Background
Despite a concerted effort by the National Institutes of Health [
1], other federal agencies [
2], and individual investigators [
3], problems persist in the recruitment and retention of patients from underrepresented populations, including racial and ethnic minorities, non-English speakers, and patients who are low-income, have low health literacy, or have substance use or mental health problems [
4‐
7]. Lack of representation from these groups limits the generalizability of research findings [
8‐
10] and may contribute to ongoing disparities in the provision of health care and inequities in health policy and funding decisions. Underrepresentation also limits the ability to tailor treatment to the populations who often carry the highest burden of disease [
11]. To address such disparities it is important to develop well-designed, culturally adapted research interventions and health promotion programs that are relevant to the population of interest [
12].
Challenges to recruiting and retaining members of underrepresented groups include choosing research questions relevant to these groups and establishing relationships with communities being recruited [
13‐
15]. In addition, members of underrepresented groups may lack experience with or knowledge about medical research [
11] and face practical barriers to participation such a longer travel times and less flexible schedules [
15‐
17]. They also may have beliefs and practices that may contrast with some elements of the study design [
12,
18]. Minority groups are also more likely to have co-morbidities, which may result in their exclusion from, and lower participation in, clinical trials [
19].
Disparities in the prevalence and severity of COPD and in access to care for low-income and racial or ethnic minority patients is well documented [
20‐
26]. Improving the inclusion of patients from these ‘at risk’ populations in COPD research is needed to address these disparities. While there is limited research on recruitment and retention of underrepresented groups for studies of people with asthma [
15,
27] there is little, if any, such research for people with COPD.
In designing and conducting a randomized controlled trial of health coaching for patients with moderate to severe COPD cared for in urban public health “safety-net” primary care clinics, we faced challenges to recruiting and retaining patients who were members of racial and ethnic minority groups, had low English proficiency, and/or had concurrent co-morbid, substance use or mental health conditions. In this paper, we report study design features intended to enhance recruitment and retention, as well as challenges to recruitment and retention that arose during the study, strategies used to meet these challenges, and rates of recruitment and retention of study.
Discussion
The AIR study employed a variety of techniques around stakeholder engagement, community partnering, building trust with patients and community and creating value for patients and clinics to maximize enrollment and retention of patients often underrepresented in clinical research due to economic, language and cultural barriers, or co-existing mental health, substance use or co-morbid disease. Our rates of recruitment (68%) and retention (82%) compare favorably to previous randomized controlled trials of similar interventions, including those with more extensive exclusion criteria. In three community-based studies which recruited patients hospitalized for COPD exacerbation, recruitment rates were 22%, 47%, and 50% and retention rates were 65%, 89%, and 92%, respectively [
47‐
49]. Another RCT of self-management support that recruited patients from primary care and specialty clinics in a large academic center reported a recruitment rate of 41% and a retention rate of 81% [
50‐
52]. Two studies that recruited from patients referred for pulmonary rehabilitation or to a COPD nurse, reported recruitment rates of 54% and 84%, respectively, and retention rates of 51% and 81% [
53,
54]. Another study, which recruited patients at the time they filled a prescription for a COPD medication, reported enrolling 69% of eligible patients with 95% completing the study [
55]. Notably, several studies had exclusion criteria that would have disqualified many patients from the AIR Health Coaching Study; for example, having diagnosis of other lung diseases, unstable cardiac, renal, or hepatic disease [
49] or having mental disorders, including schizophrenia, dementia, alcohol or drug abuse [
54].
While it is not possible to precisely link which features of the AIR study were most important for recruitment and retention, we believe, based on our experience in the AIR Study, that three strategies- incorporating patients’ perspectives, establishing community partnerships, and building rapport with patients - were key to reaching our goals for recruitment and retention. Each strategy is described below.
Incorporating patients’ perspectives
This study made a concerted effort to incorporate patients’ perspectives in multi-faceted approaches. As mentioned previously, the establishment of the advisory board, early discussions with patients from the community spirometry program, and most importantly, having a patient partner as a study team member throughout the entirety of the study helped the team view everything from the patient perspective. Having this perspective gave a realistic view of what was feasible or appropriate for the study population, such as survey length or how and where to interact with patients. Changes based on recommendations from patients made this study more sensitive to the needs and perspectives of the study population and may have helped the study avoid some of the stigma often associated with research among underrepresented groups.
Our study experience corroborates reports by others indicating that partnering with community clinics is vital to enrollment and retention of underrepresented populations [
56]. The community clinics with which we partnered are highly mission driven, so communicating the potential benefits of the study to patients (e.g., obtaining important clinical tests such as spirometry or the 6-min walk test, facilitating specialist review of cases, and providing health coaching to half of those enrolled) was critical in messaging. The study team also demonstrated their commitment to the partnership by supporting initiatives of the community clinics, such as participating in health fairs or offering in-service sessions and attending clinic huddles or holiday events. At the beginning of the study, the team committed to visiting each interested site to report on findings or offer additional training to support COPD patients after the end of data collection. The depth of this reciprocal commitment was vital to overcoming barriers of resource limitations that could otherwise have hobbled the study (e.g., finding space on busy days).
Building rapport with patients
Rapport is defined as a close and harmonious relationship in which the people or groups concerned are ‘in sync’ and is considered a therapeutic relationship between a doctor and patient [
57]. The initial appointments at baseline, lasting 2–3 h, allowed time for the research staff to build trust and rapport with participating patients. This time gave patients space to talk and ask questions. It also allowed the research staff to take their time in explaining the study, going through the survey and other measures, and allowed for breaks if needed. Having the same staff member from baseline reach out to the patient for follow up attempts may have promoted the likelihood of them showing up for follow up, which ultimately contributed to the retention rate. Establishing rapport with patients also helped to decrease the stigma often associated with research. Once patients felt they could trust the research staff, they were highly motivated to participate in the study and actually encouraged other patients to participate. Thus building rapport with individuals led to building rapport with the community. At the end of the study, all study participants and stakeholders were invited to a gathering to celebrate the successful completion of the study and to report results.
These three strategies are feasible, though each requires early planning and an investment of resources to implement. Incorporating patients’ perspectives requires identifying and meeting with patients, caregivers, and patient advocates. Often potential patient representatives can be identified by members of the clinic staff or from groups such as clinic advisory boards or disease-specific education and support groups. It is also important to advertise more widely to give patients a chance to self-nominate. Covering the cost of participation (e.g., transportation, childcare) and at least a nominal incentive is also important. The second strategy, developing a relationship with community partners such as clinic sites, also requires a sustained longitudinal commitment, preferably starting prior to the actual study. In our experience, most community clinic sites are committed to their patients and their patients’ communities. Participation in a research study may take resources, distract from patient care, or be perceived negatively by the community. Developing a positive relationship with one or more member of the clinic administrative and service staff early on is critical. Offering to help with existing activities such as health fairs or community outreach and providing educational or other resources can help build a relationship. Engaging clinic leadership in identifying ways the clinic could benefit from the study, and how to minimize the adverse impact of the study, is also important. The third strategy, building rapport with patients, can be rewarding for study staff and patients alike and does not require many additional resources. Choosing staff who have experience working with the study’s target groups is helpful, though not absolutely necessary. Providing opportunities for frequent contact between patients and the same member of the staff can help facilitate rapport.
Limitations
The lack of data available for those not enrolled in this study limits our ability to characterize these people are, what barriers they may face, and to identify strategies targeted to increase the representation of this population into research. This study was not designed to test different recruitment strategies. Thus, our results are primary descriptive and our judgements about the relative importance of patient participation, community partnerships, and building rapport are based on our experience.
Conclusions
While the AIR study included design features to maximize the recruitment and retention of patients from underrepresented groups, additional challenges were encountered and responded to during the study. Some challenges were specific to patients with pulmonary disease, such as the lack of spirometric testing. Others, such as poor health which limited access to the study, can be expected in other moderate to severe conditions. Still others, including lack of space at the clinic sites at which to meet with patients and limited patient contact information, can be found in many clinics serving underrepresented groups of patients. We hope that reporting our experience and the three key strategies we describe will help improve future studies of COPD and similar conditions. Adopting these approaches could help investigators achieve participation rates of at least 70%, providing results that can be more accurately applied to patients from underrepresented groups, who carry a disparate burden of COPD.
Acknowledgements
The study team would like to thank our clinic partners, patient advisory board, patient partners and study participants.