Recruitment of racial/ethnic minority older adults through community sites for focus group discussions

The context for this recruitment research (in this case, our intervention) is our experience in conducting one of the social science components of an ongoing interdisciplinary study funded by the National Institute of Dental and Craniofacial Research (NIDCR) and the Office of Behavioral and Social Sciences Research (OBSSR) of the US National Institutes of Health (NIH). This larger study integrates social and systems science approaches to promote health equity, and is guided by the “Ecological model of social determinants of oral health for older adults.” A variety of methodological approaches (qualitative, quantitative, spatial, and systems science) have proved useful in thinking through pathways whereby social determinants at various scales influence oral health and related health outcomes, toward promoting healthy aging [25]. A simplified “Ecological model of social determinants of oral health for older adults” [25] is presented as Fig. 1, where selected factors at the societal, community, interpersonal, and individual scales are included that bear directly on oral health and related health outcomes for both individuals and populations of older adults.

The recruitment plan builds upon previous research and practice initiatives of the authors and their colleagues. Chief among these is the ElderSmile program, a community-based initiative sponsored by the Columbia University College of Dental Medicine to promote oral and general health and conduct health screenings in places where older adults gather, including senior centers [26, 27]. In an earlier NIH-funded study conducted with racial/ethnic minority older adults who attend senior centers in underserved urban neighborhoods, we found lower levels of tooth loss than in US national samples [25]. In order to learn from what is working well in this socioeconomically disadvantaged population, we obtained subsequent NIH funding to conduct social science research, including focus group discussions with racial/ethnic minority older adults intended to gather information about factors at the community, interpersonal, and individual scales toward enhancing community- and clinic-based service delivery and improving oral and general health outcomes for older adults.

To be selected to participate in the focus group discussions, participants (individuals involved) needed to meet the following criteria: (1) aged 50 years and older; (2) attended a senior center in northern Manhattan, New York, NY; (3) speak fluent English or Spanish; and (4) self-identify as African American, Dominican or Puerto Rican. To ensure geographic representation of northern Manhattan, approximately equal numbers of participants were recruited from senior centers in each of three northern Manhattan neighborhoods: Central Harlem (largely African American), Washington Heights/Inwood (largely Dominican), and East Harlem (largely Puerto Rican).

It is sound scientific practice to segment focus groups based on important characteristics that may influence either the issues discussed or the ability of the members to build rapport [28]. For analytic purposes, however, conclusions cannot be made on one or two groups per demographic segment. Thus, there was an inherent tension between simplifying the recruitment strategy and maximizing the amount of information that could be obtained with the fewest number of focus groups overall.

In the end, a total of 24 focus groups were conducted: 12 groups of men, and 12 groups of women. In each gender block, four groups were conducted with African Americans, four groups with Puerto Ricans, and four groups with Dominicans. Ten of the groups were conducted in English and 14 were conducted in Spanish (two of the Puerto Rican groups were conducted in English). Within each gender/racial/ethnic group, half of the groups were conducted with men or women who had visited a dentist in the past year, and the other half were conducted with men or women who had not visited a dentist in the past year (see Table 1).

The investigators elected to only include older adults who self-identified as African American, Dominican, and Puerto Rican, as these are the three largest racial/ethnic groups in northern Manhattan. It was decided that segmentation on the basis of cultural identity was important to build rapport among focus group members.

Both field recruiters (implementation leaders) were bilingual (English/Spanish) and had several years of experience working with racial/ethnic minority older adults and senior center directors in northern Manhattan as the program staff for the ElderSmile network. The recruitment staff visited geographically diverse senior centers in northern Manhattan and directly approached older adults to explain the study, screen for eligibility, and solicit their participation in the focus groups (individuals involved). To achieve the desired group size of eight to ten participants, 14–16 potential participants were recruited to participate in each group. This over-recruiting was deemed necessary since focus groups frequently have poor attendance rates [29]. Based on field experience (executing the implementation), the recruitment plan was adapted by the study investigators and recruitment staff after engaged discussion of challenges posed and potential solutions at research team meetings (reflecting on the progress of the implementation), as explained next.

The ElderSmile network of prevention centers were used as recruitment sites for focus group participants, and consisted of four center types: (1) Naturally Occurring Retirement Community; (2) US Department of Housing and Urban Development Section 202 Supportive Housing for the Elderly Program; (3) New York City Housing Authority; and (4) community senior or resource center [30]. No center director (opinion leader) contacted refused to allow focus group recruitment at any site, although the level of support offered during the recruitment period varied. According to the field recruiters, the most effective directors were “warm and welcoming” (engaging) and provided direct assistance to the recruitment staff, such as suggesting potential participants and opportune times for recruitment, making announcements asking for cooperation with the recruiters, and providing private space for screening whenever possible. We refer to these directors as change agents.

The populations of the centers also varied [30]. Some participants traveled by public transportation from the Bronx because they had been attending centers in Harlem but had moved; other centers drew participants from multiple Manhattan neighborhoods [30]. Certain centers were more racially/ethnically diverse than others, while some had younger cohorts of attendees [30]. On recruitment days, some centers were full, and some were virtually empty. Attendance was, of course, weather-related as well. Certain centers had attendees who were transitioning to care for the disabled or with cognitive difficulties who were not eligible for participation in the focus group discussions. Spaces within centers for making contact with and screening potential participants also varied, raising concerns around privacy and communication, especially with hearing impaired older adults.

Two venues (inner settings) were identified for the focus group sessions--one in Harlem and one in Washington Heights--to better ensure that participants did not need to travel more than a mile by provided taxis from their residential neighborhoods. After several months, however, all groups were held in the Washington Heights location, as the Harlem location proved logistically confusing for participants, since it was situated on the third floor of an office building. The focus group participants were delighted to learn that the Washington Heights location was a community space, which offered general health information via pamphlets and the use of computers to the public. Certain participants even inquired at the front desk about services offered or brought home reading materials.

While the recruitment phase of the study was delayed by 6 months, the flexibility of the recruitment plan, the expertise of the research team members, the perseverance of the recruitment staff (implementation leaders), and the cultivation of change agents ultimately resulted in successfully executing the recruitment plan and meeting the study targets for enrollment in terms of both numbers of groups and numbers of participants (see Table 2). The original plan was to recruit participants, and then fill focus group slots. Nonetheless, when several months elapsed after determining the eligibility of older adults for the study, those recruited were sometimes no longer interested in participating or no longer available. Instead, the decision was made to recruit participants and fill focus group slots in an ongoing way. There was an initial 6-month delay in arriving at this solution, and then all of the focus groups were completed in the planned 2 years.

The effort required to meet these targets was substantial. Of 625 older adults screened, a total of 194 African American, Dominican, and Puerto Rican older adults participated in 24 focus group discussions (see Fig. 2).

Potential demographic differences were examined between the 194 focus group participants and the 76 no show, the 142 not able to be scheduled, and the 79 not interested older adults (the 73 who were never interested were combined with the 6 who were not interested when contacted for a focus group to form a single not interested subgroup). Variables were created for each of these subgroups. Pearson’s chi-squared test was used to examine differences between the participants and the no show, not able to be scheduled, and not interested older adults, respectively. All calculations assume 2-sided tests with a significance level of alpha = 0.05. The only statistically significant difference between the participants and the no show older adults is that those without a dental visit in the past year were more likely to be no show older adults than were participants (p = 0.01). The only statistically significant difference between participants and the not able to be scheduled older adults is that adults who were 65 years or older were more likely to be not able to be scheduled older adults than were participants (p = 0.01).

On the other hand, there were several statistically significant differences between those who were not interested older adults and participants. Older adults who spoke both English and Spanish were less likely to be not interested than were participants (p = 0.05). African Americans were marginally (p = 0.06) more likely to be not interested than were participants. Those without a dental visit in the past year were marginally (p = 0.06) more likely to be not interested than were participants. Adults aged 65 years and older were more likely to be not interested than were participants (p = 0.03). Finally, Central Harlem residents were more likely to be not interested than were participants (p = 0.01). These differences are consistent with the experiences of the recruitment staff in filling focus group slots. Analyses are available upon request from the authors.

Analysis of the focus group discussions is providing important information for modeling activities toward improving community- and clinic-based service delivery and guiding policies to advance health equity for older adults. In addition, reflecting upon the implementation of the recruitment plan yielded the following insights for overcoming barriers to the enrollment of racial/ethnic minority older adults that may prove transferable or adaptable to other contexts.