Genesee County, MI—In many ways, the COVID-19 pandemic has brought racial and social injustice to the forefront of public health. Because while health officials have long known that people of color are at a greater risk for poor health outcomes, the pandemic has forced swift, real-time action to prevent minority groups from disproportionately contracting and dying of COVID.
In Flint, where over half the population is Black, local health leaders have focused their efforts on eliminating these health disparities—and it seems to be working thus far. In Feb. 2021, health officials reported that Black people in Flint were underrepresented in deaths and confirmed cases, which had not been accurate at the beginning of the pandemic.
However, over the past year, a substantial amount of positive test results in Genesee County are missing race information–data that is crucial to identifying and mitigating racial disparities. With missing data, it’s more difficult for health officials to draw conclusions on how COVID affects groups of people, specifically minorities.
“We need to know how this virus is affecting different communities, and race data is a huge part of that, to know what specific communities it’s affecting, and what resources those communities need now and also moving forward,” Kayleigh Blaney, deputy health officer for the Genesee County Health Department, said.
Between March 2020 and October 2021, 38% of all confirmed cases in Genesee County are missing race data. In Flint, 31% of all cases in the same period are of an unknown race, meaning health officials have to fill in the gaps based on their best assumptions and known race data.
It’s a source of frustration for health leaders like Debra Furr-Holden, director of the Flint Center for Health Equity Solutions, an extension program of Michigan State University. She and her team have led research efforts on health disparities and COVID in the Flint area.
“The main thing I’m pushing for is the data. Because that’s what we need, that’s really where it all sort of starts and ends,” Furr-Holden said.
Furr-Holden is a member of Governor Gretchen Whitmer’s Racial Disparities Task Force and has helped implement several initiatives to address COVID disparities in Flint and Michigan.
These initiatives included launching a barrier-free COVID testing program to address the financial burdens and hardships that bar low-income individuals from accessing or seeking tests if they have symptoms.
But given the amount of missing race data, are these efforts in Flint genuinely effective? To what extent does unknown race data affect health leaders’ abilities to mitigate racial disparities and COVID?
Why Getting COVID Data isn’t Easy
In April 2021, Flint Beat partnered with JSK-Big Local News to answer a question: how does COVID-19 impact those with preexisting health conditions in Flint? Given issues like Legionnaires Disease and lead poisoning from the Flint water crisis, we wanted to know if COVID had a more severe health impact in Flint. After months of analyzing data, we ended up with more questions than answers, mainly because the data we wanted was either hard to find or didn’t exist.
We reached out to the FCHES in May 2021, asking for a breakdown of individual COVID cases by race, zip code, income, and preexisting conditions for Flint and Genesee County. We also requested data on deaths due to COVID with the same demographic information.
However, FCHES stated they had no data on preexisting health conditions. So, we decided to pivot once we could review the data that FCHES did have.
FCHES sent us data in June, but it did not include demographic information by case nor whether the patient had lived or died. Our colleagues at Big Local News scraped the back end of the Genesee County Health Disparities Dashboard, compiled and managed by the FCHES.
That gave us a spreadsheet with testing results and demographic information on a case-by-case basis without identifying information. It included whether the patient had died, their age, gender, city, zip code, race, and ethnicity.
However, some columns needed further elaboration, and after Flint Beat contacted the FCHES in July for an explanation, the disparities dashboard was removed for several weeks. FCHES stated that this was to ensure they did not violate their data agreement with the health department.
Once the dashboard reemerged, Big Local News scraped it twice again in August and October. The information was less detailed but did include race and municipality. Though we would have liked to have analyzed the first scraped dataset, it was by this time outdated. So we opted to use the most current data and focus only on race and city. It was then we noted the size of the unknown race category.
It didn’t appear that Flint Beat could cross-reference COVID data with preexisting conditions. But, in the process, we discovered another interesting question: Where was the race data?
How Does Race Data Go Missing?
The data collection process works like this. Patients fill out a form during testing which asks questions on race, age, sex, and ethnicity. Hospitals, labs, physicians, and schools are then required by law to report the patient’s information and their results to local health departments.
The data is submitted via a system called the Michigan Disease Surveillance System. If demographic information is missing, it is the job of the local health department to backfill the gaps by calling the patient to collect the data.
Chief Epidemiologist for the Genesee County Health Department Danielle Lederer, who works closely with Furr-Holden and her team, said there are many reasons why race data goes missing. However, most unknowns come from errors or oversight in labs, she said.
Furr-Holden agreed, adding that the issue is not likely caused by patients who do not want to provide their information.
“We actually think a big part of it is, it’s getting collected, and it’s not getting entered. And sometimes, when it is getting entered, when it then comes back with the result, it’s not being included or being uploaded as one of the variables that comes in with the data that then comes to the Health Department. There’s just so many leaks in the pipeline that account for this,” Furr-Holden said.
Missing race data was less of a problem during the initial stages of the pandemic, Furr-Holden said, but the issue has grown.
Below are pie graphs showing the progression of race data in Genesee County in 2020. Unknowns grow substantially between March and December.
Missing race data also increases during surges of COVID cases, as it depends on the GCHD’s capacity to call patients and backfill the information.
“When we’re getting 200 cases a day, and we only have so much staff, that gets a little difficult. But we work really, really hard to collect as much information as we can so that we can work with our partners in giving them data and making sure we keep our disparities and data dashboards as full as possible,” Lederer said.
Lederer also said when race data is missing, the health department relies on residents to answer their phones.
“During the COVID-19 pandemic, there has been COVID-19 fatigue. So, we oftentimes have people who don’t answer phone calls because they don’t want to speak with the health department or because of unknown numbers,” Lederer said.
Is Missing Race Data Systemic or Racialized?
Another way to phrase this question is whether there are more missing race cases in Flint compared to other cities in Genesee County.
While Flint has the most unknowns, 2,435 cases out of a total of 7,701, but it’s also the most populated. Per capita, Flint has the lowest rate of unknown cases, 25 per 1,000 residents. Atlas Township takes the lead at 63 unknown cases per 1,000 residents.
Narrowing this down to the ten most populated municipalities in Genesee County, Flint has the lowest rate of unknown race per capita while Fenton Township has the highest, 57 cases per 1,000 residents.
Rick Sadler, a geographer and assistant professor at MSU who works with Furr-Holden at FCHES, has been mapping missing race cases by census tract. Sadler said their research shows that unknowns are not specific to one entity or region.
“What we’ve done is examine whether there’s a difference in the number of missing cases in neighborhoods that are predominantly white or predominantly black. Every time we’ve done it, what we found is that there doesn’t seem to be a systematic error in the unknown race cases,” Sadler said. “There’s a fair amount of missing race data…. So, the odds that this is a systematic race issue are very low.”
Flint by the Numbers: What the ‘knowns’ tell us about ‘unknown’ racial disparities
Looking at the total number of COVID cases (including unknowns) by municipality, the City of Flint has the most by far, a total of 7,701 cases from March 2020 to October 2021.
But, again, Flint also has the largest population of all municipalities in Genesee County. Calculating cases per capita, Flint has the fourth-lowest case rate among Genesee County’s 28 municipalities at 80 cases per 1,000 residents. That means COVID has impacted Flint less than most of Genesee County over the past year. Atlas Township has the highest per-capita rate, 145 cases per 1,000 residents.
Among the ten most populated municipalities in Genesee County, Flint has the lowest rate per capita, while Fenton Township has the highest at 105 cases per 1,000 residents.
Lower case rates in Flint compared to the rest of Genesee County with majority white residents reflects a trend that researchers at the FCHES have been tracking over the past year: while at the start of the pandemic, communities of color were more at risk for COVID, the politicization of the virus, or rather how health-protective behaviors like mask-wearing are often associated with an individual’s political orientation, has caused whiter, suburban communities, like Fenton Township, to contract COVID at higher rates, Sadler said.
In March 2020, Black people represented 58% of cases despite being only 20% of the county’s population. The majority of those Black residents live in Flint.
By October 2021, Black residents represented only 10% of cases in Genesee County.
“I think one thing we’ve learned in the last year is that places that don’t believe in science, for example, like redder counties, have wound up with much higher case rates. And there’s some evening out of that pattern in death rates, too. The big inequality, it was at the beginning. And that was especially because more vulnerable people, older people, people with chronic conditions were contracting COVID,” Sadler said.
Furr-Holden said that because black people are now underrepresented in COVID cases, health officials are confident that racial disparities in Flint have not reemerged.
The cluster graph also confirms that unknowns are not specific to census tracts with higher African American populations.
“The assumption is that the distribution of the unknowns is the same as the distribution of the knowns. So, if half of the knowns are black, you would assume half of the unknowns are black,” Furr-Holden said.
The opposite is also true. If black people were overrepresented in known COVID cases, health officials would assume they were also overrepresented in unknown cases, Furr-Holden said, adding that because white people are overrepresented in known cases, it is likely that these communities have more missing race data.
Filling in the Gaps
Despite local health leaders’ best assumptions, unknowns remain, well, unknown. If the problem continues to grow, researchers will not address racial disparities.
Furr-Holden said it is a matter of holding testing providers accountable.
“The unknown race data problem, that’s not a fish that we can fry. The reality of it is that there are laws in place. And those laws are not being enforced,” Furr-Holden said. “Testing providers are required to collect it. The labs should be required to enter it, and it needs to be reinforced.”
Not entering data should come with repercussions, she said, such as not being reimbursed for the expenses of analyzing tests.
“I just believe that. So, we’ve done some advocacy work and tried to elevate those policy interventions. But the actual solution, in my opinion, resides in the hands of our legislative and executive branches, and they’re the ones who have shot the responsibility,” Furr-Holden said.
Meanwhile, Furr-Holden and her team at MSU will continue their data work, but residents have a part to play if they want accurate COVID data for their community.
“We just really appreciate when people are willing to speak with us about our investigations when they’re positive,” Lederer said. “Please expect an 810 number or even an unknown number….We just really need people to pick up unknown numbers and be willing to work with our COVID-19 investigations.”
(This story was in partnership with JSK-Big Local News. Flint Beat was awarded support through the JSK-Big Local News Data Reporting Grant in 2020. Our initial proposal was to analyze COVID-19’s impact on pre-existing health conditions caused by the Flint water crisis and other health issues in underserved communities. So far, the data on pre-existing health conditions caused by the city’s water crisis and other health issues in Flint was not readily available.)