The pandemic has highlighted and exacerbated two stark realities about bias in medical research: People of color are routinely excluded from research studies — and Black and brown individuals, in turn, feel distant from and distrustful of scientists and those who fund them.
The result: While people of color are at disproportionate risk for severe illness and death from Covid-19, nearly 80 percent of all data used to study how genes influence health comes from people of European ancestry.
Despite growing evidence that a person’s geographical origin is an important determinant of disease risk, progression, and response to treatment, scientists, health-research grant makers, clinical trial organizers, and researchers have failed to fully incorporate it into most medical studies. This omission undermines our understanding of health, disease, and basic human cellular biology and slows progress toward adequate treatments for people who are not white or of predominantly European ancestry.
This lack of representation has numerous causes. Donors typically don’t specify the need for diverse study participants as a requirement for research funding. Many racial and ethnic groups have a justified mistrust of scientific research because of structural racism, historical wrongdoing, and socioeconomic and health care inequities — and a lack of engagement by researchers. Physicians who refer patients to participate in studies must contend with a time-consuming patient-referral process, restrictive participant criteria, and pervasive racial biases in medical training and practice.
The combined effect is low recruitment and enrollment of participants of color in clinical trials, underrepresentation in biomedical data, and poor understanding of the health conditions that affect Black and brown people.
Incorporating data from ancestrally diverse populations is simply good science. Leaving it out not only impedes medical progress at the expense of a majority of people in the world, but also ignores a wealth of untapped scientific inquiry and discovery. During the pandemic, it has meant that the world’s health care systems are less equipped to help people of color hurt most by the coronavirus.
Inadequate community outreach is a significant part of the problem, presenting an opportunity for grant makers. The philanthropic response to the pandemic always required more than support for laboratory research. It required partnering with health providers and community-based organizations in underserved areas and with public-health departments to spread science-backed information on the importance of masking, vaccination, and testing people at most risk for disease.
Connecting effectively with those most at risk during the research process should be common practice.
Physicians should be trained to engage with a diverse cross section of potential study participants. Institutions should broaden their inclusion criteria to accept a more diverse pool of participants. And grant makers should require the incorporation of community-based participatory research in all health research they fund. Through this approach, researchers work directly with people from grassroots groups on study design, data collection, and interpretation of findings — and then jointly share the results and benefits with community members.
Advancing a Cultural Shift
Grant makers can help bring about a cultural shift in health research by not only asking researchers to increase study diversity and community engagement, but also allowing them the flexibility to put systems and people in place to make such practices the norm. Community engagement should itself be treated as a science and standard component of any research project involving human subjects — integrated into the research design from the start, not as an add-on.
Foundations can accelerate this process by encouraging research grantees to develop a clear community-engagement plan specifying how researchers will work with diverse communities to gather data in an ethical and culturally sensitive manner. Several resources are available to help with this work, including community-engagement guidelines from the Patient-Centered Outcomes Research Institute, the National Institutes of Health, and the Centers for Disease Control and Prevention.
A handful of large-scale studies with philanthropic backing demonstrate the value of targeting diverse populations from the beginning.
The Dallas Heart Study, originally funded by the now-shuttered Donald W. Reynolds Foundation and then by the Hoffman Family Center in Genetics and Epidemiology, has made ethnic diversity of trial participants a priority since it launched in 2000. To recruit people of color, it connected with them through health care providers and community-based organizations, as well as places of worship and barbershops.
Public-Health and Social-Science Expertise
If achieving such diversity is a priority, scientists should not be the ones to lead these efforts. Computational biologists have expertise in analyzing data — not building relationships with a community. For that work, experts from disciplines such as the social sciences and public health need to be included on research teams.
Health studies supported by the Chan Zuckerberg Initiative, where the three of us work, take this approach. We stipulate the inclusion of community-engagement experts on research teams and request that potential grantees submit a plan for involving the communities their work affects as part of the funding-application process.
Such parameters are a critical component of our support of the international Human Cell Atlas project, a grassroots collaborative effort to create a map of all 37 trillion human cells to better understand the cellular mechanisms of health and disease. Last year, we launched a funding project to expand the diversity of cells included in the project, with a particular focus on understudied Black, Latinx, Southeast Asian, and Indigenous groups.
We encourage other research grant makers to follow our approach. The glaring inequities exacerbated by the pandemic show why new medical treatments need to be based on data representing the diversity of humanity. The tools to do this are readily available. We just need to use them.
