Article Text

Download PDFPDF

Diversity in clinical research: public health and social justice imperatives
  1. Tanvee Varma1,
  2. Camara P Jones2,3,
  3. Carol Oladele4,
  4. Jennifer Miller5
  1. 1 Yale School of Medicine, New Haven, Connecticut, USA
  2. 2 Department of Community Health and Preventive Medicine, Morehouse School of Medicine, Atlanta, Georgia, USA
  3. 3 Harvard University Radcliffe Institute for Advanced Study, Cambridge, Massachusetts, USA
  4. 4 Section of General Internal Medicine, Yale School of Medicine, New Haven, Connecticut, USA
  5. 5 Department of Internal Medicine, Yale School of Medicine, New Haven, Connecticut, USA
  1. Correspondence to Dr Jennifer Miller, Internal Medicine, Yale University School of Medicine, New Haven, CT 06510, USA; jennifer.e.miller{at}


It is well established that demographic representation in clinical research is important for understanding the safety and effectiveness of novel therapeutics and vaccines in diverse patient populations. In recent years, the National Institutes of Health and Food and Drug Administration have issued guidelines and recommendations for the inclusion of women, older adults, and racial and ethnic minorities in research. However, these guidelines fail to provide an adequate explanation of why racial and ethnic representation in clinical research is important. This article aims to both provide the missing arguments for why adequate representation of racial and ethnic minorities in clinical research is essential and to articulate a number of recommendations for improving diversity going forward.

Appropriate racial and ethnic representation and fair inclusion help (1) increase the generalisability of clinical trial results, (2) equitably distribute any benefits of clinical research and (3) enable trust in the research enterprise.

  • Clinical Trial
  • Ethics- Medical
  • Public Policy

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Although black, Latinx and Indigenous patients have been disproportionately affected by the COVID-19 pandemic, evidence suggests racial and ethnic minorities may be under-represented in COVID-19 clinical trials.1 2 In fact, in the USA, racial and ethnic minorities are under-represented in clinical trials for most therapeutics and vaccines.3–5 Clinical trial participants are more often male and tend to be whiter, younger and healthier than real-world patients, raising concerns about the safety and effectiveness of novel therapeutics and vaccines for under-represented populations.6–8

The National Institutes of Health (NIH) and the Food and Drug Administration (FDA) have made strides to improve demographic representation in research since at least the 1980s. The NIH Revitalization Act of 1993, for example, directed the NIH to establish guidelines for the inclusion of women and racial and ethnic minorities in clinical research.9 Additionally, the statute required NIH-sponsored clinical research to address the inclusion of historically under-represented demographic groups in proposed research and in phase 3 clinical trials. Recently, the FDA published detailed guidance for research sponsors, as part of the FDA Reauthorization Act of 2017 mandate, on how to better enrol diverse study participants.10 The guidance makes recommendations, such as broadening protocol eligibility criteria and avoiding unnecessary exclusion criteria in research, to better understand the safety and efficacy of novel therapeutics in a diverse patient population.

While helpful, these efforts, like many, fail to answer a fundamental question: why is racial and ethnic representation in clinical research important? Given that demographic representation in research advances understandings of safety, efficacy and effectiveness for novel therapeutics in the population expected to use them, demographic representation is important if there is reason to believe that there will be a differential response to a therapeutic by sex, age, race or ethnicity. It is well established that there can be differential responses to therapeutics between women and men and between older and younger adults due to differences in pharmacokinetics/pharmacodynamics and drug toxicity by sex and age.11–13 Yet, in failing to explain why race, a social construct, is grouped with sex and age, biological attributes, we worry that guidance to improve representation of racial and ethnic minorities in clinical research may unintentionally endorse a biological basis for race and ethnicity. Here, we provide the missing arguments for why racial and ethnic representation in clinical research is essential. We then articulate a number of recommendations for improving the enrolment of racial and ethnic minorities in clinical research going forward.


Clinical characteristics and responses to novel therapeutics can differ by race and ethnicity. However, the underlying causes of any differences are driven by social determinants and structural racism, not inherent genetic differences.14 Racial and ethnic minorities are, for example, over-represented in poverty due to structural inequities, which results in disparities in access to quality education, healthy foods, a clean and safe environment, income and wealth, among other factors that contribute to a healthy life.15 16 This manifests in a disproportionate rate of comorbid conditions and concomitant medication use, both of which can alter the safety profile of a therapeutic.17 18 Therefore, failing to adequately include patients identifying as racial and ethnic minorities in research, can potentially limit understandings around the benefits, risks and harms of novel therapeutics for excluded and under-represented patients groups.

While race is not a biological construct, it is worth noting that evolutionary history or ancestry may be biologic constructs of value. However, ancestry is not a proxy for race and should not be used as such. Genetics research has illuminated how humans cannot be divided into biologically distinct categories and that there is more variation within racial groups than between racial groups.19 20 Although there may be some genetic similarity among members of geographically circumscribed populations, this is not generalisable to all members of a racial group. As such, ancestry may only be relevant in clinical research if used to describe a very specific population and must be explicitly defined in research protocols.21 22

Given that structural inequities contribute to racial disparities in health, one could argue that the use of race in clinical research should therefore be replaced with socioeconomic variables. If we could select trial participants based on socioeconomic status, comorbidities and concomitant medication use, would we still need racial and ethnic representation in clinical research? We argue that racial and ethnic representation in clinical research would still be an imperative for several reasons.

First, socioeconomic variables are not proxies for the ill effects of racism on health. Research suggests that racial disparities in health persist even after controlling for socioeconomic status.23 This is because living in a society that assigns value based on the social interpretation of how one looks (which is what we call ‘race’) results in differential opportunities, exposures, resources and risks by race and ethnicity.24 Differences in chemical and toxicological exposures as well as allostatic load from chronic stress can result in physiological differences that are unrelated to genetic inheritance and are instead a manifestation of social inequities inherent to a racist society.25 26 These differences cannot be modelled by any one or set of socioeconomic variables in clinical research.

Second, there may be benefits to research participation, which should not be limited to a privileged few. While bioethicists have long warned against the therapeutic misconception—the incorrect belief that the purpose of trials is to treat participants, rather than gather scientific knowledge—there is an emerging, although controversial, viewpoint that clinical research can benefit trial participants. One study found that patients enrolled in trials, even those randomised to a control arm, had better outcomes than patients receiving a standard of care but not enrolled in a trial, suggesting there may be ‘collateral’ benefits to research participation, such as enhanced medical attention and oversight.27 However, other studies do not find clinical benefits to trial participation.28 During this pandemic, participation in COVID-19 vaccine trials has led to early access to emergency use authorised vaccines for some trial participants.29 If there are benefits to trial participation, then under-representation of racial and ethnic minorities in clinical trials suggests benefits are being inequitably accessed.

Further, under-representation in clinical research may exacerbate an existing and justified mistrust in research and novel medicines and vaccines that stems from a long history of exploitation of black patients by medical researchers as well as structural racism perpetuated within and by the US healthcare system.30–32 Adequate racial and ethnic representation in clinical research is a necessary step towards fostering trust in and use of novel therapeutics among minority communities.


Below, we articulate a number of recommendations for improving enrolment of racial and ethnic minorities in clinical research based on the current literature and our collective expertise and experience working in the area.

To improve demographic representation in clinical research and advance research equity, we recommend first and foremost that all research guidance documents include an explanation of why race is categorised with sex and age in considerations of demographic representation in clinical trials. This is important for trial participants as well as for clinical trial teams to understand and reframe why diversity in clinical research is important. Not doing so risks unintentionally endorsing a biological basis for race, which has been historically used to justify and perpetuate structural racism.19

Second, we need to enhance transparency around who is enrolled in clinical research. While most clinical trials publicly report the numbers of enrolled women, few publicly report the numbers of enrolled patients identifying as racial or ethnic minorities.5 We suggest more rigorously collecting and disclosing data about the racial and ethnic identity of clinical trial participants.

In addition to collecting data about clinical trial participants, we must collect and make publicly available data about the incidence of health conditions in the USA by sex, age, race and ethnicity. Even when these data are collected, they are often not specific enough to match the indications for which drugs are approved or they are not publicly available, thereby preventing stakeholders from being able to rigorously assess demographic representation in clinical trials. An example of a solution is Flatiron Health, which has electronic health record data for two million active US patients with cancer from more than 800 geographically diverse sites. Companies like Flatiron Health could aggregate patient profiles and create free and publicly-available data for various conditions as a public service and share them through a public registry of patient profiles. This public registry, often generated from patient electronic medical record data, should not be considered proprietary, but rather a public good.

Fourth, with these data, research equity metrics should be developed to measure and track demographic representation in clinical trials by sponsor, indication and/or therapeutic. The participation to prevalence ratio (PPR), developed by Poon et al, may be one such equity metric. The PPR is calculated by dividing the percentage of a particular population among the study participants (eg, percentage of Asian adults among participants in hepatitis B trials) by the percentage of the particular population in the disease population (eg, percentage of Asian adults among individuals in the United States with hepatitis B). A PPR from 0.8 to 1.2 indicates adequate representation; a PPR less than 0.8 indicates under-representation, and a PPR greater than 1.2 indicates over-representation.5 33 Developing such metrics will allow us to hold research sponsors accountable, recognise best practices that may be generalisable, and track progress towards our goal of having a more equitable research enterprise.34

Fifth, contextual social information about research participants should be routinely collected in clinical research where possible, such as years and quality of education, household income, neighbourhood characteristics, among others. Some research has shown that despite standardisation in treatment, protocols and follow-up inherent to randomised controlled trials, socioeconomic variables contribute to wide disparities in trial outcomes, suggesting a need for more careful consideration of social determinants of health when designing randomised controlled trials.35 This will also allow researchers to develop a more nuanced understanding of the ways in which various socioeconomic factors are the drivers of racial disparities in clinical outcomes, rather than assuming that racial disparities in clinical trial outcomes are due to an inherent biological difference between racial and ethnic groups.36

Sixth, improving merited trust and appropriate patient engagement in research is essential to the research enterprise.37 Stakeholders that benefit from the research enterprise must engage genuinely with and address the concerns of minority communities regarding research participation. Engaging diverse communities throughout the full research process, from agenda setting, to trial design and pricing, through collaborative partnerships, is important for meeting the needs of different demographic groups and appropriately building trust in FDA-approved products. This includes ensuring researchers and trial recruiters are diverse. An NAACP study found black Americans were twice as likely to trust a messenger of their own racial/ethnic group than one from outside it.9 Additionally, some research has found that research and clinical professionals view black patients as less likely to enrol in clinical research and, therefore, clinicians spend less time discussing clinical trials with black patients.38–40

Seventh, racial and ethnic minorities must also be adequately represented among patients who benefit from research, such as through access to successful therapeutics. The infamous US Public Health Service Study of Untreated Syphilis in the Negro Male is only one example of many in history of racial and ethnic minorities being over-represented in research, and therefore, incurring a disproportionate burden of research risks, while bring under-represented in access to research benefits, including effective therapeutics and vaccines.41 Therefore, pharmaceutical companies should not only design strategies for a diverse clinical trial population but also for equitable distribution of the therapeutic once it is approved, in collaboration with local public health departments, health systems and community organisations.

Finally, resource commitments are necessary to accomplish the above goals. Because of structural racism, efforts to recruit and retain a diverse research workforce, including minority research participants, scientists and community leaders, to name just some examples, is no small challenge and will likely require larger efforts and budgets. Research suggests there remain structural barriers to representation of racial and ethnic minorities in clinical research including, less access to transportation, healthcare, and child care, as well as language and geographic barriers, to name a few.42–44 A sizeable and necessary investment will need to be made to address these structural barriers. It is a social justice imperative to allocate resources according to need.45 If we are truly committed to increasing representation, then we must invest in communities where there has been systematic disinvestment.

Ultimately, ‘race’ is not genetic but racism can manifest in biology. Therefore, racial and ethnic minorities must be adequately represented in clinical research. Not doing so calls into question the generalisability of clinical trial results, inequitably distributes the risks and benefits of research, and perpetuates a justified distrust in the research enterprise.

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.

Ethics statements

Patient consent for publication



  • Twitter @TanveeVarma, @millerbioethics

  • Contributors TV and JM had the original idea of the paper. TV wrote the initial draft of the manuscript with support from JM. All authors (TV, CPJ, CO and JM) were involved in editing the manuscript and producing its final version. JM is responsible for the overall content as guarantor.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Other content recommended for you