By Aman, Fall 2020.
The ease with which the COVID-19 pandemic swept through the nation and killed hundreds of thousands of people has made many Americans question the effectiveness of their country’s institutions and systems of healthcare. As the American healthcare industry is more heavily scrutinized, more disparities, unjust practices, and systemic problems come to light. One notable issue being highlighted is the disparities that exist between the COVID-19 infection and mortality rates between white Americans and Americans of color. Despite comprising only 13% of the population, Black Americans comprise more than 34% of total U.S. COVID-19 cases [1]. Black and Hispanic Americans are at higher risk of being infected, hospitalized, and killed by the coronavirus [1]. However, these incongruities are often assumed to stem from genetic differences between races, and the problem of systemic racism within American healthcare is blithely dismissed [2,3]. In reality, America’s mishandling of the COVID-19 pandemic has revealed the extent to which people of color are systematically excluded from the American healthcare system. Disparities in patient outcomes for people of color cannot be attributed to “biological difference” or rogue prejudiced individuals: systemic racism in healthcare is pervasive and multi-dimensional, encompassing not only the biases of individual physicians but also the curriculum of medical schools and the technology used by healthcare administrators for diagnosis and treatment.
Technology, particularly biomedical equipment, is often regarded as objective in its evaluation of illness. Unfortunately, technology exacerbates racial inequalities in healthcare because of the biases baked into its code. For instance, the pulse oximeter is a purportedly objective assessor of oxygen saturation levels in a user’s bloodstream, which has become an essential tool for most American households during the pandemic. However, when the oximeter was first designed and tested, it was calibrated specifically for white skin, and it is much less accurate (up to 8%) when assessing the oxygen saturation levels of people with darker skin [3]. For many people, the difference between a saturation reading of 95% and 87% could be the difference between good health and respiratory failure [3]. Additionally, oximeter tests are used by private and public health insurance agencies alike to determine whether patients are eligible to receive oxygen tanks or, early in the pandemic, COVID-19 tests [3]. In practice, this means that patients of color must present as more physically ill to access the same care as white patients.
Racial bias in medical technology is not limited to diagnostic practices and treatment: preventative measures implemented by algorithms and data-driven science also perpetuate healthcare inequalities. This was made most apparent by the significant lack of COVID-19 testing sites in Black communities in California, Texas, and Tennessee, which were all states that claimed to be using novel algorithms to determine which communities had greater testing needs [2]. In general, algorithmic technology has a history of discriminating against people of color in a medical context. For instance, a University of Chicago Medicine study investigated how one hospital used an algorithm to determine which patients required long-term care and treatment priority. According to the algorithm, only 18% of the hospital’s Black population required critical care, when, in reality nearly 47% of them did [2]. The discrepancy arises from how the algorithm was trained to identify and set aside cases with high anticipated costs for the hospital (which would more often be Black patients because they were more likely to have lower-tier health insurance plans and to not be able to pay out-of-pocket) [2]. The problem with these technologies is that they rely on imperfect status quo data to determine future measures, exacerbating discriminatory practices.
What makes the “discriminatory design” of medical technology even more difficult to address is that many doctors have no idea that this problem exists. In fact, the medical school curriculum completely glosses over systemic racism in medicine and often treats white patients (and white skin) as default [4,5]. Many textbooks used to teach dermatology severely underrepresented the presentation of diseases on non-white skin, so doctors have very little experience with diagnosing and treating those conditions [4,5]. In fact, nearly half of all dermatologists believe that they have had insufficient exposure to manifestations of disease in darker skin types. This lack of experience could have played a significant role in the spread of COVID-19 in communities of color, as skin rashes on asymptomatic COVID-19 patients of color could have been completely missed by medical professionals. A slew of recent studies have indicated that depictions of COVID-19-related skin conditions focus almost exclusively on the presentation of these diseases on white skin [5]. Of all officially published images of COVID-19 rashes, 92% were of light white skin [5]. The great dearth of information about the reaction of darker skin to COVID-19 has resulted in misdiagnosis and mistreatment of patients of color.
In addition, COVID-specific problems of medical bias are compounded by the assumptions many healthcare administrators already make about people of color. Physicians are less likely to identify and treat diseases in people of color [5,6]. In fact, physicians are more likely to dismiss the concerns of people of color as anxiety and refuse to administer treatment [2, 3]. For these reasons, communities of color are much more likely to be underdiagnosed and undertreated for COVID-19, especially since the pandemic has increased the amount of strain on the American healthcare system.
As evidenced by the tremendous diversity of ways in which racial healthcare disparities manifest, the problem of systemic racism in medicine is not limited to one set of factors but is rather all-encompassing. COVID-19 may have exacerbated many of these problems, but solving the pandemic will not address the systemic problems that made the pandemic so difficult to deal with in the first place. Only a thorough examination and systematic reorganization of the American healthcare system can resolve the pervasive problem of inequitable treatment Americans of color receive.
[1] Moore, Jazmyn T et al. 2020. “Disparities in Incidence of COVID-19 Among Underrepresented Racial/Ethnic Groups in Counties Identified as Hotspots During June 5–18, 2020 — 22 States, February–June 2020.” Morbidity and Mortality Weekly Report 69: 1122—1126. http://dx.doi.org/10.15585/mmwr.mm6933e1
[2] McCullom, Rod. 2020. “Artificial Intelligence, Health Disparities, and Covid-19.” Undark. Last modified July 27, 2020. https://undark.org/2020/07/27/ai-medicine-racial-bias-covid-19/
[3] Moran-Thomas, Amy. 2020. “How a Popular Medical Device Encodes Racial Bias.” Boston Review. Last modified August 5, 2020. http://bostonreview.net/science-nature-race/amy-moran-thomas-how-popular-medical-device-encodes-racial-bias
[4] Papier, Art. 2020. “To begin addressing racial bias in medicine, start with the skin.” Stat. Last modified July 20, 2020. https://www.statnews.com/2020/07/20/to-begin-addressing-racial-bias-in-medicine-start-with-the-skin/
[5] Lipper, Graeme M. 2020. “COVID-19-Related Skin Changes: The Hidden Racism in Documentation.” Medscape. Last modified July 27, 2020. https://www.medscape.com/viewarticle/934605
[6] Egede, Leonard E. 2006. “Race, Ethnicity, Culture and Disparities in Health care.” Journal of General Internal Medicine 21, no. 6 (June):667–669. 10.1111/j.1525-1497.2006.0512.x.