Opportunities to Increase Science of Diversity and Inclusion in Clinical Trials: Equity and a Lack of a Control

ABSTRACT

The United States witnessed a nearly 4‐fold increase in personal health care expenditures between 1980 and 2010. Despite innovations and obvious benefits to health, participants enrolled in clinical trials still do not accurately represent the racial and ethnic composition of patients nationally or globally. This lack of diversity in cohorts limits the generalizability and significance of results among all populations and has deep repercussions for patient equity. To advance diversity in clinical trials, robust evidence for the most effective strategies for recruitment of diverse participants is needed. A major limitation of previous literature on clinical trial diversity is the lack of control or comparator groups for different strategies. To date, interventions have focused primarily on (1) community‐based interventions, (2) institutional practices, and (3) digital health systems. This review article outlines prior intervention strategies across these 3 categories and considers health policy and ethical incentives for substantiation before US Food and Drug Administration approval. There are no current studies that comprehensively compare these interventions against one another. The American Heart Association Strategically Focused Research Network on the Science of Diversity in Clinical Trials represents a multicenter, collaborative network between Stanford School of Medicine and Morehouse School of Medicine created to understand the barriers to diversity in clinical trials by contemporaneous head‐to‐head interventional strategies accessing digital, institutional, and community‐based recruitment strategies to produce informed recruitment strategies targeted to improve underrepresented patient representation in clinical trials.

Nonstandard Abbreviations and Acronyms

CBPR

community‐based participatory research

CNP

Community Networks Program

DIVERSE

Diversity & Inclusion in cardioVascular trials through Enrollment and education Resulting in Sustainable Equity

FDA

US Food and Drug Administration

HTA

Health Technology Assessment

As new health care technologies, such as advances in surgical techniques, novel medical devices, and pharmaceutical drug innovations develop, clinical trials are frequently used to examine their effectiveness and provide evidence to support their adoption. However, participants enrolled in clinical trials still do not accurately represent the racial and ethnic composition of patients nationally or globally. ¹ , ²

The lack of diversity in clinical trial cohorts limits the generalizability and significance of clinical trial results among all populations. Population‐specific variable drug reactions may occur in response to the same treatment. Participant representation across diverse age, gender, race and ethnicity, sexuality, and disability statuses is necessary to confirm the safety and effectiveness of treatments. ³ Researchers have begun to address the need for diversity by looking at historical data on participant enrollment and formulating recruitment plans to target historically underrepresented populations. ⁴ Nonetheless, a 2021 analysis found that very few trials achieved levels at, or above, nationally representative proportions for many race and ethnic demographics based on the US Census. ⁵

A review of 379 National Institutes of Health–funded clinical trials found that only 48% of the studies provided complete participant racial/ethnic information. ² Between 2010 and 2020, the percentage of individuals who did not select White alone on the US Census increased from 27.6% to 38.4%. ⁶ Although Census trends highlight how non‐White, underrepresented racial/ethnic groups in the United States are steadily growing in size, the US Food and Drug Administration (FDA) reported that 75% of participants enrolled in clinical trials approved in 2020 were categorized as White individuals, indicating how clinical trial participant pools fail to reflect the public demographic. ⁷ Additionally, these shortcomings in clinical trial recruitment also extend to lines of gender: in a review of 862 652 participants across 740 trials, only 38.2% were categorized as women. Despite the 2020 FDA‐published industry guidance on enhancing diversity, certain demographics are still underrepresented in clinical research trials. ⁸ Multiple publications have suggested strategies for increasing participant diversity, but there are little data that rigorously compare these strategies. ⁹

In December 2022, the FDA released the Food and Drug Omnibus Reform Act, with new mandates regulating the recruitment and retention of diverse participants in clinical trials. Outlining changes to reform clinical trials, the mandates focus attention on decentralized clinical trial development, increasing diversity among the clinical trial population, and efficacy of digital health technology in clinical trial recruitment interventions. To that end, the FDA intends to require clinical trial sponsors to submit diversity action plans outlining the intent and concomitant actions intended to recruit diverse clinical trial patient populations. Each sponsor's diversity action plan must disclose an enrollment goal, which is informed by demographic information concerning the patient population of the disease or condition, comorbid conditions of the population of interest, and barriers to enrollment of diverse populations such as population size, geographic location, and socioeconomic status. The diversity action plan should outline sponsor interventions to meet the enrollment goal, such as demographic‐specific outreach and enrollment strategies, study‐site selection, clinical study inclusion and exclusion practices, and any diversity training for study personnel. ¹⁰

As patients, payers, and purchasers struggle to grapple with the rising costs of care within a new landscape of treatment, the question of inclusion and the value of care among all patient populations must be addressed. ¹⁰ To advance diversity in clinical trials, a robust evidence base for the most effective strategies for recruiting diverse participants is needed. To that end, the American Heart Association created the Strategically Focused Research Network on the Science of Diversity in Clinical Trials to focus attention on the critical issue of patient representation (Figure 1). ¹¹ The Science of Diversity in Clinical Trials represents a multicenter, collaborative network created to understand the barriers to diversity in clinical trials (Figure 2). To add dimension to the faces and communities that we hope to reach, the collaborative work of investigators from Stanford School of Medicine and Morehouse School of Medicine created the Diversity & Inclusion in cardioVascular trials through Enrollment and education Resulting in Sustainable Equity (DIVERSE) Network Center program (Table 1). ¹² This program has designed contemporaneous head‐to‐head interventional strategies assessing digital, institutional, and community‐based recruitment strategies to better understand barriers surrounding underrepresented patient population recruitment and retention in clinical trials and establish general best practices and recommendations grounded by evidence (Figure 3). The DIVERSE Network Center is one of the 6 centers that make up the Strategically Focused Research Network on Science of Diversity in Clinical Trials; the function of the Science of Diversity in Clinical Trials in establishing a collaborative foothold and centralized body of evidence is the first step in challenging the historical and modern barriers surrounding the participation of underrepresented individuals in clinical trials and health equity gaps.

Figure 1. Strategically Focused Research Network: Science of Diversity in Clinical Trials. ¹⁰¹ .

Figure 2. Interpersonal, institutional, and political relationships that overlay research recruitment and retention.

DIVERSE indicates Diversity & Inclusion in cardioVascular trials through Enrollment and education Resulting in Sustainable Equity.

Table 1.

TIDE and TOTAL Study Design (DIVERSE SFRN SDCT) ¹⁰²

Type	Title	Specific groups addressed	Outcome measure	Enhancement strategy	Effect size
Trials with Inclusion, Diversity, & Equity	TIDE	Underrepresented populations	Clinical trials accrual	Patient‐level intervention strategies using community presenters, digital health technologies, and UberHeatlh	TBD
Trial of Sites to Increase Diversity in Clinical Trials	TOTAL	Underrepresented populations	Clinical trials accrual	Site level intervention strategies using peer navigators, Google advertisements, population‐based registries	TBD

SDCT indicates Science of Diversity in Clinical Trials; SFRN, Strategically Focused Research Network; TBD, to be decided; TIDE, Trials with Inclusion, Diversity, & Equity; and TOTAL, Trial of Sites to Increase Diversity in Clinical Trials.

Figure 3. TIDE and TOTAL intervention network graphic.

AI indicates artificial intelligence; CA, community ambassador/peer navigator; DIVERSE, Diversity & Inclusion in cardioVascular trials through Enrollment and education Resulting in Sustainable Equity; DM, digital media (Google ads & Meta ads); R, registries; TIDE, Trials with Inclusion, Diversity, & Equity; TOTAL, Trial of Sites to Increase Diversity in Clinical Trials; and UC, usual care.

The strategies presented in our review represent several evidence‐based recruitment and intervention approaches that aim to improve diversity in clinical trial patient populations. Examples of interventions include (1) community‐based interventions, (2) institutional policy, and (3) digital social media and web platforms such as Facebook and Google focused on inter‐ and intrapersonal communications. Individually these interventions have shown promise to increase underrepresented patient population representation within clinical trials. However, there are no current studies that comprehensively compare these interventions against one another. A study of the comparative effectiveness of each of these interventions that is informed by different patient cultural and socioeconomic contexts can improve recruitment efforts targeted to specific underrepresented populations.

Community‐Based Interventions: Community‐Based Participatory Research

Engagement of patients in the communities where they live is one of the most effective strategies to increase diversity in clinical trials. The limited national and regional availability of clinical trials for which a patient is eligible and the probability that a patient participates are issues organized around patient location. As such, the infrastructure and community relationships that contextualize patient outreach and involvement in clinical research are critical in reaching underrepresented participants. ¹³ , ¹⁴ , ¹⁵ Establishing relationships within communities is at the center of community‐based participatory research (CBPR), in which the community partners with community‐based organizations to develop and implement clinical research within the community. The framework incorporates strategies targeted to community‐based (engaging the community) and place‐based (focusing on a geographic area) interventions. ⁴ , ¹⁴ , ¹⁵ , ¹⁶ This partnership is critical for disseminating information that promotes community knowledge and community participation within research incentives.

Community‐based models describe how effective engagement and collaborations between academic institutions and the surrounding communities focused on improving health outcomes are based on trust, consistency, sustainability, and cultural context. ¹⁷ , ¹⁸ , ¹⁹ The Morehouse Model is a successful framework that focuses on the nuances of community engagement and the importance of community members as facilitators and advocates. ¹⁸ By leveraging established relationships and the valued spaces around which communities gather, research interventions create a tenable route to better shared outcomes. In the same vein as CBPR, Community Networks Program centers function within communities as an intermediary between researchers and patient communities to increase community engagement in an iterative research and outcome development process to decrease health disparities. The differences in environmental contexts and socioeconomic barriers within each population and collaborator priorities are important contexts within which Community Networks Programs exist. In a comparative analysis between 2 Community Networks Programs, Arroyo‐Johnson et al surveyed 2 geographically disparate, participant programs to determine common operational conditions and processes at the program and project level that led to effective CBPR success. Despite differences in operational goals at the program and project level, quality principles continued to stimulate effective community engagement and CBPR partnerships, increasing patient engagement and improving health literacy. ¹⁶

Wright et al also underscore the importance of culturally competent engagement among the lesbian, gay, bisexual, transgender, queer/questioning (LGBTQ) community in CBPR success. In their study, they underscore the potential for CBPR to engender effective change through establishing key partnerships within the LGBTQ community and grounded in the core value RESPECT: “Realize Every Single Person's Experience Carries Truth; Realize Every Single Person's Experience Can Teach. We honor and value the diversity we bring.” By doing so, the group set the stage for honest and open dialogue wherein each partner was both expert and student, open to ask questions without shame and to increase the collective knowledge of the group. This honest and open dialogue was needed for project‐level success in addressing the needs of aging LGBTQ seniors, who often experience worse outcomes than their heterosexual counterparts. ¹⁷

Understanding the relationship between the economic, social, and physical relationships that impact the health of patients, and working to incorporate stakeholders from the community in community benefit initiatives will ensure the sustainability of these projects. Contemporary CBPR projects centered on the American Heart Association Life's Simple 7 framework, which delineates 7 modifiable health behaviors and clinical factors to promote cardiovascular health, have shown preliminary effectiveness in improving individual and composite Life's Simple 7 indicators among diverse patient populations. ¹⁸ , ¹⁹ , ²⁰ , ²¹ In a systematic review of peer‐reviewed literature examining CBPR among African Americans, intervention strategies that focused on dietary change or physical activity with coach‐led community groups and faith‐based interventions demonstrated effectiveness in reducing blood pressure, with an average decrease of 8 mm Hg in systolic blood pressure. ¹⁸ Additionally, 2 other studies which emphasized community engagement utilizing pharmacist‐led blood pressure control intervention in Black barbershops or therapeutic lifestyle changes in Black churches noted a 27‐ and 6‐mm Hg reduction in blood pressure at the community level, respectively. ²¹ , ²²

Maximizing the capacity of community engagement strategies will allow academic and health care organizations to address and prioritize lifestyle risk factors and social determinants of health specific to the communities they serve. In addition, combining internal resources horizontally across hospital systems increases the investment capital for infrastructure development and produces measurable results where health inequities are most concentrated. This allows interventions to transition from small‐scale, isolated, and short‐term events into institutional and transformational health care delivery reform programs. ²³ , ²⁴ , ²⁵ Community engagement and community benefit coordination across hospital systems in coordination with local public health departments can compound and leverage analytical, administrative, managerial, and public health resources to produce equitable gains for all participants involved. This also allows for the provision of population health services on a larger and more integrated scale, thus decreasing the duplication of services between health care structures. ²⁶ , ²⁷ , ²⁸

Promising Institutional/Policy Interventions

Institutional infrastructure development, coordination, tracking, management, and quality improvement together often represent an interventional strategy that is easily accepted, but difficult to produce and sustain. Institutional policies not only need to be functional to appropriately coordinate and track patient activities, but they must also be robust and pliable enough that they can respond to shortfalls, acting collectively to respond to anticipated and unanticipated challenges. ²⁹ , ³⁰

Institutional Changes: Community Engagement and Interpersonal Relationships

Through community and organizational collaboration, researchers have highlighted the importance of intentional materials for recruiting underrepresented populations, such as Black, Indigenous, people of color and/or women. Through these community‐based methods, 1 notable trial was able to recruit a participant population of 67% women, with a majority identifying as women of color. ¹⁵ The development of institutional‐level policies that direct interventions toward increased representation orient research organizations and researchers to the future; however, the results of these efforts will depend upon effective institutional collaboration and patient–researcher understanding and communication. ¹⁴ , ¹⁵ , ³⁰ , ³¹

Building trust within a community requires that an institution conduct its research in an open and honest manner with research representatives sensitive to the concerns of patients and capable of discussing research ethics and the historical abuses of racial and ethnic minority groups in research. ⁴ , ³¹ , ³² , ³³ , ³⁴ Establishing a firm ground of mutual understanding and collaboration will establish a relationship grounded in mutual benefit. For example, to increase racial and ethnic minority group enrollment, the National Cancer Institute's Minority‐Based Community Clinical Oncology Program increased outreach efforts to health care institutions that serve many underrepresented individuals with cancer. ³⁵ Similar to other interventions, its success was largely dependent upon system‐ and hospital‐level factors such as increased access to open trials that matched the clinical characteristics of the patient population of interest and development of local physician and cancer advocacy groups, which in turn increased willingness to enroll or refer underrepresented patients to clinical trials. However, the success of the intervention was limited by inadequate funding for outreach activities, limited system support for staffing and education of sexual, gender, racial, and ethnic minority investigators, and appropriate funding for protocol‐related drugs and related services for uninsured patients. ³⁵

At the patient–physician level, racial and ethnic differences in communication present themselves in the form of patient‐related communication behaviors, communication behaviors of physicians, and communication behaviors centered on clinical trial information itself. For example, in a study utilizing video‐recorded interactions between discordant physician–patient racial interactions in an oncology clinic, linguistic analysis revealed that discussions with African American patients were significantly shorter than those for White patients, and clinical trials were less frequently mentioned. ³⁵ , ³⁶  When clinical trials were mentioned, oncologists spent significantly less time discussing the trial, less time discussing risks and benefits associated with trial participation, less time discussing alternatives to trial participation, and more time focusing on the voluntary nature of the trial compared with White patients.Because institutional strategies require goal‐oriented alignment at each level of patient engagement, the physician represents the most central and proximal influence to improve patient enrollment and initiate discussions that may overall result in patient enrollment. Institutional attention was focused on the physician by creating structured learning programs directed at improving patient‐centered communication. The programs included structured questions, and physician‐focused communication skills training to address trial‐related informational and relational communication with patients. Additional goals aimed to improve physician attitudes concerning clinical trial recruitment, and interpersonal activities between physician–patient teams to decrease bias, increase trust, and align goals such that physician and patient share the same common purpose. ³⁰ , ³⁷ , ³⁸ One study noted success through multilevel, center‐wide policy change, by combining infrastructural process control coordination with clinical investigators, data analysis, and reporting utilizing Protocol Review and Monitoring Committees. Independent committees were able to monitor proportional accrual across demographics. ²⁹ The efforts resulted in an increase in racial and ethnic minority group accrual to therapeutic trials from 12.0% in 2005 to 14.0% in 2010 and a significant increase in the number of racial and ethnic minority individuals participating in all trials (therapeutic and nontherapeutic). ²⁹ The infrastructural capabilities, adequate‐resource funding for data collection and management, staffing for enrolling, managing, and tracking participants, and efficient Institutional review boards are all necessary prerequisites to accomplish system‐level goals. ²⁹ , ³⁰ , ³¹

Institutional policy changes can address the language of exclusion criterion and the groups excluded from the eligible patient population. In the 2020 guidelines for enhancing the diversity of clinical trial participants, the FDA recommended, among others, adjustment of restriction criteria based upon priorities within trial phase, and utilization of adaptive clinical trial designs that allow for prespecified trial design changes based upon interim findings. These strategies allow attention to phase‐level priorities and acknowledge that the priorities within the trial and the exclusion criteria characterizing those priorities may be untenably stringent at the next phase of testing; and early consideration to pharmacokinetic sampling establishes safe dosing in women who become pregnant, to ensure sufficient assurances of safety in patients choosing to continue in the trial where the risks to participant and fetus are clearly explained and understood, and the benefits of continued participation outweigh the risks. ⁸

Cooper‐Patrick et al noted that patient–physician interactions in race‐concordant relationships were associated with significantly higher participatory decision‐making ratings among patients than those in race‐discordant relationships, ³⁹ and later noted significant differences in the quality of information exchanged between African American versus White patients during audiovisual recorded tapes. ⁴⁰ In the latter, measures of physician verbal dominance during interactions were significantly higher, and patient‐centered decision‐making was lower among African American patients compared with White patients, even when controlling for patient and physician demographic characteristics and how long the physician knew the patient. ³⁹ Active patient engagement in the medical dialogue and less physician verbal dominance are sensitive markers of patient‐centered interaction dynamics. ⁴¹ , ⁴² Furthermore, previous studies have noted that a patient‐centered approach, grounded on shared decision‐making, was associated with increased adherence and trust among racially discordant visits. ⁴³ , ⁴⁴ , ⁴⁵

By constructing for physician's situation‐specific plans for discussing a trial with a patient, interventions hope to increase the likelihood that physician participants will engage in said behavior, and by layering skills, the interventions build upon one another to establish a solid base on which effective, clear communication grounded in shared‐decision making can begin. ³¹ , ⁴⁶ These skills build knowledge and willingness among physicians to engage patients in making informed decisions regarding participation in trials. Because the physician is often the first stakeholder to engage patients and the representative with whom patients are most familiar, establishing strong buy‐in at the physician level is often a vital component in increasing patient engagement, participation, and retention in clinical trials. ³¹ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ , ⁵⁰ , ⁵¹

Community‐Based Initiatives: the Built Environment and Community Ambassadors

Community‐based interventions are collaborative strategies that enlist multiple stakeholders in a targeted community. ¹⁸ , ⁵² Community members trained by and working with researchers have been shown to improve underrepresented patient and participant engagement. ⁵³ , ⁵⁴

The shift to holistic care services that address the needs of patients outside of the clinical setting has spurred efforts in surveillance, regulation, and quality assurance methods that provide wrap around services to the patients for whom hospitals are responsible. States such as Massachusetts, Minnesota, Washington, and Vermont have adopted this model of care, integrating the care they provide with the communities that they serve. These fundamental service transformations have wide‐reaching health impacts and represent an opportunity for integration of primary health services alongside community‐based engagement strategies to expand the access and reach of recruitment interventions across sociodemographic strata among vulnerable patient populations. ²⁸ , ⁵⁵ , ⁵⁶ , ⁵⁷

Innovations within the environment can produce robust results. For example, utilization of members of the community as street outreach workers at Baltimore's Sinai Hospital and other community‐based centers in Baltimore's Safe Streets program is an attempt to mediate conflicts between parties involved in a shooting before discharge, whereupon increased escalation of violence is likely without intervention. The community initiative has shown meaningful results in its ability to decrease homicides and nonfatal shootings. The program is overseen by the Baltimore City Health Department and run by community‐based organizations, and it was developed based upon the successful intervention program CeaseFire seen in Chicago, where violence interrupters present at hospitals work to integrate hospital functions and community realities to decrease gun violence. ⁵⁸ , ⁵⁹ In another study that randomized Mexican Americans to hypertension education by community health workers versus controls using language‐appropriate education materials, community members within the intervention group reported significant improvement in dietary salt/sodium intake at 9 weeks postintervention. ⁶⁰

Advancing health equity requires coordinated actions between the institution‐based internal and community‐based external environments. Research indicates that system affiliation, larger hospital size, and hospitals with higher profit margins have been associated with more positive results from community intervention. Larger hospitals' pre‐existing organizational arrangements are positioned in a more favorable situation to move forward with community health needs assessment implementation. ¹³ Resource limitations have often been limiting factors for CBPR outcome success. Enhancing the capacity of community‐based intervention strategies allows researchers to prioritize factors related to the cultural and socioeconomic contexts of different patients. ²⁰ , ²⁸ , ⁵⁵ , ⁵⁶ , ⁵⁹ , ⁶⁰ , ⁶¹ For example, under the Urban Health Initiative, the Chicago Anchors for a Strong Economy represents a collaboration between anchor institutions (eg, universities, cultural institutions, hospitals, government, businesses) and small businesses to build economic vitality across Chicago's neighborhoods. The goal of the collaboration was to increase local spending, facilitate new contracts, grow a network, and facilitate training within Chicago. Through Chicago Anchors for a Strong Economy more than 236 businesses aligned in the creation of new contracts, created more than 180 new jobs, and $50.6 million in revenue has been committed to small businesses through its multiyear anchor contracts. ⁶² , ⁶³ Combining internal resources horizontally across different public and private health systems increases the opportunity for investment capital for infrastructure development to produce measurable results where health inequalities are most concentrated. By doing so, these interventions can transition from small‐scale, isolated, and short‐term events into institutional and transformational health care delivery and trial recruitment reform programs.

While no research has been completed concerning the specific outcomes related to patient enrollment, several examples demonstrate how logistical support by integrating services into traditional community locations, such as grocery stores, has increased community health awareness and benefitted health outcomes (Data S1). ²⁸ , ⁶¹ , ⁶² , ⁶³ Many qualitative studies and interviews indicate that providing reimbursements for and transportation to clinical trial sites are a means to increase underrepresented participant enrollment. Although several trials have incorporated transportation reimbursements, it is unclear what specific effect logistical support has on patient recruitment. ¹⁵ , ⁵² , ⁵³

Promising Digital Interventions

Artificial Intelligence‐Enhanced Honest Broker and Research Registries

Policy changes and the advent of digital tools and media opened an avenue for improving outreach for institutions by leveraging health care messaging, digital advertising, and research registries. Previously cited hurdles to clinical trial participation include lack of understanding, culturally appropriate information, and awareness of clinical trials. ⁶⁴ Digital advancements allow us to reach diverse populations by using technology to intentionally mitigate these barriers. ⁶⁵ , ⁶⁶ In cardiovascular trials specifically, along with being a new avenue of underrepresented patient outreach, digital health has also resulted in substantial benefits to improving participant health outcomes. ⁶⁷ , ⁶⁸ , ⁶⁹ In terms of technology's role in participant outreach, honest broker entities have become an integral part of connecting eligible patients to research trials. ⁷⁰ Honest brokers use artificial intelligence to tap into de‐identified patient databases and make the participant pool increasingly more robust by identifying large sets of potential participants based on their health records. Although literature identifies the role of honest brokers in ethically de‐identifying and utilizing electronic health record data to outreach to participants, there is little research on the direct effect of honest broker use on underrepresented patient population enrollment in trials. ⁴⁸

A research registry is a database of potential participants who agree to be contacted for clinical trials for which they are eligible. Research registries may improve clinical trial accrual of diverse groups of participants, though this strategy has not yet been tested compared with other potential strategies. ⁷¹ In addition to institutional registries such as the Stanford Research Registry, ⁷² there are registries for LGBTQ+ ⁷³ participants, as well as Asian, ⁷⁴ Black or African American, ⁷⁵ and Hispanic or Latino ⁷⁶ participants. These registries provide a pool of willing potential participants that can be leveraged to increase the number of underrepresented groups in trials.

Inter‐ and Intrapersonal Interventions: Social Media and the Internet

Inter‐ and intrapersonal interventions using digital technologies have been proposed as strategies to increase racial and ethnic minority group enrollment in clinical trials. These targeted advertisement tools can tailor content to recruit a multitude of audiences. For example, tech companies have proprietary algorithms that allow researchers to set target demographics (eg, age, location, education, relationship status, interests, income via zip code) that can improve the reach of recruitment materials. ⁷⁷ Such methods of outreach have been utilized to reach rural and remote populations that were previously hard to reach using traditional methods. Previous studies have mapped out and identified these hard‐to‐reach communities and have had success using the mentioned social media filtering tools to exclude ineligible or uninterested audiences and target their ideal demographic instead. This specific targeting allows for study teams to tailor their outreach and use their budget efficiently by not using resources on populations that would not qualify for the study. ⁷⁷ , ⁷⁸ In a study evaluating the effectiveness of online outreach methods targeting African American women with untreated hypertension, Staffileno et al found that of 176 inquiries, 44% were from advertisements posted on the university website and 13% were from Facebook or Craigslist. In terms of number of inquiries, these online methods outperformed traditional methods. Additionally, online strategies had less of a burden on staff time and resources, highlighting the convenience and wide reach of online outreach methods for cardiovascular trials. ⁷⁹ These findings are congruent with other similar interventions and highlight social media as a feasible and effective recruitment method. ⁸⁰ However, not all studies using social media outlets such as Facebook have been successful in reaching target underrepresented minority group enrollment goals. ⁷⁸

Recruiting LGBTQ+ young adults through mobile applications and social media has also been successful; nearly 20 000 sexual and gender minority people were recruited to The PRIDE (Population Research in Identity and Disparities for Equality) Study, the largest online, longitudinal cohort study of the health of sexual and gender minority people. ⁸¹ A separate study comparing online and in‐person recruitment of LGBT individuals for baseline survey data showed that the screener survey completion rate of individuals who clicked on the social media ad was higher than those who were asked to complete an intercept survey at LGBT nightclubs and bars. ⁸² Overall, social media recruitment was faster and less expensive, although there were more data quality issues.

Similarly, online ad platforms help studies develop campaigns around patients' interests, locations, and other inclusion/exclusion criteria. Google Ads have been used to increase participation in clinical trials, and studies have been published showing how these features can precisely target potential participants. ⁸³ Additionally, advertisement data can be leveraged to identify audiences and recruit specific demographics more heavily. By cross‐referencing ad performance data with site recruitment statistics, a trial can identify which advertising strategies will be most effective for meeting enrollment targets. With this modern method of outreach, it is important to acknowledge that comprehensive regulations around such approaches are still in development as companies navigate consumer privacy. Some social media companies have recently removed ad targeting for what they call sensitive topics. ⁸⁴ One consequence of these restrictions is that advertisers will not be able to directly identify and recruit patients based on their health conditions; however, recruiters can still use more roundabout ways to reach their desired audiences. ⁸⁵

Recruitment Strategies Not Included

While there are many potential strategies that have been implemented over the years, not all are included in this review. We decided not to include cold calling, sending letters/faxing without making direct/personal contact with the community, materials featuring subjects from only 1 demographic, and physician‐/staff‐focused interventions. These methods did not have sufficient support for their effectiveness because recruitment methods for underrepresented participants were difficult to implement and scale. Enlisting research staff and physicians as direct recruiters were not included because this puts an additional burden on these already full roles. It is important to note that the study materials should center on the participant. This includes featuring accessible language, representative and culturally sensitive graphics, and clear content. ³

FDA Policy and Diversity in Clinical Trials

We hope to expand the available area of knowledge and cost‐effectiveness research globally by expanding the bottleneck of patient recruitment and retention. In doing so, this work can potentially serve as a foothold to improve binary comparisons between drugs or devices, as well as those assessments of medications or devices in comparison with behavioral interventions or policy‐based interventions. ⁸⁶ The body of our research can also expand research capability in multiple ways. Potential avenues include: increasing FDA provision for conducting randomized controlled trials, structuring studies to validate surrogate outcome measures, and functioning as a long‐term portal for data, safety, recruitment, and retention measures that improve the durability of clinical benefits associated with specific interventions. ⁸⁷

For a pharmaceutical company to remain sustainable, the revenues that it generates from the reimbursement and pricing system must be at least equal to that for the growth and investments into research and development. Broadly, the overall industry productivity in terms of the number of new compounds over the last 11 years has been flat, averaging 30 to 35 new compounds per year. In contrast, industry studies have pointed to increased research and development costs at an average of 8.5% per year, with the latest estimate of the average cost per new molecule entity, although contentious, equal to $2.6 billion. ⁸⁸ , ⁸⁹ , ⁹⁰ In line with these trends, the decline in average economic returns for new drugs has been falling to values near zero for the latest cohort that could be analyzed. In order for firms to maintain viability and the industry number of firms to remain the same, the average price (eg, per unit of health gain) must rise. ⁸⁸

Policy considerations for the short‐term budget impact and long‐term benefits of a product across multiple, different patient populations must be integrated into the US system if it hopes to respond to increased market pressures on drug manufacturers to produce specialty drugs that target specific populations. For example, in the pharmaceutical industry, the subset of purchasers who represent the patient population for whom a medication is effective will value the medication more highly than the population for whom the medication has no use. This unequal valuation leads to higher prices demanded by the pharmaceutical company since demand is already low and tailored to a specific population. However, without equitable representation in clinical trials, the costs and value of medical therapies are not equitable or consistent across and even within specific patient populations, and the risk of adverse side effects increases. ⁸⁸ , ⁸⁹ , ⁹⁰ , ⁹¹ , ⁹² , ⁹³ , ⁹⁴

From the standpoint of the FDA, the Food and Drug Omnibus Reform Act reflects commercial considerations. In terms of government representing a major payor of most medical expenditures and therapies, the FDA represents the gatekeeper around which medical therapies are validated. ⁹³ , ⁹⁴ The development of innovative technologies and devices engenders and necessitates the development of new methods and technologies of evaluation. As new therapies and devices continue to progress, new methods and technologies dedicated to measuring and improving evaluative measures will need to be developed, and the FDA itself must be cognizant and thoroughly experienced with these developing technologies and measures of evaluation. As sponsors continue to develop new medications and devices, the diversity on which they base the outcomes of their trials, and the efficacy of their interventions must be grounded in data. ¹⁰

A thorough Health Technology Assessment (HTA) focuses on long‐term viability of our health system and concentrates efforts on delivery system reform and payment mechanisms that consider the long‐term sustainability and value gained by patients and US markets from sacrifices made in the present. ⁹¹ , ⁹² However, such an assessment is not possible without first having an equitable stake of land in those clinical trials instructing medical therapies. While the grounding of assessments evaluating medical efficacy will still be fixed upon the long‐term benefits and cost‐effectiveness of products to US markets, HTA tools to evaluate true opportunity costs and evaluate budget impact thresholds relative to drug or device cost will prove critical for the present and future. ⁹¹ As such, the fundamental issue of patient enrollment and diversity within clinical trials is fundamentally also a discussion of opportunity costs. Addressing and working to correct the systemic issue of patient recruitment and representation grounds that discussion ethically and literally such that patient stakeholders have a voice, because the consequences of silence are exclusions that translate to patient outcomes.

To increase the comparability and quality of cost‐effective analysis, the Second Panel in Cost‐Effectiveness in Health and Medicine recognized that the health sector and societal evaluative perspectives and value frameworks must be incorporated into the cost‐effective analysis to allow researchers to properly account for the health‐related effects and non‐health‐related effects secondary to budget impact thresholds. ⁹⁴ To that end, the International Community for Pharmacoeconomics and Outcomes Research Task Force made 6 recommendations for clarifying value in different health plans to maximize health gain and well‐being for a given budget. Using international data from HTAs outside the United States, the task force recommended supplementing HTA evidence evaluations with lessons pertaining to governance, communication, trust, politics, and the persistence of local practice patterns. ⁹⁵ , ⁹⁶ As such, an HTA organization can fill the void in current cost‐effectiveness analyses by incorporating a more complete societal perspective in its cost‐effective analysis and account for consequences outside of the health sector. ⁹⁵ , ⁹⁶ , ⁹⁷ , ⁹⁸ , ⁹⁹ , ¹⁰⁰ , ¹⁰¹

An improved recruitment and retention strategy represents an opportunity to develop rigorous scientific evaluations and transparent, fair, predictable, and efficient evaluation procedures for innovative technologies and medications. To accomplish this, the resources and leadership of government, professional organizations, and citizens are pivotal to creating an organization capable of decreasing uncertainty and increasing the predictability, efficacy, and value of endorsed medications and scientific evaluations.

Overcoming Barriers to Advancing the Diversity in Clinical Trials

A major limitation of previous literature on clinical trial diversity is lack of control or comparator groups for different strategies. In addition, most studies examining recruitment strategies have targeted a single underrepresented group (ie, Black or LGBTQ+ populations) rather than considering multiple underrepresented groups (ie, Black and Hispanic and women and LGBTQ+ groups) and the intersectionality of diversity (ie, Black, Hispanic, LGBTQ+ women). Several strategies may have utility in improving enrollment of underrepresented patient populations in clinical trials. However, multilevel, interventional strategies with action plans that coordinate and leverage system capabilities alongside health care institutions, health care professionals, and participant communities present the best opportunity to improve the recruitment of underrepresented participants.

Institutional interventions are often complicated to coordinate, require consistent quality improvement evaluations, staffing, and are dependent on buy‐in from representatives of the organization and stakeholders with whom they do business. However, once implemented their benefits are well‐enduring so long as institutional fervor and funding allow. However, institutional intervention represents a specialized niche among interventional strategies because they can operate within the constraints of public interest, yet also extend beyond operating rooms and into private investment strategies. Consequentially, they improve interventional approaches for patient recruitment, coordinating alongside other primary stakeholders with significant stock in patient outcomes. Because institutions contextualize and localize patient health care experiences, collaborations within this sphere can produce outcomes several times over from an initial input. As the whole of the bureaucratic system moves with 1 mind and singular, coordinated, and intentional action, they disperse project interventions horizontally either within their own system or across different public or private health sectors to accomplish tasks for which they are most skilled, and which they can sustain. By doing so, institutional interventions can embody a clear and concise initiative and combine the varied resources and relationships to which they have access. The development of institutional‐level policies capable of directing this transformation such that it yields results that align the structures and functions of organizations, engage public policy and practice, and orient state and local governments for the future will be a measure of efficacy between institutional inter‐ and intraindustry collaboration.

CBPR strategies in coordination with institutions present a viable path for patient and community engagement, orienting community and institutional relationships through shared, goal‐directed practices and decisions, and improving the infrastructure outside of the organization itself to operationalize relationships and magnify results. The spillover effects of these relationships produce increased patient awareness, communication, and engagement in discussions needed for collaborative success. ¹⁰² The significant resources to accomplish and coordinate activities at the community level have limited the capacity and value of community‐based interventions. Because these interventions move outward from traditional locations that organize health care, community‐based relationships and institutional collaborations are an opportunity to engage patients directly and foster patient education and health literacy. This method creates relationships that foster trust, partnership, and a shared space in which true collaborative work occurs with communities instead of work on communities. While these relationships may require considerable time and resources to establish, the opportunity costs may well be worth the investment, and may be offset by collaboration with other similar or disparate health care institutions (Data S1).

Lastly, digital interventions offer a feasible and cost‐effective means of focusing interventions on specific patient populations, but they come at the risk of poor patient targeting, low investment, and unclear retention. However, the utility of a centralized database or computer function capable of registering and accessing possible participants and automating selection by which studies participants are suited has its promise. Better still, if it automates which of those patients are more likely to agree to clinical trial testing when asked. However, the capabilities and limitations of machine learning are systems that need to be tested and calibrated across different patient demographics and testing sites, and currently, no such system exists.

Conclusions

Each strategy may have benefits that outweigh the costs if organized with clear goals and directives. In differentiating which strategies are effective and contextualizing these answers within specific patient and institutional contexts across a wide variety of facilities, our forthcoming study under the American Heart Association DIVERSE Strategically Focused Research Network–Science of Diversity in Clinical Trials hopes to add to the current body of research on clinical trial and diversity, provide actionable methods by which organizations can address health inequalities, and improve patient outcomes for the large portion of patients who are underrepresented in current clinical trials.

Sources of Funding

This work was supported by the American Heart AssociationGrant: 20SFRN35360178; Award ID: TOTAL: 946917; TIDE: 946918.

Disclosures

None.

Supporting information

Data S1