Improving diversity in study participation: Patient perspectives on barriers, racial differences and the role of communities

Abstract Introduction The lack of racial/ethnic diversity in research potentially limits the generalizability of findings to a broader population, highlighting the need for greater diversity and inclusion in clinical research. Qualitative research (i.e., focus groups) was conducted to identify (i) the potential motivators and barriers to study participation across different races and ethnicities; (ii) preferred delivery of education and information to support healthcare decision‐making and the role of the community. Methods Patient focus groups were conducted with 26 participants from the sponsor's Patient Engagement Research Councils selected through subjective sampling. Recruitment prioritized adequate representation across different race/ethnic groups. Participation was voluntary and participants underwent a confidential interview process before selection. Narrative analysis was used to identify themes and draw insights from interactions. Experienced research specialists identified emerging concepts, and these were tested against new observations. The frequency of each concept was examined to understand its importance. Results Based on self‐selected race/ethnicity, participants were divided into five focus groups (Groups: African American/Black: 2; Hispanic/Latino, Asian American, and white: 1 each) and were asked to share their experiences/opinions regarding the stated objectives. Barriers to study participation included: limited awareness of opportunities to participate in research, fears about changes in standard therapy, breaking cultural norms/stigma, religion‐related concerns and mistrust of clinical research. Participants identified the importance of transparency by pharmaceutical companies and other entities to build trust and partnership and cited key roles that communities can play. The perceptions of the African American group regarding diversity/inclusion in research studies appeared to be different from other groups; a lack of trust in healthcare providers, concerns about historical instances of research abuse and the importance of prayer were cited. Conclusion This study provided insights into barriers to study participation, and also highlighted the need for pharmaceutical companies and other entities to authentically engage in strategies that build trust within communities to enhance recruitment among diverse populations. Patient or Public Contribution The data collected in the present study was provided by the participants in the focus groups.


| INTRODUCTION
Carefully designed randomized clinical trials, along with complementary findings from observational and real-world evidence studies, can demonstrate the efficacy, safety and treatment outcomes of therapies in a certain patient population. [1][2][3] Ideally, participants in research studies should closely mimic the demographic diversity of the general population or the prevalence of a specific disease. [4][5][6][7] The responses to a particular medication might differ among population subgroups based on several factors, such as age, sex, genetic profile and ethnicity. [8][9][10] Although race and ethnicity are interchangeably used, 'race' refers to a person's physical characteristics, and 'ethnicity' includes culture, beliefs and language in addition to the racial ancestry. A single geographical area can have few basic races, but a number of ethnic groups. 8 Genetic variability in ancestral populations, along with shared environmental factors, is known to alter the response to drug therapy among different ethnic subgroups. 9 Therefore, the inclusion of different races and ethnicities in research is critical to enable the development of better treatments and better ways to fight diseases that often disproportionately impact diverse communities.
Recent studies have demonstrated that study samples do not adequately represent the race and ethnicity of those affected by various diseases. 7,[11][12][13] For instance, more than 40% of cancer clinical trials in the United States do not reflect the incidence rates among diverse racial/ethnic groups, resulting in an over-representation of whites. 11 Under-representation of non-white participants has been widely reported in US-based clinical and observational studies, including those for type I diabetes, 14 type II diabetes, 15 Duchenne muscular dystrophy 16 and rheumatoid arthritis. 17 Importantly, the 2020 Census reported a decrease in the white alone population by 8.6%, a 276% increase in the multiracial (two or more races) population and 23% and 88.7% growth in the Hispanic/Latino and Black/African American populations, respectively, when compared with the 2010 Census. 18 By 2045, it is projected that less than 50% of the US population will be comprised of non-Hispanic whites. 19 Study findings with homogeneous samples (e.g., those based on age, gender, or race/ethnicity) are less generalizable to the broader population. 7,20 While medical care and scientific knowledge have advanced significantly over the past decade, many studies continue to have a majority of white participants. 21 African American, Latino and other diverse racial/ethnic groups are often less represented in these studies and can have worse health outcomes as a result. 7,22,23 Studies featuring an adequate representation of diverse racial/ethnic groups therefore are critical to reducing healthcare disparities. [24][25][26] Studies across multiple therapeutic areas have reviewed this lack of racial/ethnic diversity and have identified which psychosocial and cultural attributes act as barriers and motivators to study participation. 7,[27][28][29][30][31][32][33] For example, having a lower income, poor education or a lack of research study resources in a native, culturally appropriate language may be barriers to participation. 30,34 This study sought to understand what other factors may exist to motivate or prevent someone from participating in research.
There is a growing focus on community-based participatory research (CBPR) to achieve diversity in the study population. 35 CBPR follows the bottom-up approach (starting with community members to identify salient issues important to a particular population) instead of the traditional top-down approach (where external organizations identify an agenda that may not be reflective of the needs of a community). This approach maximizes community participation and patient retention as it considers community members as invested partners in the intervention and outcomes. 36,37 Research supports that the CBPR approach is effective in recruiting, retaining and improving behavioural and health-related outcomes in disadvantaged communities. [38][39][40][41] US Food and Drug Administration (FDA) recommends a patientfocused drug development approach to ensure that the views, needs, preferences and interactions of patients are captured and meaningfully integrated throughout the lifecycle of a medical product. 42 Although it is crucial to embrace patient-focused research, the existing literature reveals an unmet need in this area. 33,43 Qualitative research provides an insight into the experiences of participants and enables researchers to understand rich explanations and descriptions in local contexts. 44 This type of research can help capture underrepresented patient experiences and may enhance the engagement of diverse patient groups in clinical research. 45 In this study, focus groups with a diverse group of patients across multiple disease areas were conducted to better understand potential motivators and barriers to study participation across different races and ethnic groups. This study also assessed the preferred delivery of education and information to support healthcare decision-making and the role of the community. The PERC program constitutes a diverse group of participants who suffer from chronic disease, are self-aware of their condition and provide their input through a structured series of specific research activities.
Participants were recruited through outreach to patient advocacy organizations, online advertising websites, social media, and physician referrals. Recruitment targeted 'everyday' participants with a variety of healthcare experiences; some participants were very involved in healthcare decision-making, others less. Participants were not exclusively experts or advocates in their disease area. The screening and recruitment process for this study allowed participants to self-identify race and ethnicity, and these identifiers were taken into account when designing and analyzing research. Subjective sampling was used, and demographics such as race/ethnicity were often prioritized in the recruitment process to ensure adequate representation. Individuals were subjected to a confidential and thorough screening and interview process before selection by a series of questions, assessing prior participation in any research study, self-identification in terms of race/ethnicity, and the highest level of education received. Of patients who expressed interest, on average, approximately 64% were interviewed; 57% were eligible to participate and 42% were invited to become PERC participants. Council members were compensated for their time participating in research activities and each research opportunity was voluntary.
At the time of this study, the PERC program consisted of 108 participants, 30% of which self-identified as African American/Black, 8% as Hispanic/Latino and 7% as Asian American.

| Procedures and study groups
Research questions were developed to address the key study objectives before conducting the focus groups. The questions were formulated by a research specialist who possessed experience in culturally appropriate research methods and patient literacy and were structured around validated health behaviour principles that were used by the research team across similar studies. A discussion guide was developed by a research specialist with additional review by senior researchers and sponsor representatives. Participants involved in these sessions had one of the following self-reported disease conditions: peripheral arterial disease, venous thromboembolism, cardiovascular disease, inflammatory bowel disease, ankylosing spondylitis, psoriatic arthritis, prostate cancer, multiple sclerosis or pulmonary arterial hypertension. Information from five focus groups was gathered and analysed; each discussion was conducted for 90 min on different dates.  Table 1). Nine participants had previous experience participating in clinical research.
Participants were informed that no treatments would be provided, and they could withdraw at any time. Additionally, a consent and release form was signed by the participants that communicated confidentiality and Health Insurance Portability and Accountability Act (HIPAA)-compliant practices. All data were deidentified; thus, no ethics board review was required. The purpose of this study was to collect personal perspectives and qualitative insights from the participants. The study was also conducted in accordance with the Helsinki Declaration of 1964 and its later amendments. All sessions were conducted virtually, and participants joined from their homes.
For each focus group, participants were first introduced to the types of clinical research to provide the foundation for discussion and the basic role of the participant ( Figure 1). Participants were then asked to share their experiences and opinions regarding participation in research studies, factors that they considered as motivators or barriers to study participation, and sources of information with a focus on understanding the factors enhancing trustworthiness.

| Data management and analysis
All discussions lasted 2 h each and were audio-recorded and transcribed. Following the sessions, recordings for transcription were submitted through a transcription firm experienced in transcribing medical market research. 46 The research team utilized conceptualization and conversational (narrative) analysis to identify concepts in the data from this study. A senior research specialist with experience in narrative analysis and drawing insights from interactions with diverse populations directly observed the data (during data collection, and subsequently, through transcript analysis) and identified concepts, with support from a research associate. Using an iterative process, analytic insights were tested against new observations; concepts were refined as those continued to emerge from the data.
This process continued until all the data were analysed and the frequency of each concept had been examined to understand its importance. The resultant concepts were then reviewed by additional SHEA ET AL.

| Barriers to study participation
In our discussion, fear was found to be the primary obstacle to research study involvement. Specifically, participants feared: side effects (short-term or long-term) due to untested medications;  -African American/Black … Generally speaking, white men tend to make me a little nervous because I feel spoken down to at times.

-Hispanic/Latino
If the doctor were a Christian or not, it would have absolutely no effect on me.
- This study provides valuable insights for increasing diversity and inclusion in research. While qualitative research is a critical tool to gather the patient voice, inherent limitations exist. There is a tendency for the emergence of socially acceptable opinions and bias development within the group due to the dominance effect and groupthink.
Small sample sizes cause the research to be mainly exploratory in nature, as the number of participants in each group and the duration of focus group discussions do not allow testing of the data saturation point. 53 Prior participation in research studies by some participants might indicate increased openness to clinical research, knowledge and awareness that could shape their perspectives either positively or negatively. Further, responses were not coded, which could have resulted in an interpretation bias. Therefore, an opportunity exists for future research to address these limitations around data collection and analysis. The drafting of this article was planned only after data analysation and was intended for hypothesis generation, not confirmation. In addition, the individuals who agreed to participate in this study were likely to be relatively more health-engaged or actively aware of their disease, which could potentially limit the generalizability of these results to broader patient populations.

| CONCLUSION
Our study generated significant qualitative patient perspectives regarding potential barriers to participation in research studies, highlighting racial and ethnic differences, and the role of engaging with communities to help overcome these barriers to achieve diversity and inclusion in studies. These focus groups highlight the importance of programs that foster bidirectional collaboration between pharmaceutical or other entity sponsors and community members. The findings and recommendations presented in this study could be further integrated into the research study recruitment framework and processes to support an engagement strategy, with an enhanced focus on the intersectionality of the 'whole patient' to build trust and partnership with patients and their communities.

AUTHOR CONTRIBUTIONS
All authors contributed to data analysis, drafting or revising the article, gave final approval of the version to be published, and agree to be accountable for all aspects of the work.