely serves as a borderline between the city’s predominantly white neighborhoods to the north and the predominantly black or Hispanic neighborhoods in the south.
In North Dallas there are 20 testing sites. Southern Dallas has nine, with a third of those sitting within a mile of the interstate.
Dallas County, home to the city of Dallas, has the second-highest COVID-19 case count in the state.
In three other cities, the trend was the same: More testing sites were in whiter neighborhoods. In each city, NPR ranked neighborhoods based on their density of white residents. The analysis then focused on the half with the lowest density of white residents to see how many testing sites that are open to the public are located in those areas:
- In Fort Worth, six out of 18 testing sites were in the neighborhoods with the lowest density of white residents.
- In Austin, nine out of 25 were in these neighborhoods.
- In El Paso, five out of 18 were in these neighborhoods.
To determine where permanent testing sites were located, NPR contacted health care providers in those six largest cities in Texas — representing a total population of 7.9 million by most recent Census estimates — and reviewed government, health care and news websites.
The testing site locations include facilities where residents can go to get diagnostic testing for active infection, such as urgent care clinics, hospitals and drive-through testing sites. They don’t include sites such as doctor’s offices and some hospitals that may provide tests for admitted patients but are not available to the public. Mobile sites, where locations regularly change, are also not included. Reporters analyzed the racial breakdown in areas where testing sites are located using census tract data.
NPR offered city and county health officials the opportunity to review the findings and point to additional testing sites.
These disparities could point to a larger problem in Texas, says Nuzzo. Without good testing access in places where minorities live, the state could be missing cases. She notes that Texas has a low rate of positive cases — a measure that can indicate a community is conducting sufficient testing — but it also has one of the lowest rates of testing among states.
“It’s either that there’s just not that much infection in Texas, or that there is infection, but they’re testing the wrong people,” says Nuzzo. “And it’s been hard for me to believe that there’s not that much infection in Texas.”
The Texas Division of Emergency Management and the Department of State Health Services did not make anyone available for an interview on the state’s testing plan.
The national picture
Some communities across the country are starting to identify and tackle the issue, but only four states currently publish racial breakdowns of who is being tested: Nevada, Delaware, Kansas and Illinois. In New York City, authorities have identified hotspots in low-income and minority neighborhoods and have set up testing sites at over 70 faith-based organizations to target those communities. And in Wisconsin, where some of the first reports of stark racial disparities emerged, free testing is now available to all black, Hispanic, and Native American residents.
At the federal level, the House of Representatives passed a wide-ranging coronavirus relief bill this month that would, in part, fund testing centers in minority communities. Continuing negotiations on the bill, however, make it unlikely that it will pass the Republican-controlled Senate in its current form. Democratic Rep. Bobby Rush of Illinois has introduced a separate bill that would fund testing sites, prioritizing those around coronavirus hot spots and in underserved communities.
“We don’t know the numbers, we don’t know the number of people who have not been diagnosed,” says Rush, who represents a mostly black district in Chicago, recalling his early response to Chicago’s outbreak. “We were operating blindly regarding the individual community response to COVID-19, and the white community was better informed.”
The House Energy and Commerce committee also announced Tuesday it would hold a hearing on states’ testing plans, including access to testing for minority communities.
But local communities might have trouble piecing together what testing access looks like. Local public health departments have dealt with years of underfunding, and the federal government isn’t keeping track of and publishing its own list of testing sites. States may not have complete data either. Texas has its own state map, but some sites located by NPR were missing or out of date.
“For the most part across the U.S., it’s really, really hard to rely on those government websites for the information,” says Tarryn Marcus, who leads a volunteer effort called Get Tested COVID-19, which is attempting to compile a national database of testing locations. “There’s a lot of inconsistency, a lot of outdated information — stuff that was true maybe a month ago or a few weeks ago but it’s no longer true.”
Marcus’ organization is attempting to fill in the gaps with their database, which was compiled over two months by more than 80 volunteers. Though it’s still not complete, their data reinforce what NPR found in Texas cities. Marcus’ team analyzed the sites they do have to look at where there were large numbers of people without a nearby testing site. Across the country, most of the areas with gaps, she says, were in blue collar communities with low income.
Cities in Texas vary in their approach
Not all cities in Texas have a disparity in testing site locations. In San Antonio, the state’s second largest city, testing sites are equally distributed by race. Houston, the state’s largest city, has more testing sites in minority neighborhoods than white neighborhoods. The city says it took into account risk factors for disease, as well as CDC testing data, to determine where to target its testing.
But in Dallas, recent research by the UT Health School of Public Health identified parts of South Dallas as vulnerable to COVID-19 hospitalization because of higher concentration of some chronic conditions like asthma, diabetes and obesity. These were the same areas — predominantly minority neighborhoods — that NPR’s analysis show to lack testing sites. The state does not provide detailed data on the races of coronavirus patients, and Dallas County’s is incomplete.
“If we’re looking to protect our most vulnerable in the population, then we need to concentrate some testing resources in the areas where they’re more likely to live,” says Dr. Stephen Linder, one of the UT Health researchers.
Dr. Philip Huang, director of Dallas County Health and Human Services, said many of the testing sites in north Dallas are at private sector providers. That’s where most of Dallas’ hospitals are as well.
Huang said the county has tried to focus public sector testing efforts towards underserved communities as much as possible.
“We have been trying to definitely target those resources that we have control over into those areas,” he said, including two federally-funded drive-thru sites, one in South Dallas and the other downtown. Combined, they can test 1,000 people a day. “A lot of the disparity … is reflecting some of that private sector availability of tests in the northern part of Dallas compared to southern Dallas.”
In an attempt to fill the gaps, in mid-May the city started sending health care workers to residents’ homes if they don’t have transportation to get to a testing site. That effort has averaged 20 tests per day, according to figures provided by the city. Other cities have set up temporary sites that are open for days at a time before moving on to the next location.
Even the public health systems, which exist to fill these gaps, present obstacles for someone trying to get a test. Parkland, the public hospital system in Dallas, does have clinics in South Dallas that can test, but you have to be a patient already or be seen through the ER. You can also be referred to Parkland through one of the area’s federally qualified health centers, but those centers don’t do testing themselves.
For cities that have testing disparities like these, the solution involves not just opening an equal number of testing sites in minority neighborhoods, but actually focusing on those that are most at risk, says Dr. William Owen, a medical school administrator whose work has focused on health care access for racial minorities.
In other words, Owen says, communities should make sure it’s easy to get tested in places carrying the heaviest burden of the disease.
“That’s where I’m going to plot my testing sites,” he says. “Direct your resources specifically while they’re limited — which is what they are right now — to where you think you’ll get the most hits.”
Other barriers to testing
Location, of course, isn’t the only factor that determines whether at-risk people will get tested for the disease. Other barriers can keep someone from being tested, including whether a doctor needs to refer you to a testing center or whether you can just walk in; how much the tests cost if you’re uninsured; and how much time you have to wait at a testing site. Cost could be an issue in a place like Texas, which has not expanded Medicaid and has the highest uninsured rate in the country. Lack of transportation and concerns about being able to work after testing positive can also keep people away from testing sites.
These challenges can be especially difficult for people of color to navigate: They’re more likely to be uninsured and, in some places, more likely to work in essential services, making it difficult or impossible to spend hours traveling or waiting at a testing site.
“This disease has absolutely highlighted many disparities that we have in our country when it comes to public health,” says Angela Clendenin, an instructional assistant professor of epidemiology at the Texas A&M School of Public Health. “And it boils down to access.”
One way to deal with these challenges is to work with community leaders and groups that people are already familiar with to encourage them to get tested, says Owen, the medical school administrator. Leaning on faith-based leaders to stress the importance of testing, along with setting up makeshift testing sites at community centers like churches, could bring large-scale public health efforts to the doors of people who might otherwise be excluded.
A South Dallas church, for example, hosted a two-day mobile testing event last week, sponsored by the state and National Guard.
“How do you do voter registration? You do it at the local high school. You maybe have someone in front of the grocery store in the neighborhood. You have somebody at the church,” Owen says. “If I were the mayor, I’d say, ‘Great, let’s have the arena. But by the way, we’re going to have a small testing center at the big churches.'”
NPR gathered addresses of permanent testing sites from county, city and state websites, health care providers and news reports. NPR verified these sites by contacting health care providers and city officials by phone or email. City and county officials were offered the opportunity to review the findings and point to additional testing sites. Mobile and temporary testing sites are not included in this dataset. Sites that only perform antibody testing are not included.
NPR geocoded testing site locations on an online geocoding platform to determine what census tract they were within. A reporter verified each location with a geocoding “accuracy score” of below 0.9. For each city, the analysis included only census tracts within the city’s official boundaries, which do not include suburbs. The Census Bureau provided median household income and demographic data, which were used along with data from the Social Vulnerability Index (SVI). The main demographic measure referenced in this story was the percent of the population who identify as ‘white alone.’ For income, percent white and SVI, NPR calculated the number of sites and distance from a site for tracts above and below the city’s overall median for that measure. Medians referenced are medians of census tracts, not population medians, and may therefore differ slightly from population medians.
The dataset used in this analysis is available on Google Sheets.