Health disparities throughout the COVID-19 pandemic have been misattributed to different minority groups due to outdated and insufficient data collection on race, ethnicity, and the social determinants of health. Data policies must be updated to support burgeoning relationships with community organizations that serve these groups and improve societal wellbeing.
Drs. Janice Bowie (JB) and Darrell Gaskin (DG) are professors at the Bloomberg School of Public Health with expertise in health disparities and the social drivers of health. Below they discuss the perpetuation of outdated public health data policies toward minority groups that have driven miscommunication, confusion, and even animosity throughout the COVID-19 pandemic.
JB: Our minimum national standards for racial and ethnic categorization, which are set by the U.S. Office of Management and Budget, leave out a wide range of people because the government hasn’t kept up with changing demographics. The availability of the denominators for certain groups wasn't even obtainable in the recent census. We have to figure out how to think about, collect, and disaggregate data on social drivers of health. In order to do that, we have to do a better job of thinking about multi-racial groups, as well as many other racial and ethnic populations. Currently, we don't have that data. Everybody falls into four or five overly broad categories.
DG: Along with recording race, ethnicity, language, and gender, we need to do a better job of collecting data on social risk factors: income, neighborhood, level of educational attainment, health literacy, food insecurity, and housing insecurity, because oftentimes problems that have to do with social risk factors get conflated with race, ethnicity, and culture. If you are housing insecure and you delay seeking healthcare because you’re prioritizing figuring out where you're going to live tomorrow, that has nothing to do with your race and ethnicity.
Sometimes we see behaviors that are more prevalent in one group than another, and then we ascribe that behavior to culture, when it's actually about attending to basic needs. Low-wage earners have had a difficult time during this pandemic. It was likely not particular to Black or Hispanic low-wage earners, but we don't collect data on white low-wage earners. Then, when we report the data just by race, white low-wage earners will respond two ways: (1) that's a Black problem, not my problem, and (2) I have my own problems, and nobody cares.
JB: We collect it, but I don't think we connect it. We don't always want to know that problems affect the normative power groups or privileged populations. It's much easier to assign the blame and perpetuate the myth and misinformation around the minority populations of Black, brown, and Native American people, and those who may have a different sexual orientation or gender identity than the “majority.” Conversations about systemic racism in healthcare are important and have to happen at the levels of data, training, treatment, and policymaking. All of these areas intersect, and we can't fix one without paying attention to the others.
DG: I would first try to make sure that data can be captured and analyzed by our electronic medical records (EMRs), even if we just made sure that we collected the relevant detailed race and ethnicity data. One size won’t fit all. The same demographic categories won’t make sense in California and Vermont for instance.
The same should be done with social risk factors. Questions about social risk factors should be embedded in our EMRs, and people should be asked those questions. Otherwise, we presume that someone had access to healthcare, had a choice to do something, and did not do it. Then we ascribe that to some part of their identity. The National Vital Statistics System has begun to incorporate educational attainment in the death record, so that we get some idea about social position, not just race and ethnicity.
JB: We also have to acknowledge that we have poor data quality. We need to figure out why it's poor and how we can improve it. I would want to know, if you're collecting it, what are you collecting it for, how will you be using it, how will you know if there's been an impact, and who does it impact. We need to think about what data is relevant for what it is that we are trying to either understand, change, or advocate for — some policy or initiative. We talk about data in the same way we talk about vaccine hesitancy — in a broad way that means many different things. We have to think about integrity, quality, usage, application, and disaggregation of data by racial and ethnic groups and by social drivers.
DG: Data is often proprietary. Sometimes the people who need access to the data in order to make better decisions don't have access. If healthcare providers and community-based organizations that serve these underserved communities had access to some of the data that other organizations collect and were able to then use that data to inform their own decision-making around their patients and community, we could get better results. It appears as though the relationship with providers in underserved communities is not equal. They have a lot of knowledge about what's going on with their circumstance up front, but they don't perceive academic public health experts as partners that they can work with and trust. Giving them data that they can use to serve their patients better would be much better, as opposed to the current method of criticizing overworked providers for not handling the situation the same way we would without our power, resources, and opinions.
JB: It is institutional will and commitment. We have to do a better job of removing the barriers to entry in every respect. In the case of COVID-19, the barriers to testing, vaccines, treatment, and recovery. We also have to spend time thinking about how to be accountable to the data that we collect, disseminate, and use to inform consumers. Doing so may also reduce the burden on data analysts and ensure accuracy and adequate interpretation for the general public.