Research Fail: Lack of Representation of Some Harms All

This is not a new problem. The lack of minority representation in clinical trials is an ongoing source of health disparities for non-Caucasian patients, the Kentucky Neuroscience Institute recently reported. But this persistent lack of attention to addressing the representation gap runs the risk of further widening the divide in socioeconomic disparities related to health.

The lack of minority representation is a problem in emerging studies for multiple sclerosis, as recently reported, and also for emerging cancer drugs. Perhaps even more concerning is the gap in representation for genetic research, which may represent the future of personalized medicine.

In 1993, the Public Health and Welfare Act first enumerated requirements to ensure that women and minorities were included in research, which have been reiterated in subsequent versions of the act. It also required consideration of research outcomes based on sex and race/ethnicity, recognizing that attention to health outcomes on the basis of variability in sex and race may be important to advancing health of individuals from diverse groups.

Despite these requirements and federal guidance on how to collect race/ethnicity data, in over two decades, there are continued concerns about the lack of minority representation in research.  

Over these same two decades, research suggests that minority adults are underrepresented in research. There are variable reports in research on children, with studies showing both over and underrepresentation of minority children.

In addition, many studies do not contain enough information about race and ethnicity to appreciate whether outcomes are different by these markers. There may be other factors beyond race and ethnicity that are important socioeconomic markers, which receive less attention, including income, health and research literacy, age, health insurance status and education. Even less is known about why individuals from diverse groups fail to participate in research.

The lack of diversity in research generally may be particularly problematic for future efforts to develop individualized health care. The All of Us Research Program, launched in 2016, is a national effort to bring precision medicine to all diseases by enrolling a diverse cohort of 1 million or more individuals to advance health through research. Critical to the success of the program, and genomic research generally, is the ability to enroll diverse participants with respect to their lifestyle, environment and biology.

Yet, historically, genomic research has failed to capture diversity. Between 2009 and 2016, researchers noted the failure of genomic research to significantly increase inclusion of minority populations in research.

As the All of Us program has only recently begun enrollment, more time will be needed to know whether representation of diverse groups is adequate. However, the story of the last two decades suggests that we still have work to do to avoid continuing to perpetuate health disparities by limiting our strong evidence base to a non-diverse population.

A robust understanding of barriers to research engagement is critical, as is addressing newfound barriers. Existing research identifies factors related to participation, which require attention including age and education level.

Many scientists have argued for implementing known strategies in low health literacy populations to address some of these concerns. Simplifying consent forms and conversations is a key step in the right direction and required under the recent updates to the federal regulations regarding federally funded research. Strategies such as alternative media formats are debated but may have added value and require further study.

More needs to be done to identify other barriers to adequate representation. In these evaluations, we must reconceptualize how we think about diversity to include not only race/ethnicity and gender but other key factors related to social determinants of health that may impact how an individual will respond or not to a health intervention. We need to ensure inclusion of individuals from different income levels, with different primary languages, as well as groups that are geographically diverse to account for environmental factors that may relate to research outcomes.

To fully address inadequate representation of diverse groups, we must also turn attention to whether individuals from diverse backgrounds have access to the kinds of institutions where research engagement is possible.

In addition to tracking enrollment of individuals from diverse groups, we must pay equal attention to how often we are unable to engage such groups either due to unavailability or other barriers. The failure to approach individuals for inclusion guarantees their nonparticipation and will not be aided by strategies that address why individuals may say no to participation.

Efforts to address why individuals from diverse groups may refuse research participation must be strengthened. Trust in research is an often-cited reason for nonparticipation by diverse groups, particularly in precision medicine research. Indeed, trust is an important component of interpersonal relationships.

Distrust may result both from one’s own negative interpersonal experiences with the health system and also from systematic injustices in research such as the medical experiments conducted by Nazi investigators, the syphilis study at Tuskegee, and the case of Henrietta Lacks, that have eroded public trust in research.

When interacting with potential research participants, it is important to recognize the vulnerability of certain groups making decisions about research participation and to communicate sensitively, while respecting that the decision-making ultimately rests with the potential research participant. Encouraging public engagement in research at all stages and robust oversight can also bolster trust.

Fully embracing that proportional and desired participation is the goal will help to guide successful interactions. Both over- and underrepresentation are problematic as overrepresentation may lead to exploitation of certain groups and further erode trust in research, while underrepresentation denies certain groups the ability to benefit fully from the results of research.

As All of Us continues its efforts for enrollment to advance the future of personalized medicine, it should do so by promoting adequate opportunities, enabling authentic interactions that support decision-making by potential participants and upholding an ultimate goal of enrolling enough of us to ensure benefits to all.


Erin Paquette is a pediatric critical care doctor, lawyer and ethicist, assistant professor of pediatrics at Northwestern University’s Feinberg School of Medicine, adjunct professor at NU’s Pritzker School of Law, a Pediatric Critical Care Scientist Development Scholar and a public voices fellow through The OpEd Project.

Morning Consult welcomes op-ed submissions on policy, politics and business strategy in our coverage areas. Updated submission guidelines can be found here.

Morning Consult