Highlighting the Unequal Burden of COVID-19: Informing Possible Pathways for Intervention

Introduction

Reporting of health disparities and the unequal burden of COVID-19 began early in the epidemic in both the popular and scientific press. For example, National Public Radio and the Centers for Disease Control (CDC) reported racial disparities in COVID-19 outcomes by mid-April of 2020, only a month after the WHO declared COVID-19 a pandemic. Several other timely reports on racial disparities appeared in the popular press that same month [1,2,3].

The rapid recognition of disparities in COVID-19 is in stark contrast to the experience during the H1N1 pandemic. It was almost two years after the WHO declaration of the H1N1 pandemic that the American Journal of Public Health published “the first empirical examination of disparities of H1N1 exposure, susceptibility to H1N1 complications, and access to health care during the H1N1 influenza pandemic” [4].Similarly, recognition of racial disparities in H1N1 by the popular press was not as timely, with only a few pieces published within the first 6 months of the WHO declaration of the H1N1 pandemic [5,6].

Reporting on health disparities during the COVID-19 pandemic has shined a spotlight on long-standing structural inequities and the health effects of racism and socioeconomic status. While this public and scientific recognition of COVID-19 inequities is critical, the alarm bells must be followed by rigorous and actionable research on interventions that help mitigate the disparities [7]. This post describes example research questions, possible approaches, and tools and resources to help facilitate research and action.

Example Research Questions

To intervene effectively, we must first understand which factors may contribute to specific populations experiencing a higher burden of COVID-19. Disparities in disease burden may be driven by unequal exposure (e.g., due to occupational differences), susceptibility to complications from an infection or survival after infection (e.g., due to pre-existing conditions or co-morbidities), access to health-promoting resources (e.g., high-quality health care) [4,7], or some combination of these.

Disparities in exposure. Research suggests that people of color and low-income individuals are more likely to have jobs that can’t be done virtually, rely on public transportation, or live in densely populated areas [4,7]. These conditions lead to greater interaction with people outside one’s immediate social circle, resulting in higher risk of COVID-19 exposure. What interventions might reduce the disproportionate exposure to COVID-19?

Disparities in susceptibility. Many racial or ethnic groups already experience health disparities in underlying medical conditions associated with developing complications from COVID-19 (e.g., heart disease, chronic obstructive pulmonary disease, uncontrolled asthma, diabetes, and high body mass index) [8]. What interventions might reduce disparities in chronic conditions or mitigate the consequences of COVID-19 for people with underlying medical conditions?

Disparities in access. Limited access to timely and high-quality medical care increases the likelihood of COVID-19 complications and deaths. Barriers to timely, high-quality care may include inadequate health insurance, fear of accessing care, lack of a previously established primary care provider or other regular source of care, or limited resources in the facility where care is received. What interventions might improve access and quality of medical care received for disadvantaged groups?

Possible Approaches

The rapidly shifting nature of COVID-19 responses combined with the simultaneous implementation of interventions at multiple levels of government makes it challenging to isolate the outcomes of any one intervention or determine its effectiveness. However, some research designs are better suited than others for understanding the different drivers of disparities in COVID disease burden and how to address them. We highlight possible research approaches that may aid in identifying which interventions are effective in addressing the drivers and/or outcomes of disparities in exposure, susceptibility, and access to COVID-19, as well as examples studies.

Addressing disparities in exposure. The timing of when a new policy or practice is implemented can be leveraged to draw causal inferences between the intervention and measured outcomes. An interrupted time series (ITS) approach is particularly well suited for this type of analysis. The ITS design allows researchers to compare trends immediately before and after policy implementation. For example, assessing the impact of statewide mandatory mask policies using an ITS approach demonstrates ten-times fewer excess COVID cases in states that reopened with mask mandates versus those without [9].  

Addressing disparities in susceptibility. An instrumental variable (IV) approach allows the causal effects of an intervention or exposure to be isolated by disentangling the part of the intervention/exposure that impacts the outcome from the factors that do not. For example, an IV approach can be useful in assessing the contribution of air pollution to COVID-19 outcomes. Such an assessment, with wind direction as the instrument, demonstrates that U.S. cities with worse contemporaneous air pollution have higher rates of confirmed cases and deaths from COVID-19 [10]. Prior research suggests that people of color and low-income groups are more susceptible to the effects of air pollution, and adverse health effects, due to segregation in housing [11,12,13].

Addressing disparities in access. Randomized controlled trials (RCTs) are often preferred when feasible because randomization helps reduce the chance that factors outside of the intervention of interest are affecting outcomes. It would be unethical to manipulate access or quality of resources available to people in the wake of COVID-19. However, randomized trials that were already underway prior to the COVID-19 outbreak present an opportunity for understanding whether an intervention is effective in the face of public health and economic shocks. For example, the Stockton Economic Empowerment Demonstration (SEED) RCT – the country’s first city-led guaranteed income (GI) pilot – had been underway for about a year when the pandemic hit. Researchers are capitalizing on the RCT that was already in place to evaluate whether receiving a GI inoculates individuals against the effects of COVID-19.

Putting Evidence into Practice

When designing studies to provide insight into health disparities in COVID-19, we must consider the generalizability of the findings that different research approaches offer. Observational studies that often use convenience samples may provide expedited findings but can systematically under-represent low-income individuals, certain racial/ethnic groups, and hidden or marginalized populations, thereby biasing, confounding, or incompletely representing the burden of COVID-19 for important subgroups. While often considered a gold-standard approach, RCTs can often be too costly, unfeasible, or unethical to implement. Quasi-experimental designs (e.g., ITS and IV), however, may prove useful for providing timely answers to questions about health disparities in COVID-19 without compromising the generalizability of a study’s findings because they allow for better control of confounding variables thereby increasing the validity of a study's findings. 

As mentioned, the rapidly shifting nature of COVID-19 responses complicates our ability to isolate any given intervention's causal impact. We need timely data representative of the health and social interventions implemented during COVID-19 to make policy and programming decisions and elevate the most effective and promising interventions. E4A is currently funding one such endeavor to establish a comprehensive, publicly available COVID-19 U.S. State Policy database – CUSP. The CUSP database documents the implementation dates of health and social policies in the wake of COVID-19 and its economic ramifications in all 50 states and the District of Columbia. The CUSP database offers an essential source of data for evaluation of policies implemented across different places and times shape outcomes such as COVID-19 cases and deaths, as well as financial distress, food insecurity, housing insecurity, and mental distress, particularly for populations disproportionately affected by COVID-19 and its economic ramifications. Stay tuned for more information about the development and possible uses of the CUSP database in a future E4A blog post.

Tools & Resources

There has been considerable recognition of COVID-19-related health inequalities and disparities by the CDC and other institutions. The list below highlights tools and resources especially useful for tracking, assessing and understanding disparate impacts of COVID-19. (Note, this selection of tools and resources is not exhaustive but rather highlights a few examples in popular and scientific press). 

• The CDC has a webpage explicitly describing the health equity considerations related to COVID-19, a health tracker of COVID-19 cases and deaths by race and ethnicity, and a webpage on hospitalization and death by race/ethnicity. 
• Emory University has created a COVID-19 Health Equity Interactive Dashboard.
• John Hopkins has an interactive web-based dashboard to track COVID-19 in real-time.
• Northwestern has published resources for funding, publications, data, and teaming for COVID-19 research.
• The COVID Tracking Project, a volunteer organization, launched by The Atlantic, has a COVID-19 Racial Data Tracker.

The timeliness and frequency of reporting on health disparities related to COVID-19 are impressive. However, to reduce the immediate and long-term impact of COVID-19 on the groups traditionally experiencing health disparities, evidence must be collected and put into action.

References

1. Mays, J.C. and Newman, A. (2020, April). Virus is twice as deadly for Black and Latino people than Whites in N.Y.C. New York Times.
2. Aleem, Z. (2020, April). New CDC data shows Covid-19 is affecting African Americans at exceptionally high rates. Vox.
3. Edwards, E. (2020, April). African Americans 'disproportionately affected' by coronavirus, CDC report finds. NBC News.
4. Quinn, S. C., Kumar, S., Freimuth, V. S., Musa, D., Casteneda-Angarita, N., & Kidwell, K. (2011). Racial disparities in exposure, susceptibility, and access to health care in the US H1N1 influenza pandemic. American Journal of Public Health, 101(2), 285-293.
5. Smith, S. (2009, August). Cases of swine flu higher among city blacks, Hispanics. Boston.
6. Gonzalez, J. (2009, October). Swine flu's bigger impact on blacks and Hispanics is not being addressed. New York Daily News.
7. Bibbins-Domingo, K. (2020). This Time Must Be Different: Disparities During the COVID-19 Pandemic. Annals of internal medicine, 173(3), 233–234.
8. Koma, W., Artiga, S., Neuman, T., Claxton, G., Rae, M., Kates, J., & Michaud, J. (2020, May). Low-Income and Communities of Color at Higher Risk of Serious Illness if Infected with Coronavirus. Kaiser Family Foundation.  
9. Kaufman, B. G., Whitaker, R., Lederer, N. M., Lewis, V. A., & McClellan, M. B. (2020). Comparing associations of state reopening strategies with COVID-19 burden. Journal of General Internal Medicine, 1-8.
10. Austin, W., Carattini, S., Mahecha, J. G., & Pesko, M. (2020). COVID-19 mortality and contemporaneous air pollution (No. paper2016). International Center for Public Policy, Andrew Young School of Policy Studies, Georgia State University.
11. Tibuakuu, M., Michos, E. D., Navas-Acien, A., & Jones, M. R. (2018). Air pollution and cardiovascular disease: a focus on vulnerable populations worldwide. Current Epidemiology Reports, 5(4), 370-378.
12. Jones, M. R., Diez-Roux, A. V., Hajat, A., Kershaw, K. N., O’Neill, M. S., Guallar, E., ... & Navas-Acien, A. (2014). Race/ethnicity, residential segregation, and exposure to ambient air pollution: the Multi-Ethnic Study of Atherosclerosis (MESA). American Journal of Public Health, 104(11), 2130-2137.
13. Morello-Frosch, R., & Lopez, R. (2006). The riskscape and the color line: examining the role of segregation in environmental health disparities. Environmental Research, 102(2), 181-196.