Let’s talk about bias. According to a recent study published in Nature, clinical trial participation has a lot of room for improvement when it comes to unbiased representation across national demographics.
In what may be the first-of-a-kind analysis of clinical trial participation, researchers from the University of Hawaii looked at more than a decade of clinical trial data to determine how well demographic data on those trials reflect the broader US population. The results showed “significant yet varied” disparities and small improvement over time, managing to confirm well-known realities and dispel a few myths of clinical trial diversity at the same time. Read on for a discussion of results and unanswered questions around these key findings:
Data insights from US clinical trials conducted between 2008 and 2019 included:
- Trial demographics did not accurately represent national demographics
- Failure to report race and ethnicity data was prevalent
- Females, Hispanics, American Indians and Alaskan Natives (AIAN), and Asians were underrepresented, while Blacks, Native Hawaiians, and Pacific Islanders were over-represented
- Blacks and Native Hawaiian and Pacific Islanders (NHPI) were found to be overrepresented overall in most clinical trials reviewed in this analysis
The biggest surprises?
Results suggest recruitment patterns often paralleled the baseline demographics of the particular illness but not in every disease category. Across the board in this analysis, there’s an over-representation of Blacks, Native Hawaiians, and Pacific Islanders. A closer look at some of the therapeutic areas reveals that Blacks are still under-represented in some disease areas, such as oncology, outside of prostate, breast, and gynecologic indications. To a larger extent, Blacks are also underrepresented in otologic and dermatological disorders, ones that tend to be overrepresented by Native Hawaiian and Pacific Islanders (NHPI), multi-racial groups for otology and NHPI, and whites for dermatologic disorders. These findings help debunk widely held assumptions about low participation in clinical research from Black communities due to a long history of medical mistrust. It also suggests that existing efforts around outreach, education, and trial support designed to help increase participation among Black communities have demonstrated initial success in indications such as diabetes and kidney disease, but additional exploration and refinement of efforts are needed. Clinical trial representation is a nuanced issue. Further indication-level and other meta-analyses of trials will help determine a clearer and more specific picture of representation, particularly in disease areas that disproportionately impact Black populations, such as Alzheimer’s disease and various cardiovascular disorders beyond heart disease and hypertension.
Now we know the extent of the problem. Sort of…
This review may be the start of a baseline, but many gaps remain. Overall, the analysis included 55% of the whole data set (2977 out of 5388 clinical trials). 36% of these trials reported ethnicity, and 53% reported race, representing major gaps in data reporting in spite of expectations and pressure from regulators and industry initiatives.
- The analysis did not include multi-center studies with international locations due to the inability to extract US sites from these global trials.
- The gender proportion for statistical analysis did not include transgender participants because of a lack of participants.
Awareness is the first step required for change, and knowledge of the data gap that needs to be filled can serve as an urgent call to action among researchers to better address health disparities. As public demand for health equity continues to amplify, increased collection of data may naturally follow as part of that trend. Discussion in the report acknowledged that “Some of the trends observed may have represented improved reporting rather than changes in demographic representation over the years.”
Where to go from here
Patient recruitment alone is one of the biggest challenges in clinical research. It’s multiplied with these additional layers of complexity: finding diverse trial populations who mirror the disease population and are also willing to participate in research. The application of systematic approaches to patient-friendly research, participation incentives, and behavioral sciences to attract broader trial populations is still in its infancy, but the biopharma industry is certainly beyond the pilot stage.
There’s also a growing interest in socioeconomic status and clinical trial participation, more specifically, how education and income affect a patient’s ability to participate in research. With a dearth of data on this topic and increasingly powerful AI applications in clinical research, this is an untapped area with urgent needs.
Broader ethnic categories – .4% (nearly 2500 participants) reported their ethnicity as unknown. 1% (more than 7400 participants) reported their race as unknown, which raises additional questions. How do participants want to report their race and ethnicity? It’s unclear how they were asked and what choices were made available. There’s a growing body of work around inclusive demographic questions that can be applied in clinical research.
First step: explore demographic improvements over time in women and Hispanics for potential application in other underrepresented groups.
For women, representation improved significantly, from underrepresentation (48% in 2008) to overrepresentation (64% in 2018). Hispanic participation followed a similar pattern, from 11.6% in 2006 to 18.7% for trials that began in 2016. While this trend can possibly be correlated to the rise of the educated and empowered digital patient, it’s not the case universally. With some improvement over time in certain types of trials, namely Tier 2, which is comprised of trials that tended to include a combination of these criteria: multi-center, 200 or more participants and reporting of race and/or ethnicity), AIAN and Asians remained underrepresented over the duration, with Asian populations as the most underrepresented ethnic group. The disparities here can possibly be attributed to many factors, including trust over historical medical misconduct, cultural behavior, financial and logistical barriers, and language restrictions. As a research community, we can build upon initiatives that demonstrate improvement in clinical trial demographics over time, but we must continue to fund primary research and exploration to inform practical solutions that go beyond what’s already been tried to help close the research gap.
Author: Matthew Maxwell, Chief Marketing Officer
