Race Still Unknown in One of Five COVID Cases

by Carol J. Bova

On July 27th, Bacon’s Rebellion asked the question, “Why is VDH Stockpiling Cases as Unknown Race”? The Northam administration had expressed concerns since March about the disparity of racial impacts from COVID-19. Yet 24% of confirmed cases at that time still had not been classified by race or ethnicity.

More than a month later, that percentage has barely budged. Between Aug. 1 and Sept. 5, 19% of COVID-19 cases had no racial or ethnic identifier.

On August 26, the Virginia Department of Health (VDH) discussed on its COVID-19 blog why the missing information is important and announced a new method to address the problem.

Good information on disparities in disease incidence, outcomes, and social and economic consequences, is necessary to guide and develop an appropriate response. However, efforts to study these disparities have been hampered by missing data. Almost a quarter of confirmed cases are missing race and ethnicity data. Accounting for this missing data is essential to understanding COVID-19 and to facilitate research into health disparities. Social Epidemiologists from the Office of Health Equity used imputation techniques to estimate race and ethnicity for cases missing that data.

The blog post described the process used to estimate the racial composition of COVID-19 cases that were not originally reported and showed the results. Although VDH will continue to use unimputed data on the Daily Dashboard, staff from VDH Surveillance and Investigations will post new results of the imputed numbers for research purposes as they are calculated.

But imputed data is not as good as real data.

Since August 1, labs were required to include the patient age, race, ethnicity, sex, residence zip code and county with other required info on test results, so why hasn’t the VDH Daily Dashboard improved on the number of cases with unreported race?

Who is responsible for the lack of information – are the labs failing to report as required or is VDH failing to input the data?

Share this article


(comments below)


(comments below)


26 responses to “Race Still Unknown in One of Five COVID Cases”

  1. That Latino rate is so sad given they don’t make up near that percentage of our state’s population. But as a percentage of “essential workers”, especially in meat processing plants? Yeah, exactly.

    1. Fortunately, the number of Latino deaths from COVID-19 is not in the same proportion as cases. It is 10% and in line with population numbers, although more Latino deaths are in the ages of 34-65 than other racial/ethnic groups. The White deaths are at 54%, Black at 26% and Asians at 5%.

      1. I’d hazard to guess that as a demographic on the whole, Virginian Latinos are on the younger side which would help lower the death rate vs other groups.

        1. The death statistics are identified by the same groups as the cases and by age ranges. See my post here on BR July 11, “The Shocking Number of Hispanic COVID-19 Deaths for Ages 35 to 64” that explores the issue.

    2. idiocracy Avatar

      It wouldn’t have anything to do with the propensity for Latinos to live in overcrowded conditions, would it?

      1. That certainly may be a factor, but other groups have members living in overcrowded conditions also.
        If researchers had more complete and accurate information, they could compare the factors across racial groups to come up with theories and approaches to test to reduce the impact.

  2. Reed Fawell 3rd Avatar
    Reed Fawell 3rd

    “Since August 1, labs were required to include the patient age, race, ethnicity, sex, residence zip code”

    Lets drop the ‘race’ and ‘ethnicity.’ Why not? We are all humans, after all.

    1. Not all humans are alike. Ethnic/cultural differences affect lifestyle choices and consequently health. Diet choices that result in obesity and diabetes are an example. I don’t agree with the practice of substituting ethnicity for race in health matters for Latinos. The racial component does matter for developing treatments. Some drugs affect people of different races in different ways, just as some drugs act differently between men and women. That’s why research studies and CDC use Hispanic/Black, Hispanic White, Non-Hispanic Black, Non-Hispanic White. I don’t know if grouping all Asian and Pacific Islanders in one package is valid or not, but it’s done.

  3. Steve Haner Avatar
    Steve Haner

    The CASE data often lacks “racial” classification. But the hospitalization reports and death certificates are far more complete. Those are the things that matter more. Nobody has any idea, none whatsoever, how many “cases” there really are, because so many people who get this show no symptoms or minor symptoms and never get tested. Oh, and wait, most of the people who have died are NOT in those minority categories? My my, don’t get that impression from the news, do you…..

    1. We’re getting distorted case info because of political correctness in trying to separate Hispanic ethnicity as a race.

  4. LarrytheG Avatar

    I notice on the VDOE School Quality Profiles for enrollment, they use the following categories:

    black, hispanic, white, asian, multiple races ,american indian, native hawaiian

    here’s an example:


    also – I’ve seen multiple categories for Hispanic…

    next post is from PEW:

    1. Steve Haner Avatar
      Steve Haner

      Wow, we’ve evolved a race since 1492, when the first Spaniards reached the New World and started breeding with the locals. See? It’s not biology, it is culture……

    2. Larry, because a group of people have a mistaken idea confusing race and ethnicity or culture doesn’t make it a fact.

      1. Steve Haner Avatar
        Steve Haner

        Now, as we all have to recognize, people can just define themselves now. No actual basis in logic or science required. Don’t get me started on gender, for example……

        1. I have a quixotic preference for biological fact when it comes to science and medical research.

          1. Reed Fawell 3rd Avatar
            Reed Fawell 3rd

            Where does a black person begin and end? Where does a white person begin and end?
            Half white? Half Black? And what other combination thereof? Throw in a little Asian, bit of Nordic, a tint of Italian, Serb, and Moroccan, Mexican, too, including in the latter, God knows what? At what point does all the mixed up nonsense end?

          2. Reed,
            That’s where the categories of Two or more races and Other Race come in. When genetics are blended, research from the other categories may or may not apply.

          3. Reed Fawell 3rd Avatar
            Reed Fawell 3rd

            But Carol, is that fact or fiction, those claims of whose white and/or black? Or is it mostly illusion? Surely we are rapidly shading into homogeneity, which in fact is quite the reverse of racial distinctions, since most of those claims now are based on skin color, on myths, on bias, and preferences, often driven now by chronic obsessions.

        2. I self-identify as a fighter pilot, but those bigots at Oceana Naval Air Station still won’t let me fly one of their F-18s.

          1. Reed,
            It’s a growing challenge for researchers trying to match study and control groups. I don’t know if there are any studies on proportion of blended heritage. Sometimes differential drug impacts are only discovered when people react badly and a recognizable pattern develops.

            COVID does not have a universal course from what’s been documented so far. There are definite differences in who is more susceptible to the worst complications. There’ll be a post soon discussing some of this.

      2. LarrytheG Avatar

        Oh I agree. I’m really asking how such determinations are made. Are they self-reported or is someone who is taking the data making that call?

        I know in some situations – it’s provided by the applicant – not determined by the person processing the applicant.

        How does this work at VDH?

        1. Either doctor’s order or lab asking or both.

  5. I am very interested any official analysis of the hispanic situation…certainly seems like the NoVA COVID experience revolves around the hispanic community. Probably as essential workers and then I guess it is obvious that, for all demographics, household spread accounts for many cases.

    A few weeks ago they said Wash DC was trying to round up 500 past COVID victims volunteers to interview and interpret how the spread was happening, but only maybe 80 had volunteered at that time of the report.

    1. LarrytheG Avatar

      Demographics are taken from clients at various things like volunteer taxes and signing up for health insurance and the applicant is allowed to self-identify and one of the choices on some of the forms is “prefer not to answer”.

      I don’t know in the VDH case. Do we know if people self-identify and have the option of not saying?

Leave a Reply