Addressing Challenges in Recruiting Diverse Populations for Research: Practical Experience from a P20 Center

Abstract

Background:

Improving the recruitment and retention of underrepresented groups in all research areas is essential for health equity. However, achieving and retaining diverse samples is challenging. Barriers to recruitment and retention of diverse participants include socioeconomic and cultural factors and practical challenges (e.g., time and travel commitments).

Objectives:

The purpose of this paper is to describe the successful recruitment and retention strategies used by two related studies within a National Institute of Nursing Research-funded P20 center focused on precision health research in diverse populations with multiple chronic conditions (MCC), including metabolic syndrome (MetS).

Methods:

To address the complexity, biodiversity, and effect of MetS and MCC, we developed culturally appropriate, multipronged recruitment and retention strategies for a pilot intervention study and a longitudinal observational pilot study within our P20 center. The following are the underlying principles that guided the recruitment and retention strategies: (a) flexibility, (b) active listening and bidirectional conversations, and (c) innovative problem-solving.

Results:

The intervention study (Pilot 1) enrolled 49 participants. The longitudinal observational study (Pilot 2) enrolled 45 participants. Women and racial/ethnic minorities were significantly represented in both. In Pilot 1, the majority of the participants completed the intervention and all phases of data collection. In Pilot 2, the majority of participants completed all phases of data collection and chose to provide biorepository specimens.

Discussion:

We developed a recruitment and retention plan building on standard strategies for a general medical population. Our real-world experiences informed the adaption of these strategies to facilitate the participation of individuals who often do not participate in research—specifically, women and racial/ethnic populations. Our experience across two pilot studies suggests that recruiting diverse populations should build flexibility in the research plan at the outset.

Improving the recruitment and retention of women and underrepresented racial/ethnic minority groups in all research areas has been a priority for the National Institutes of Health (NIH) for more than 2 decades. The NIH Revitalization Act of 1993 and subsequent amendments mandate that “it is imperative to determine whether the intervention or therapy being studied affects women or men or members of minority groups and their subpopulations differently” (NIH, 2000). Addressing health equity hinges on the representation of diverse groups. It is important to note that diversity is more than race and ethnicity. It is essential to include socioeconomic status, education, environment, sex, and other characteristics that affect health.

Progress has been made in increasing the participation of underrepresented populations in research, with specific improvement in the enrollment of women in large studies (Feldman et al., 2019). However, there remains a significant chasm in enrolling diverse ethnic and minority women and men into research. Reasons cited for ongoing recruitment challenges are similar across research foci. These reasons include socioeconomic and cultural factors, awareness and opportunity barriers, and potential health care provider bias (Nicholson et al., 2015). More attrition in research studies for racial/ethnic minority populations than White populations is related to participant burden and includes study duration, time commitment, travel, and potential participation costs (i.e., hourly wage loss; Denny et al., 2020).

Inclusive and equitable recruitment and retention in research are essential in populations with multicontextual risk factors for multiple chronic conditions (MCC) who, as a result, experience poorer health outcomes (Buttorff et al., 2017; Virani et al., 2021). Metabolic syndrome (MetS) is a set of interrelated health conditions highly prevalent in the U.S. and places individuals at increased risk for MCC (Virani et al., 2021). Compared to non-Hispanic Whites, Black women (OR, 1.20; 95% CI, 1.02–1.40) and Black men (OR, 0.74; 95% CI, 0.62–0.89) are more likely to be diagnosed with MetS and experience significant barriers to its management (Moore et al., 2017).

The National Institute of Nursing Research (NINR) funds P20 exploratory centers to build new research teams for interdisciplinary research in MCC. Our P20 center is based in one of the world’s most diverse cities and is uniquely situated to focus on precision health research among diverse populations. Accordingly, the overarching research goal of our P20 center is to understand the root causes, underlying determinants, and inequalities of health outcomes related to MetS and MCC to develop precision health strategies for health equity. While the P20 funded pilots’ recruitment plans included standard methods, we developed culturally appropriate, flexible, multipronged recruitment and retention techniques. The New York University’s institutional review board (IRB) approved Pilot 1 (18-00707) and Pilot 2 (S18-00447). The recruitment period for these two pilots was from January 2019 through March 2021, with a 4-month pause in recruitment during this period due to COVID-19 research restrictions. This paper shows the recruitment and retention strategies used by two pilot studies of our P20 center and presents common challenges and efforts to ensure protocol fidelity across the projects. Descriptions of Pilot 1 and Pilot 2 are presented in Table 1.

Table 1.

Common and Distinct Components of Pilot 1 and Pilot 2

	Pilot 1	Pilot 2
Pilot Study Component	Extending sleep to reverse the metabolic syndrome in short- sleeping, middle-aged adults: A feasibility and acceptability study	Predictors of Morning and Evening Fatigue Profiles in Adults with Metabolic Syndrome
Purpose	Determine the feasibility and acceptability of a 12-week sleep health intervention in short-sleeping, diverse middle-aged adults at risk for MetS.	Describe fatigue in diverse adults with MetS and evaluate for biobehavioral characteristics that define subgroups with higher fatigue severity
Design	Single arm pretest/ posttest intervention study	Prospective longitudinal study with repeated measures
Methods	At home sleep evaluation with actigraphy and surveys (weeks 1 and 2 and weeks 11 and 12); Clinical research visit; Symptom surveysa, daily sleep diaries (weeks 1 −12); qualitative interview (week 13)	Symptom surveysa weekly upon awakening and at bedtime for four weeks; Clinical research visits to collect MetS clinical criteria, blood collection for circulating biomarkers and genomic DNA (weeks 1 and 4)
Inclusion Criteria	≥ 1 objectively confirmed MetS clinical criteriab
	Age 35–60 years Accelerometry-confirmed short sleep (≤6 hours/night)	Age ≥ 18 years
Exclusion Criteria	Pregnant/lactating Current chemotherapy or radiation therapy
	Moderate-severe depression Alcohol abuse Shift work or trans-meridian travel Habitual napping Untreated diagnosed OSA Sleep-promoting medication	Major Depressive or Psychiatric disorder T1D or T2D CVD other than HTN
Recruitment Methods	Community and clinical placement of IRB approved paper and video flyers Electronic health record messages Screening for the other pilot Emails to health care providers for referrals
	Contacting past study participants Research registries Education sessions	On-site recruitment

Note. CVD = cardiovascular disease; HTN = hypertension; IRB = institutional review board; MetS = metabolic syndrome; OSA = obstructive sleep apnea; T1D = type 1 diabetes; T2D = type 2 diabetes.

^aSymptom surveys: anxiety, depression, fatigue

^bObjectively confirmed MetS clinical criteria: Abdominal adiposity (waist circumference > 120 cm [men]; > 88 cm [women]); Hypertension (Blood pressure ≥ 135/85 mmHg); Hyperglycemia without diabetes diagnosis (Fasting glucose ≥ 110 mg/dL); Hyperglyceridemia (Triglycerides ≥ 150 mg/dL); Hypercholesteremia (HDLc < 50 mg/dL [men] < 40 mg/dL [women])

Methods

Cross-Pilot Recruitment and Retention Strategies

Within our P20 center, the pilot principal investigators (PIs) collaborated on recruitment and retention strategies, prepared materials, incorporated screening questions that they could share across projects, and provided referrals when appropriate. The PIs planned a multipronged recruitment strategy. This strategy included (a) advertising in clinics with IRB-approved recruitment flyers, (b) using the electronic health record (EHR) to send IRB-approved study recruitment messages to potential participants bimonthly, and (c) sending emails to health care providers for referrals. Clinic advertising included posting a version of the recruitment flyers on electronic “video bulletin boards” in the New York University-affiliated dental clinic and distributing paper recruitment flyers in general medical clinics and provider offices. The PIs developed EHR queries based on pilot inclusion/exclusion criteria in collaboration with data management staff. The PIs received lists of eligible patients from the health system associated with the P20 center and then sent IRB-approved recruitment emails through the EHR.

Weekly meetings of the P20 center pilot research team supported recruitment and retention across the projects. The pilot PIs shared research assistants and a program coordinator. This deliberate team design facilitated cross pilot recruitment and the identification of challenges and opportunities across the pilots. During the meetings, the team discussed trends in recruitment effectiveness and challenges to generate solutions that could apply across projects. Throughout the pilots’ participant recruitment stage, enrollment yields were calculated as the ratio of the number of participants enrolled in the research study to the number of participants screened from each recruitment strategy. In this way, we quantified the effectiveness of each strategy. (Davis et al., 2018).

Pilot 1 Distinct Recruitment and Retention Strategies

Pilot 1 distinct strategies included (a) contacting participants in the registry of the investigators’ past studies, (b) posting study information on the P20 Center social media site, (c) using Research Match, a secure online registry that connects potential participants to research studies (Harris et al., 2012), and (d) community outreach.

Pilot 2 Distinct Recruitment and Retention Strategies

Pilot 2 distinct strategies included (a) maximizing study team availability to potential participant email or phone queries and requests to meet in person, (b) approaching potential participants in the waiting room for scheduled dental or health care visits, and (c) speaking openly with potential participants about barriers to research.

Results

Cross-Pilot Referrals

Both pilot projects used the same inclusion criteria for MetS but had different exclusion criteria (Table 1) facilitating pilot-to-pilot referrals. When potential participants were not eligible for Pilot 2, the research team asked a screening question (e.g., “Do you sleep less than 7 hr a night?”) and referred those who met that criterion to Pilot 1 (n = 10). Pilot 1 offered information to individuals not eligible for the sleep intervention after the first in-person screening visit (n = 32). This cross-referral process resulted in three participants enrolled in Pilot 1 after a referral from Pilot 2 and one participant enrolled in Pilot 2 after a referral from Pilot 1.

Pilot 1

Pilot 1 telephone-screened 162 potential participants. Of these 162 individuals, 81 signed an informed consent to begin a multistage screening process, resulting in a final sample of 49 participants eligible for the sleep intervention (43% Black, 41% White, 4% other races, 12% Hispanic ethnicity, 57% women). See Figure 1. This sample represents 82% of the projected recruitment goal of 60 participants. EHR messages accounted for the largest number of telephone-screened contacts (n = 94) and generated an enrollment yield of 35% (n = 33). Of these participants, 59% were racial/ethnic minorities (n = 19). Past study participants accounted for 18 telephone-screened contacts and generated an enrollment yield of 50% (n = 9); 56% were racial/ethnic minorities (n = 5). Clinic advertising accounted for 26 telephone-screened contacts and generated an enrollment yield of 23% (n = 6); two participants identified as a racial/ethnic minority. Research registries accounted for two telephone screens and yielded one racial/ethnic minority participant. Community outreach (i.e., visiting food banks, hair salons, churches, and educational programs via the local K–12 school, webinar presentations) yielded no screening contacts. The EHR recruitment strategy was the most effective for recruiting ethnic/minority persons in Pilot 1.

Figure 1. Pilot 1 and Pilot 2 Recruitment and Enrollment Flow.

Abbreviations: CVD = cardiovascular disease; T2D = type 2 diabetes; OSA = obstructive sleep apnea

Retention rates during the 12-week sleep intervention were very high. Pilot 1 retained 41 participants (84%). One participant was lost to follow-up, three withdrew due to study burden, and four became ineligible due to new onset mental health diagnoses, inadequately treated obstructive sleep apnea (OSA), and shift work. Adherence to the intervention was very high, as evidenced by daily sleep diary completion rates averaging 6.7 diaries per week and a 95% attendance rate at the weekly intervention sessions.

Pilot 2

Pilot 2 screened 112 potential participants (43.8% Black, 31.3% White, 12.5% other races, 23.2% Hispanic ethnicity, 64% women), and enrolled 45 (48.9% Black, 20% White, 31.1% other races, 13.3% Hispanic ethnicity, 78% women) See Figure 1. This sample represents 75% of the projected recruitment goal of 60 participants. Clinic advertising accounted for the largest number of screening contacts (n = 68) and generated an enrollment yield of 45% (n = 30). It is important to note that of the 45% enrollment yield from the flyers, 40% (n = 12) were recruited in person; most were racial/ethnic minorities (n = 11). EHR messages resulted in 40 contacts and generated an enrollment yield of 35% (n = 14). One participant was enrolled from a Pilot 1 referral. Health care provider referrals resulted in three potential participant contacts but no enrollment. We hypothesize that the effectiveness of the video and paper flyers at clinics is related to our availability and maximized flexibility meeting in person and our individualized description of procedures associated with this pilot’s research strategy.

Retention rates during the month-long study were very high (n = 43, 95%). None of the 45 enrolled participants withdrew. Two participants were lost to follow-up due to a lack of reliable contact methods (e.g., lost access to phone and email). Pilot 2 included biomarker collection for analysis and an optional biorepository. There was a 98% opt-in rate for the biorepository. To date, no participants have withdrawn biorepository consent.

Discussion

The goal of our P20 pilot studies was to ensure diverse representation of participants, particularly women and racial/ethnic minorities. Both pilot studies recruited over 40% Black, 10% Hispanic, and 50% women participants, successfully meeting this goal. At the outset, our recruitment and retention plan was developed based on standard strategies for a general medical population. Our real-world experiences informed the adaption of these strategies to facilitate the participation of individuals who often do not participate in research—specifically women and diverse populations. The following are the underlying principles that guided the successful recruitment and retention strategies contributing to each pilot’s success: (a) flexibility, (b) active listening and bidirectional conversations, and (c) innovative problem-solving.

Principle 1: Flexibility

Flexibility was integral to recruiting diverse participants for both pilots. There are practical barriers to recruiting and retaining diverse participants into longitudinal and intervention studies. Multiple research appointments, associated time (i.e., work and family responsibilities), and required resources (e.g., transportation costs) may limit participation interest.

Flexible research teams who are available to describe the study when potential participants contact them may increase enrollment (Falcon et al., 2011). Once enrolled, flexibility is essential to reduce practical barriers (e.g., scheduling in-person data collection around participants’ daily responsibilities, preexisting health appointments, and transportation needs) and support retention (Falcon et al., 2011).

We learned that participants often traveled to their health care appointments on a day off from work and incurred travel expenses due to public transportation or ambulette services. The research team’s flexibility and availability to the potential participant demonstrated respect for their time and the value of their research involvement. Several participants commented that seeing the value the team placed on meeting them when and where they were available made them feel important to the project.

In Pilot 1, flexibility included modifying the inclusion/exclusion criteria to recruit a diverse sample of individuals diagnosed with MetS. See Table 1 for a complete list. The inclusion/exclusion criteria were modified to include persons at risk for OSA but not currently diagnosed with OSA, as well as persons diagnosed and adherent to OSA treatment. The modification was justified because direct and indirect associations between sleep health and MetS suggest that behavioral sleep interventions may benefit individuals diagnosed with MetS regardless of OSA. Also, OSA is often undiagnosed in racial/ethnic minorities with poor treatment adherence rates (Billings et al., 2011). Since health disparities exist in the treatment and management of other chronic conditions (Ezenwa et al., 2006), it is plausible that similar differences exist for treating and managing poor sleep via medication use or cognitive behavioral therapy for insomnia. Therefore, modifying the sleep-promoting medication use and insomnia criteria was essential to ensure a more representative sample of people suffering from poor sleep, including those who could not afford or access treatment. With this strategy, 21 persons who would have been excluded were able to participate in the study.

In Pilot 2, the study team’s flexibility maximized responsiveness to potential participant email or phone queries. When the potential participants called the contact number to ask about the study, they often waited at a clinic appointment to watch the video monitor and see the recruitment flyer. These potential participants wanted information when contacting us; they did not want to schedule another appointment. We provided an overview of the study during such a call and asked if they would like to be screened. Participants asked for an in-person meeting 40% of the time. In-person meetings were facilitated by the proximity of the PI and research staff to the recruitment settings (e.g., within four blocks of the study team’s office). We prioritized going to their site, demonstrating their importance to the study and how we valued their time. When we met the potential participants in person, we had a 100% enrollment rate of those who met screening eligibility.

Recognizing the importance of in-person meetings, the research team routinely visited waiting rooms of clinics and physician offices. Distributing paper flyers in the waiting areas was an opportunity for individuals who might otherwise not have been responsive to research participation to ask questions, meet the research team, and consider enrollment. The openness of potential participants to engage in conversation about research indicates the importance of personal connections for the recruitment of diverse participants and resulted in 20% of these informal conversations resulting in formal screening for study inclusion. In-person recruitment—in which a study team member approaches a potential participant—demonstrates high recruitment rates (Tutino et al., 2019). However, participant autonomy and choice must be emphasized to limit the potential for acquiescence or coercion, especially with vulnerable diverse populations (Ramsay et al., 2020). Flexibility maximized participant inclusion and gained valuable data on the studies’ acceptability and feasibility without compromising the integrity of the research.

Principle 2: Active Listening and Bidirectional Conversations

Active listening and bidirectional conversations were identified as another underlying principle contributing to the successful recruitment and retention of diverse populations. Barriers to participating in intervention research for racial/ethnic minorities exist on several levels, including intrapersonal, interpersonal, and community/environmental (Joseph et al., 2015). Our team built a person-centered approach to recruitment and enrollment. Our goal was to build trustworthiness without promoting acquiescence bias. For example, in addition to the standardized study description provided to each subject in the consent process, we also asked potential participants about their transportation to the clinic, the schedule of future appointments, and other life responsibilities that may challenge their ability to participate in the research. Our questions aimed to proactively identify specific practical barriers for participation so individuals could make an informed choice about the effect of the research schedule on their lives. We focused on identifying barriers, collaboratively evaluating the feasibility of how their schedule could fit the research protocol requirements, and explained how we could and could not adapt the research protocol. These discussions provided individuals with details to examine the practical barriers to their participation to decide if participation was overly demanding or “doable.”

For instance, individuals may be reluctant to participate in sleep intervention research if they perceive that the intervention will not be effective due to situational factors (e.g., caregiving or work responsibilities). Approaches that fail to address situational factors may be viewed as irrelevant and lead to poor recruitment and retention, particularly for racial/ethnic minorities (Joseph et al., 2015). Hence active listening and bidirectional conversations are essential underlying principles to advance recruitment and retention of underserved groups for intervention research

In addition, the Pilot 1 study team sought advice from the institution’s Patient Advisory Council, a diverse group of community members. The study team was particularly seeking feedback regarding the study flyer messaging. Initial study recruitment flyers led with lines such as, “Are you sleeping less than 7 hr per night?” The Patient Advisory Council noted a gap in people’s understanding about the importance of sleep for health. The Patient Advisory Council recommended that the flyers be educational and communicate a sense of urgency. The video and paper flyers for Pilot 1 were revised (e.g., “Worried about obesity? Sleep less than 7 hr a night increases your risk!”) and improved recruitment rates.

Active listening and bidirectional conversations also contributed to the high retention rates throughout the 12-week sleep intervention. The participants’ data and concerns drove the intervention. Weekly progress was calculated from participants’ sleep diaries and tracked across key sleep health metrics (e.g., total sleep duration, sleep regularity). This information was shared graphically and verbally with participants during the weekly sessions, and questions were encouraged.

In both pilots, the team monitored the weekly email survey results. If a participant did not answer the survey within 24 hr of the expected completion, a team member would follow up. The goal of our conversations was to facilitate data collection and support retention. We asked how they were if they had questions about the surveys before reminding them to complete a missed survey. Consistently, the participants offered that they had not completed the survey because of competing priorities or problems with their technology. Participants would either complete the survey during the call or independently after the call.

A research protocol, such as Pilot 2 that includes collecting biomarker data, adds psychological barriers forged by historic misconduct and discrimination (Samayoa et al., 2020; Sheridan et al., 2020). Biomarker collection does not automatically increase participant risk (U.S. Department of Health and Human Services, Office for Human Research Protections, 2015). Historically, diverse populations have been reluctant to participate in research, especially studies that include biomarker collection (Jaiswal & Halkitis, 2019). In diverse populations, the perceived risk due to potential misuse of the specimens and distrust of researchers are significant psychological barriers to recruitment (Samayoa et al., 2020; Sheridan et al., 2020). Distrust in medical research stems from the exploitation of African Americans in the Tuskegee experiments and the legacy of Henrietta Lacks (Skloot, 2010). Corbie-Smith et al. (2004) described African Americans’ distrust in the transparency of study protocols and feared that the risks were greater than was disclosed. Strategies to address these psychological barriers require open bidirectional, nonambiguous conversations about the biospecimens collected methods, purpose, and use (Samayoa et al., 2020). During recruitment, active listening and responsiveness to questions about the biomarker collection and study procedures may facilitate participation through education and building trustworthiness (Samayoa et al., 2020). We attribute our success in obtaining biorepository consent to opening the conversation with participants about past exploitations. It is important to note that we engaged in bidirectional discussions, taking time to answer specific questions about using all potential participants’ biospecimens, regardless of race/ethnicity. We built trustworthiness with authenticity, respectful complete explanations that gave the participants the power to consent or decline.

We also employed this strategy when addressing when a participant’s research MetS criteria lab results were outside clinical standardized normal limits or depressive symptom screening results required clinical evaluation. The PI called the participant with the results and instructed them to contact their primary care provider. However, some participants (n = 4 Pilot 1; n =19 Pilot 2) did not identify a primary care provider. We coordinated referrals to the free and low-income clinics that served as recruitment sites for these participants. Building rapport through active listening and bidirectional conversations is integral to clinical care; it is integral to implementing research. In general, people are more likely to persevere in a study when treated with respect and kindness (Kim et al., 2014).

Principle 3: Innovative Problem-Solving

Innovative problem-solving addressed structural, social, cultural, and religious barriers for participants and limited resources for small pilot studies. Structural inequities (e.g., transportation, communication methods, unpredictable work schedules, housing) are barriers to the inclusion of underrepresented communities in research (Nicholson et al., 2015). Small pilot studies often lack resources that larger projects must recruit diverse populations. Innovative problem-solving helped address potential barriers for participants that would have prevented enrollment or retention without compromising study integrity.

Sleep interventions face unique recruitment and retention challenges related to burdensome data collection and study-related travel. Data collection methods are burdensome because participants must complete daily sleep diaries, wear devices (e.g., accelerometers) for extended periods, and contact study personnel regularly. These structural barriers have hindered research participation in intervention studies for underrepresented populations (Nicholson et al., 2015). Technology promises to overcome structural obstacles and thereby assist in recruiting and retaining diverse participants. Emerging research suggests that the uptake of technologies is greatest in minority and economically disadvantaged populations in the United States (Rice & Katz, 2003). Pilot 1 planned for and used an IRB-approved videoconferencing platform. We shipped wearable technology used during the intervention (e.g., accelerometers) to and from participants to minimize study-related travel burden for participants further. There were no participant costs associated with the videoconference platform or mailing equipment. The videoconference platform is part of the institution’s technology infrastructure. Removing structural barriers through technology contributed to the successful recruitment and retention of racially/ethnically diverse participants.

When recruiting a diverse sample, religious or cultural practices can challenge participation. Our team generated creative solutions to resolve specific participant issues to include diverse participants in our study. We adapted the scheduling of participant research appointments to respect religious observance requirements, such as fasting, menstruation limitations, or interactions with the opposite sex. For example, during Ramadan, we continued to recruit participants, then scheduled them to begin the study 1 week after the fasting observances ended. When a participant canceled an appointment due to the commencement of her menses, which, according to her culture, did not allow her to have blood drawn, we asked her how she would like to proceed with study participation while explaining the requirements biomarker collection. We worked to respect religious and cultural issues, maximizing the inclusion of diverse participants.

Insecure housing limits potential participant recruitment and enrollment. Since we recruited from free or low-cost clinics, we were contacted by potential participants who did not have a stable address or phone. However, some did have access to the public library computers and internet and could thus receive emails. We explained the need to log out of their emails when leaving the public computer to maintain their confidentiality. We worked with them to develop a contact system through emails, updating their “burner phones,” and ensuring our availability when they would email or call. A burner phone is an inexpensive, prepaid mobile phone that can be used for communication and then disposed of when no longer needed. To facilitate call responsiveness and accessibility, we forwarded the study contact phone number to our private mobile phones using a free internet phone number that improved our availability while maintaining the privacy of our mobile number. These participants were able to reach us when they had access to their methods of communication. If we failed to respond promptly, they would have been lost to follow-up.

Other examples of creative problem-solving involved adapting our environment and procedures to facilitate the research process. For instance, when participants came to the office to discuss the study protocol, it became evident that the environment was not conducive to some adults with MetS due to abdominal adiposity. To make the environment more respectful and accessible to these participants, we obtained a wider, sturdier chair for their comfort. We also learned that our emailed gift card incentive process was not practical for those participants who had insecure housing. Therefore, we adapted our incentive method to gift cards without fees to deliver in person at the final study visit.

To support ongoing recruitment and retention in future P20 center pilot projects, we have developed best practice protocols available to our college research community. Examples of effective recruitment material, messaging, procedures, and recommendations for unique situations are stored on the P20 center data repository. When the pilot studies resumed recruiting after the COVID-19 research pause, technology-supported recruitment, and enrollment were used. The team used telephone and video calls to describe the study to participants and answer their questions. While six individuals were willing to enroll for Pilot 2, they were unwilling to come to the clinical research site for biomarker collection due to the potential exposure to COVID-19. Since alternative methods of biomarker collection (i.e., in-home phlebotomy) were deemed not feasible for reasons including the exposure risk and cost, the study was closed to enrollment.

Through resourceful problem-solving, we addressed numerous unanticipated social, cultural, and religious barriers to enrolling and remaining in the research study. Structural inequities (e.g., transportation, communication methods, unpredictable work schedules, housing) are barriers to the inclusion of underrepresented communities in research (Nicholson et al., 2015). Our pragmatic yet creative problem-solving helped address potential structural barriers without compromising study integrity. Social, cultural, and religious barriers required that the study team collaborate with a participant to identify possible solutions to unique problems. Respecting sociocultural needs entailed recognizing that the participant was the expert on the matter (e.g., fasting as a religious belief) and was a partner in identifying a solution. We discussed these challenges and potential solutions during weekly team meetings and examined the study effect. This robust dialogue across the study teams enabled us to maintain study fidelity while increasing inclusion and retention.

Conclusion

In summary, our recruitment and retention efforts suggest that to recruit and retain diverse participants successfully, we need to move away from strategies based solely on “the way it is usually done,” which have become barriers in real-world research. Our real-world experiences informed the adaption of these strategies to facilitate the participation of individuals who would not usually be able to participate in research, specifically, women and diverse populations. Comparing enrollment yields for specific recruitment strategies identifies divergent results between the two pilots. For example, Pilot 1 successfully recruited a diverse sample using EHR messaging. In contrast, EHR messages were not as effective as in-person conversations to recruit diverse participants for Pilot 2. Our experience suggests that approaches to recruiting women and racial/ethnic populations should build flexibility in the recruitment plan at the outset and strive to include individuals rather than exclude those who may have unique needs. The lessons learned from our P20 center’s first two pilots will support ongoing inclusive recruitment and retention in future P20 center pilot projects.

Table 2.

P20 Center Recruitment Principles and Strategies

Principal	Strategies
Flexibility	Accommodating participant preferences for in-person/on-site or email recruitment contact Scientifically justified inclusion criteria modifications for maximal inclusion Research accessibility based on participants’ schedule
Active Listening and Bidirectional Conversations	Proactively identifying specific practical barriers for participation Educational messaging about study purpose based on patient advisory committee Contacting participants when weekly surveys not completed to identify barriers and solutions to complete data collection Acknowledgement and responsiveness to questions and concerns about the methods, purpose, and use of collected biospecimens Referral process for low-income free follow-up
Innovative Problem Solving	Weekly research team meeting to focus recruitment efforts based on enrollment yields and address challenges Schedule flexibility based on religious or cultural practices IRB approved video conferencing Mailing study related equipment Forwarding contact phone to internet phone for increased availability of study team for participants without stable address or phone Adapt study environment for participants with specific needs (i.e., wider, sturdier chairs, incentive methods)