No experimentation without representation
Looking at the importance of diversity in clinical trials
In the wake of Derek Chauvin’s trial last week and America’s continued struggle for racial equality, I decided to write about one manifestation of existing inequities in the context of healthcare — the underrepresentation of minority groups in clinical research.
Despite the increasingly diverse makeup of America’s population, minorities — specifically Black and Hispanic Americans — are chronically underrepresented in clinical trials. An FDA report from 2011 found that African Americans make up 12% of the broader population but only 5% of clinical trial participants, while Hispanics comprise 16% of the population but only 1% of clinical trial participants.
Diversity in research is key for safety, access, and equity
Proper representation in clinical trials is an important issue for multiple reasons. First, since clinical trials are used to evaluate the safety and efficacy of new interventions, underrepresentation can limit the amount of evidence for certain racial groups. This can pose a serious safety risk to patients in cases where the response to a new treatment varies depending on race, which is the case for about 1 in every 5 new FDA drug approvals.
Second, underrepresentation directly hinders access to quality care. At the most basic level, clinical trials provide an opportunity for patients to gain access to promising, investigational treatments which in some cases might be life-saving. Lower participation among minorities means fewer chances for them to benefit. There are also downstream consequences — the underrepresentation observed during clinical development may persist even after approval with real-world use. If there is a smaller evidence base to support the safety and efficacy of a treatment in minorities, physicians may be more reluctant to prescribe it for those populations and patients may be more hesitant to adopt it.
A lack of diversity in clinical trials is not only a problem in and of itself, but also because it reinforces existing health disparities across racial groups. These disparities have been well-established — compared to white Americans, Black Americans are 66% more likely to report having fair or poor health, Hispanics are almost three times as likely to be uninsured, and Asian Americans are 40% more likely to be diagnosed with diabetes. As such, there’s an argument to be made that rectifying the status quo will require overcompensating for current inequities, and that with clinical research, our goal should actually be the overrepresentation of disadvantaged populations.
Mistrust, resource constraints, and internal biases can get in the way
Proper representation in clinical trials has proved to be a challenge because there are several layers of barriers that get in the way. One commonly studied obstacle is mistrust of the clinical trial process; although this concern is shared across minority groups, the exact nature of the concern can vary between races. As an example, African Americans are often distrustful because of their belief that research will benefit white Americans and not people of color (an entirely justifiable concern given the deplorable history of human experimentation in the US). Asian Americans, on the other hand, tend to be more concerned with the informed consent process and the fear of relinquishing their rights.
Challenges can arise at an even simpler level, in the form of language barriers for Asian and Hispanic Americans. Rightly so, the informed consent process for clinical research must be conducted in a language the patient can understand. This means, however, that enrolling patients from immigrant populations often requires interpreters, who can be hard to come by. A recent study published in Health Affairs found that nearly one-third of US hospitals fail to offer interpreters to patients even for routine medical care, let alone clinical research.
Unsurprisingly, time and resource constraints also hinder minority participation in research. Clinical trials require extra doctor visits, and being able to carve out time for these visits may prove challenging, especially for single parents or patients working multiple jobs. There’s also the financial cost of being involved, including the cost of transportation to and from each visit, which can easily exceed the compensation patients receive from the study.
Barriers exist on the other side of the research picture as well, with study investigators and sponsors. Two physician surveys conducted by the Tufts Center for the Study of Drug Development found that the racial makeup of study investigators has an impact on the diversity of enrolled patients, highlighting the importance of ensuring representation among investigators. Physicians are also subject to their own implicit biases, which can influence their clinical decision making and specifically, who they decide to enroll into a study.
Lastly, study sponsors can design studies which unnecessarily exclude patients with certain co-morbidities. They are, in fact, incentivized to do so — a more homogeneous patient population can lead to less noise in the data and more power to draw conclusions. However, given existing health disparities across racial groups, these exclusion criteria can end up disproportionately filtering out minorities. One such example is with cancer studies that often exclude patients with HIV, a condition that is 8 times more prevalent in Black Americans than whites. Despite cancer being a leading cause of mortality in HIV patients, a review by the American Society of Clinical Oncology (ASCO) found that only 1.7% of oncology trials allowed enrollment of these patients.
Improving representation will require a multi-faceted response
In response to this broad challenge, regulators have ramped up their efforts to increase diversity in clinical research over the past 25 years. In 1999, HHS issued a report recommending the capture of race and ethnicity data in government sponsored research, and in 2001, the NIH adopted a policy requiring women and minorities to be included as research participants unless their exclusion was justified by federal guidelines. In 2005, the FDA released its first guidance on the topic, urging sponsors to collect race and ethnicity data in investigational trials and include them in submissions.
Despite these initiatives, disparities remain, and this is punctuated by significant variability in the degree of representation across different studies and products. In Nov 2020, the FDA went a step further by providing prescriptive recommendations on how sponsors can improve diversity. One thing that stands out in the guidance is how the agency highlights the importance of avoiding unnecessary exclusion criteria and calls out the role that precedence plays in maintaining these criteria:
FDA also recognizes that some eligibility criteria have become commonly accepted over time or used as a template across trials, sometimes excluding certain populations from trials without strong clinical or scientific justification.
Medical journals also have a role to play in improving clinical trial diversity. Given their importance in disseminating research findings, journals are uniquely positioned to influence study design choices and impose guidelines on the information ultimately reported. As a first step, journals should require reporting of race and ethnicity data for all in-human trials. An even more impactful development would be if race and ethnicity data were added to the CONSORT Checklist, a minimum set of reporting criteria for randomized controlled trials. Since numerous journals adhere to CONSORT guidelines, an addition to the checklist would be passed through to countless journal requirements as well.
Earlier we talked about how transportation costs and logistics are an important barrier for clinical trial participation. It’s also a hindrance for healthcare in general, as an estimated 3.6 million Americans don’t receive proper medical care due to transportation issues each year. Various initiatives have been started to tackle the latter, including an exciting partnership between Lyft and Blue Cross Blue Shield to integrate ride-sharing services with care delivery. Since research visits are typically not covered by insurance, new programs will need to be launched to address the former — in the case of Lyft, perhaps a program that allows them to partner with study sponsors directly.
Another area where technology can help is with overcoming language barriers. The past year has demonstrated the viability and convenience of telemedicine, and one particular application which warrants more adoption is with interpreter services. For clinical sites that don’t have in-house interpreters, having access to on-demand interpreters would make it significantly easier to recruit non-English speaking patients for trials. Such a platform could also have the advantage of supporting a much broader range of languages than an in-house team.
In many ways, the race to develop a COVID-19 vaccine over the last year underlined the importance of diversity in clinical trials. The vaccines are intended for the general population, so understanding their safety and efficacy in all racial groups is paramount. The pandemic’s disproportionate impact on minorities highlighted the importance of ensuring an equitable development process and rollout. And ultimately, the success of the vaccines will rely on widespread adoption, which requires a high degree of trust and understanding of the data.
Fortunately, the phase 3 trials for the Pfizer-BioNTech and Moderna vaccines were significantly more representative than the average study — almost 10% of participants were African American (compared to 12% in the broader population), and Hispanic Americans were overrepresented. It’s no coincidence, though, that this was attained amidst a national spotlight being cast on race relations and loud calls for diversity during development. As with the COVID vaccine trials, the broader goal of diversity throughout clinical research will only be achieved through a combination of awareness, activism, and perseverance.