Research

Diversity in Clinical Trial Research

Best practices for outcome equity

BY Maya Palmer and Courtney Jordan Baechler, MD, MS

Clinical trials in the United States are predominantly made up of white, middle class, male patients. As a result, many individuals are left behind in the pioneering of the future of medicine. Furthermore, many trials lack appropriate demographic data contributing to an even greater disparity in understanding the extent of the exclusion of BIPOC (Black, Indigenous and People of Color), women and gender nonconforming individuals in trials. 

Research trial execution must achieve a representative sample, requiring greater focus in improving diversity in trials. Unfortunately, even the research sites dedicated to equity, aiming to be leaders in cutting-edge medical research, fail to supply necessities that would aid in improving diversity in trials. In this article, the quantification of English fluency at a multicenter research facility and the importance of multilingual resources are explored as one approach to improving diversity in research.

Racial Diversity and Study Efficacy

Marginalized communities are being denied the opportunity to participate in innovative and potentially life-saving care through research. Inadequate representation can have unknown adverse effects, harming understudied patient populations even after clinical trial approval. Although it should require little qualification, medical interventions not tested in a reflective sample have poor study efficacy. Representation in research can capture variations in biological and pharmacological processes that result from lived experiences (epigenetics, stress resilience, nutrition, education) and genetic variation. Poor study efficacy results in adverse outcomes for patients. For example, FDA-approved interventions have had consequences on patient well-being that should have been caught during trial testing rather than retrospectively. Studies observing ACE inhibitors or ARBS between 2012 and 2018 did not capture higher rates of angioedema in Black Americans. It was not until 2020 that the prevalence of angioedema was understood for Black Americans, thus impacting the use of these medications as first-line antihypertensives and stressing the importance of symptom awareness for Black patients. Similarly, the dosage variation of warfarin was not utilized for genotype-guided dosing until 2013, largely again due to the lack of diversity in early trials. Interest in the effect of diverse bioprocesses has become pertinent to consider when evaluating the efficacy of a medical intervention. Remember, racial categories are a social construct and do not reflect genotype, but increasing racial diversity in trials does show an increase in general genotype variation as well as epigenetic environmental influences. Research has new opportunities with the advancement of algorithms and statistical capabilities to evaluate more complex variations of medical interventions among patients. Maximizing treatment effectiveness and avoiding adverse outcomes for patients start at a representative research sample.

Medical interventions not tested in a reflective sample have poor study efficacy.

Demographics and English Fluency at a Multicenter Research Institution

Demographics of potential research participants based on incoming clinical visits to a multi-center site were collected for five months (November 2023 to April 2024). Variables such as interpreter status, preferred language, race, sex and country of origin were observed.

The patient population across research sites were mostly white (92%, n = 28,380). 51% of patients were female (n =13,980). 1.6% of patients required an interpreter (n = 477). The top preferred languages of individuals requiring an interpreter were Russian (27%), Somali (25%), Spanish (16%) and American Sign Language (6.1%).

Given that increasing diversity in research patient samples was of interest, the relationship between race and interpreter status was observed. We define “racially underrepresented in research” as individuals who identified as Black or African American, Asian, American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, or two or more of the stated races including white. We found that racially underrepresented individuals in research were significantly more likely in need of an interpreter.

Increasing Racial Diversity and Multilingual Resources

Inclusivity is a foundational component of successful research. Through the quantification of patient demographics, our research center gained a better understanding of the population it serves. The inability to approach individuals who do not speak English hinders an inclusive approach to research participation. Below are several ways in which multilingual resources could be implemented for clinical trials.

Interpreters Negotiation of interpreter services should be made with sponsors during study start-up at site initiation. The use of interpreters from clinics/hospitals can be baked into follow up and enrollment visits, allowing customization of when research interpreters are specifically needed. In-person interpretation will not always be necessary, in which phone call or online interpretation could be deemed ethical depending on risk of study. Because we are one of very few countries that allows pharmaceutical and device companies to advertise, it seems that the funds needed to pay for interpreters is likely possible. Universities have a unique opportunity to provide tracks for multilingual students to participate in research translation services. As the population of premedical students continues to diversify, including languages spoken, students have expressed interest in volunteering in clinical research and broadening patient inclusion. Furthermore, hiring practices that engage in community hiring increases diversity of research staff and thus staff knowledge of highly specialized, practical, community-focused patient outreach. The more the staff reflects the community, the easier it is to have multilingual researchers, and a broader patient population. Overall, the use of interpreters in research is underutilized and is an important necessity in improving the scope of patient recruitment.

Materials Industry sponsors should provide consent forms and patient-facing materials in several languages, allowing sites to pick and choose which languages they need to serve their community. If unavailable directly from the sponsor, translation and IRB approval could be budgeted by the sites. Research institutions and sponsors should aim to optimize their websites containing research information for automatic translation. Inclusive research study advertising should be included in study start-up and should never be in only English. Furthermore, previous research findings on how various racial, ethnic, and gender groups engage in materials should be utilized when creating patient advertising/resources. Although ethically different from commercial advertising, research resources should mimic the same tenacity shown in reaching individuals across cultures to buy certain products. Inclusive multilingual resources created at the beginning, rather than as an afterthought, heavily influence a patient’s ability to choose to participate in research.

Intentionality and Quantification Consistent consideration by both the sponsor and research site of how recruitment operations function as well as quantification of who is being recruited will create a harmonious opportunity to both optimize and improve patient outreach. Quantification of patient demographics will also assist in understanding unique barriers depending on site location and increase transparency of what patient populations are at risk of exclusion. Intentional inclusivity and measurement of diversity quotas will help guide future improvements of increasing diversity in studies. We are at a critical time in medical research and healthcare delivery — we should do more than admire data and be complacent with whoever is the first to be recruited. Achieving specific set standards of recruitment that are truly reflective of a community is a necessity.

Institutional Paralysis, Bias and FDA Standards 

Although multilingual resources are a tangible investment in improving diversity in research, it is only one step toward improving a representative research sample. Lack of diversity in research is multifaceted. Increasing heterogeneity will require efforts spanning research staffing, study start-up, study execution and community leaders. Authors from “Improving Representation in Clinical Trials and Research: Building Research Equity for Women Underrepresented Groups (2022)” clearly outline the various approaches of improving equity in research.

Institutional Paralysis 

Another factor to consider in the challenge of improving diversity in research is “institutional paralysis,” in which larger systemic changes such as access to healthcare are perceived as beyond the scope of the company therefore paralyzing any internal efforts towards improving diversity. At our site, 92% of patients visiting the clinics were white. This is in stark contrast to the racial diversity of the clinical locations. Systemic barriers such as lack of universal healthcare, transportation and community engagement all contribute to the lack of racial diversity in clinics, impacting a researcher’s ability to approach racially diverse patients. However, research institutions should avoid the paralysis of systemic barriers. All research studies are responsible to enact what agency they do have to increase diversity at their sites, while also advocating for the larger removal of systemic barriers that impact diversity upstream. The multifaceted nature of increasing diversity does not excuse research institutions from participating in the change.

Inclusivity is a foundational component of successful research.

In our data, 12% of our BIPOC patients need multilingual resources. If we are unable to support this small number of patients, how will we be ready with the near future of medicine proving to have an even more diverse patient population? A representative research sample is of top priority, and opportunities such as multilingual resources are essential to pushing the needle forward. At a minimum, as scientists and physicians we should ensure we are collecting these demographics and making innovative attempts as outlined above at utilizing existing resources, as well as leaning on those resources that could be easily negotiated to improve outcomes.

Challenging Bias 

The challenges research institutions face from improving diversity in trials are not from the recruitment and participation of BIPOC patients. It has been robustly supported that BIPOC patients are more likely to participate in research and have better follow up rates than white participants. The obstacles research institutions face in improving diversity in studies stem from historical discrimination and oppression of BIPOC including unconscious or intentional tailoring of patient outreach to benefit enrollment of white participants only. Harmful narratives such as an added difficulty of enrolling BIPOC patients place blame on the participants themselves rather than the institutional discrimination research upholds in exclusive trials. The daily navigation of insurance coverage, translator scheduling and transportation resources only gets easier as we make them priority focuses during study start-up. To increase diversity in trials, research institutions must be willing to do the work that focuses specifically on breaking down the challenges that systems originally made to make access to health care difficult for BIPOC patients.

Medical Mistrust and Research Another form of bias is the assumption that extra efforts to approach BIPOC patients are futile because of medical mistrust. Medical mistrust is a consequence of historical oppression, systemic barriers and experiences of racism. Medical mistrust is an invalid reason not to approach patients for research. The definition of medical mistrust was never supposed to become a reason for not approaching patients, it aims to serve as an understanding on how clinical outcomes such as commitment to treatment plan and health outcomes are influenced. Refusal to approach patients because of assumed medical mistrust is not only harmful (revokes agency and consent by never giving patients the opportunity to make decisions themselves) but is directly adding to the mistrust (unequal treatment is perceived as racism and furthers mistrust). Staff concerned about medical mistrust must challenge themselves to take the responsibility of addressing the mistrust with the patient.

Rejection to changing health care and research systems to be more inclusive based on preconceived notions that efforts will not make a difference of a patient’s mistrust is discriminatory and directly perpetuates mistrust. If inclusivity has not been built into research, what have we done to show that patients can trust us? There has been little effort to rectify the systemic barriers that prevent BIPOC patients from accessing research just as easily as white patients. The lack of action to disrupt these barriers creates an intentional selection of white patients. Awareness of the barriers but lack of action furthers mistrust as such indifference overtly communicates to BIPOC patients that white patients are more deserving of improved health outcomes and new medical advancements. This is evident through disparities observed in disease prevalence, severity and mortality in Black and Hispanic patients. When research is designed to test treatment options for white patients only, the health of white patients are being selected over BIPOC patients.

FDA Standards 

It is time to move from the 2020 guidelines to required standards for racial diversity in research. The 2020 guidelines are not enough to push research practices toward more representative samples, or we would have seen stronger evidence of diverse samples already. FDA requirements emphasize to research sites that this is a critical investment in conducting trials, which helps redistribute financial priorities and building of long-term infrastructure for supporting all research participants. Interpretation services should be mandated, forcing sites to adapt their infrastructure to inherently fit the needs of LEP or ASL patients. Mandates such as diversity quotas will also help encourage appropriate efforts to recruit more diverse patients. It is important to consider, however, that close ethical committees and checks should be represented by actual examples. Based on historical ethical misconduct, sites expected to maintain diversity quotas should be carefully monitored. Cutting corners at sites to hit diversity quotas will be detrimental and dangerous. Unethical recruitment to reach diversity quotas should be quickly rectified by FDA oversight, with built-in safety nets for anonymous reporting from both patients and staff.

Conclusion 

Research enrollees should directly reflect the community we serve. Medical interventions should aim to be effective across all demographics and should from the beginning represent these populations even if it takes longer to enroll participants or if costs are higher in the study start-ups. We pay the price for not having an inclusive study when ongoing disparities in clinical care outcomes have changed only marginally over decades. Gaps or success in medical intervention effectiveness cannot be captured if tested on a select population. Sponsors and research institutions must prioritize increasing diversity in study samples. In the example of our multicenter research facility, quantification of patient demographics clarified pitfalls for increasing diversity such as improving multilingual resources for patients. This is the beginning of what will be a systematic process for us to hold each other accountable going forward. In order for us to see changes in who is represented in trials, who is having favorable outcomes in clinical care, it’s critical that we change our processes to be truly representative for all.



Maya Palmer, is a clinical research associate at the Minneapolis Heart Institute Foundation and is entering medical school this fall.

Courtney Jordan Baechler, MD, MS, is the medical director for health equity and health promotion at the Minneapolis Heart Institute Foundation.