Author’s Note

This work began in May 2017, at an MRCT Center Bioethics Collaborative, a neutral forum during which a number of clinical research stakeholders convened to discuss diversity in clinical trials. The attendees of that meeting, represented by a multi-stakeholder group of industry, academia, government, and patient advocacy, agreed that the participant population enrolled in a clinical trial ought to be representative of the general population at a minimum and, optimally, of the intended population for the intervention. If study populations are skewed, if they lack diversity, then the safety and efficacy, effectiveness, and value of medical interventions—the biological heterogeneity of treatment effect—cannot be adequately investigated and understood. Justice issues also influence diversity – or lack thereof – of study populations, with a fundamental unfairness perceived if specific populations are either disproportionately burdened, or unfairly excluded, from study enrollment. The attendees at that meeting agreed that, despite the understood necessity as a matter of science and ethics, underrepresentation of gender, sex, ethnic, and racial minorities in drug development, and in clinical research more generally, persists.

The extent of the problem, particularly for underrepresented and underserved populations, came into stark relief in the U.S. with the first publication of FDA Drug Trial Snapshots in January of 2015.[1] Drug Trial Snapshots reports on the demographics (sex, age, race, ethnicity) of patients who participated in the pivotal trials of either new molecular entities (NMEs) or Biologics License Applications (BLAs) that led to product approval in that year. The report is truly a “snapshot,” dependent upon the vagaries of the drugs and biologics approved in one year by one regulatory agency. With that limitation, the publication from the Center for Drug Evaluation and Research, developed in part in response to 2012 Food and Drug Administration Safety and Innovation Act (FDASIA 907), revealed striking disparities in participation by sex and race. In 2015, of 45 novel drugs approved, and with over 105,000 enrolled participants, only 40% of patients were women, and strikingly only 5% were African American. However, over the two year time frame of 2015 and 2016, 67 products were approved with dramatic variation by therapeutic area: the percent Black or African-American patients included was less than 3% of the total in trials of products for both cardiovascular diseases (2.50%) and oncology (2.74%) while 24.18% of participants were Black or African-American in psychiatric disorder trials.[2] Thus, racial diversity in clinical trial participation and drug development was possible, it just was not occurring and apparently not prioritized.

These stark and sobering observations led to a robust discussion at the Bioethics Collaborative: diverse representation is a principle of justice and of a just society, and our collective failure to achieve diversity is a solvable albeit difficult problem. Since that time, there have been numerous additional reports in both the scientific literature and the public press recounting the lack of diversity in clinical trials, across the spectrum of demographic dimensions of diversity: race, ethnicity, sex, gender, the elderly, the young, and genetics, as well as non-demographic variables such as comorbidities, polypharmacy, organ dysfunction, etc. Importantly, the impact of social determinants of health on health outcomes in clinical trials cannot be measured in the absence of validated methods for categorization, which have not yet been universally adopted.[3]

It should be understood that some dimensions of diversity (e.g., age, sex) represent biological differences, while others (e.g., race, ethnicity) represent social constructs, not fundamental biology. Race and ethnicity may, however, serve as surrogates, albeit inadequate and often flawed surrogates, for other factors such as genetic allelic frequencies, environmental factors, and social conditions, and analysis of study populations using those constructs can identify underrepresentation about which we as a society should be deeply concerned. The mission of health regulatory agencies is, in part, to protect the public health of its population—all its people, of every demographic—by ensuring the safety, efficacy, and security of drugs, biologics, vaccines, devices, and other products. Inclusion of all populations is necessary for reasons of justice, health equity, and trust.

Understanding the problem, and finding approaches to mitigate underrepresentation, requires focus and commitment, and a larger workgroup was formed in the fall of 2017 to address diversity, inclusion, and equity in clinical research. The group was comprised of representatives from academia, industry, patients, participants, advocacy organizations, regulatory agencies, non-profit organizations, and others. The group has worked steadily, meeting monthly, over the last 2+ years, and it has grown accustomed to often uncomfortable and challenging, but always respectful, conversation.

While the problem seemed important and relevant over the last few years, it is today an imperative that is foundational to society and to medicine. The COVID-19 pandemic exposed great inequities in health: Black, Latinx, Pacific Islander and some vulnerable (e.g., homeless, incarcerated, aged, institutionalized) populations have been disproportionately affected by the SARS-CoV-2 virus, and the disease has greater severity and mortality among those populations. This disproportionate impact appears to be related to comorbidities (e.g., hypertension, diabetes, obesity), access to healthcare and prompt testing, inequities in healthcare delivery, immune compromise secondary to chronic stress and other factors, exposure risks (e.g., density of living quarters, dependence on public transportation, work requirements), and/or potential genetic differences, among other factors. The fact that we do not know the relative contribution of these underlying factors, and lack data addressing them, exposes the degree of the problems we face today.

In this time of a global pandemic crisis came yet another example of racism in the United States with the tragic and painful death of George Floyd, and that death followed Ahmaud Arbery, Breonna Taylor, and countless others before them. In the U.S. and internationally, the world is rising to decry inequalities in power, opportunity, access, and, importantly, health.

This moment is a—long delayed—call to action. Eliminating racism and racial inequalities begins with eliminating disparities in health, and that necessarily demands deliberate and purposeful inclusion in health research that itself will help lead to equitable access and outcomes. This document addresses one part of that manifest inequity. To address that inequity successfully requires inclusion of diverse populations in research to advance the science – science that can then help create and implement data-driven, impactful solutions.

While we would like to believe that the work presented in this document is timely and relevant, in fact it is long overdue, and it is only a beginning. We appreciate that. We also know that we must start somewhere. The real work lies ahead, and for that, every member of the clinical research enterprise must commit, engage, and respond for real change to be lasting and impactful. We must value not only the imperative to understand biological differences but also the need to improve the health of all populations, eliminate disparities, and advance health equity.

Barbara E. Bierer, MD

June 15, 2020


1 U.S. Food and Drug Administration. 2015-2016 Drug Trials Snapshots Summary Report. Available at [Accessed 14 June 2020]
2 U.S. Food and Drug Administration. 2015-2016 Global Participation in Clinical Trials Report. Available at [Accessed 14 June 2020]. Note: the percent (number/total) of participants enrolled in the pivotal trials differed across the therapeutic areas: cardiovascular diseases 2.50% (1,415/92,329) , oncology 2.74% (211/7,480), and psychiatric disorders 24.18% (1,405/4,405).
3 The absence of data on social determinants of health is a significant deficiency, foreclosing opportunities for correlative studies that may be meaningful and potentially more relevant than other factors.