Reducing minority health disparities is a national health priority as reflected in the Healthy People 2010 initiatives (www.healthypeople.gov).1 In addition, the National Institutes of Health2 and the National Centers of Excellence in Women's Health3 have placed a strong emphasis on recruiting and retaining women and minorities, who have been largely underrepresented in health promotion and clinical trials research, in order to help reduce these health disparities. Despite this focus, there is little information guiding researchers in the evaluation of impediments to successful recruitment and retention among minority populations. Moreover, there is little known about successful strategies for younger African American women. Few studies describe recruiting strategies with older African American community-dwelling women,4-6 but these results are not generalizable because recruiting challenges for younger women may be very different.
Effective recruiting and retaining of female participants from minority groups are essential for research studies aimed at decreasing or eliminating health disparities. The purpose of this article is 2-fold: (1) to describe existing literature regarding recruitment and retention methods previously used with minority populations, and (2) to report the effectiveness of these methods to recruit and retain a sample of young (aged 18-45), inner-city African American women for participation in an 11-week physical activity intervention study.
The incidence of cardiovascular risk factors (high blood pressure, physical inactivity, obesity) and the prevalence of cardiovascular disease, hypertension, and diabetes are higher in African American women than in Caucasian women (American Heart Association [AHA]).7,8 Physical inactivity is a key risk factor that can be modified to reduce cardiovascular disease risk and improve health. The Healthy People 2010 initiatives1 specifically identify African American and Hispanic women as populations with low levels of physical activity. Fewer than half the women in the United States are involved in regular, sustained physical activity of sufficient duration and intensity to reduce risks of chronic diseases, and minority and older women, in particular, are among the least active.1 According to AHA statistics,7 the prevalence of physical inactivity is higher among African American compared with Caucasian women (55.1% vs 38.3%, respectively). Because health promotion efforts can reduce the prevalence of cardiovascular risk factors and disease progression, developing appropriate strategies to promote healthy lifestyles among women, and specifically minority women, is essential. Greater emphasis to include adequate numbers of minority women in clinical trails is important to help facilitate valid analyses and interpretation of health promotion interventions. Therefore, research with African American women is sorely needed.
Within the Chicago metropolitan area, 26.1% of the population is African American, double that of the US total census.9 African American women living in Chicago are quite heterogeneous, with varying demographics including socioeconomic status and level of acculturation based on community area. Wilcox et al10 found that when working with minority women, within-group differences were large and not just based on ethnicity alone. Differences based on socioeconomic status, level of acculturation, primary language spoken, and geographic residences make broad generalizations difficult.
Fitzgibbon et al11 reported different recruiting strategies used in 2 studies in the Chicago area involving 2 different African American populations. The first study, "Hip Hop to Health" recruited participants aged 6-10 and their caregivers, largely women, from a population with a median income of $12,000. In this group, direct presentation was the most effective recruiting strategy. Due to fear of exploitation, participants preferred contact by known and trusted community agencies. The second study, Fat Reduction Intervention Trial in African Americans (FRITAA), recruited households with a median income of $30,000. Alternatively, for this group, telephone recruiting and neighborhood canvassing were the most effective recruiting strategies.
Yancey et al5 described experiences recruiting African American women for a health promotion program entitled "Eating and Exercising for a Cancer-free Life" (EECL). By conducting a pilot study, the researchers were able to collect data and ethnically tailor the program targeted to African American women in urban Los Angeles. Findings showed that different recruiting strategies were needed based on age, economic status, educational attainment, and body size. For example, overweight women needed evidence that the program would benefit "women of their size" versus "leaner, thin women."
Lee et al12 evaluated the effectiveness of different strategies to recruit minority women to a home-based walking program. They categorized the strategies into active and passive methods. Active methods were those that targeted specific samples in defined areas, whereby researchers personally approached participants. Examples of active methods included telephone, face-to-face, and direct mailings. Passive methods were those that employed materials used to heighten the awareness of the target sample, thus providing a method for participants to contact the researchers. Examples of passive methods included newspaper advertisements and notices at targeted agencies. The results of this study were that passive methods were more cost-effective and yielded a larger number of participants than active methods (97 vs 29 participants, respectively).
Wilbur et al13 recruited a sample of women aged 45-85 for a study to test a walking program using active and passive recruiting methods. The primary recruiting sites were workplaces that employed varying ethnic and socioeconomic status groups close to the data collection site. Active recruiting strategies included personally meeting and enlisting help from management personnel at the workplace sites and church leaders to gain trust in the community. Continuous follow-up occurred during the data collection period. Passive recruiting methods included notices by e-mail to employees and flyers placed throughout the workplace. Recruiting was successful using these methodologies, but the success by method was not reported.
In conclusion, theses studies demonstrate that successful recruitment of African American women requires custom tailoring based on varying subgroup demographics. Socioeconomic factors, sense of mistrust, and participant age are key elements to consider when recruiting and retaining African American women. From study design conception, research questions and data collection instruments should be designed to meet the needs, experiences, and geographic location of the targeted minority population. The use of trusted community agencies and community leaders to gain trust was important for recruiting in the previously cited studies.
Barriers to Recruiting and Retaining Minority Populations
Researchers conceptualize quality research projects with the intention of recruiting minority populations. However, implications in the literature10,11,15 suggest that researchers cannot avoid some race and class bias in organizing research from their own value system, even when trying to be culturally sensitive to minorities. Development of culturally sensitive recruitment materials targeted to specific minority groups is an important detail often missed. Moreover, researchers may rely on preexisting data collection instruments, which may be more efficient, but not culturally sensitive, for the population being studied. There may not be funds, time, or resources for the labor-intense demands of tailored recruiting strategies designed to meet the specific needs, experiences, and environment of the target population.
Barriers to study recruiting have been identified,5,6,11,14,15 and should be considered (and possibly augmented) when designing studies involving minority populations. A combined, though not exhaustive, list of these barriers include:
- Research instruments and interventions lack tailoring for the target minority population;
- Research sites are distantly located, unfamiliar to participants, and hard to reach using public transportation resources;
- Research staff is untrained to be culturally sensitive and seems more interested in the research project rather than the participants' well-being.
The difficulty in retaining participants greatly increases the overall study length, presents threats to internal and external validity, and can deplete finances. Davis et al16 described a number of factors from community-based clinical trials that affect retention of participants. These included: (1)number and timing of follow-up visits; (2) treatment complexity, stress and emotional reaction of participants to the time for participation in longitudinal studies; (3) lack of research staff skills; and (4)inequitable incentives for continued participation.
Banks-Wallace and Conn17 found that attrition rates of African American women in short-term intervention studies including a physical activity component ranged from 9% to 47% (mean = 21%). Several reasons why women stopped participating included: (1) research staff working against participants' traditional views of health, (2) research staff not taking time to listen to participants' personal history, (3) multiple family roles imposing stress and burden on participants, and (4) traditional family beliefs (eg, high-fat cooking) that restricted participation in components of the study.
Lastly, participating in clinical trials often requires skills such as self-discipline, observing and repeating events, goal setting, and reading and writing on forms.18 These skills may be less familiar due to lack of formal education, thus making study continuation difficult for participants.
Successful Recruiting and Retaining Strategies
The Women's Health Initiative Workshop proceedings10 and several studies11-14 have documented successful recruiting and retaining strategies. These strategies are categorized into groups and examples of each strategy are provided below. Additionally, a few examples of how these strategies were employed are referenced.
- Importance of community:
- Place research centers where the target population interacts and have accessible transportation;
- Wilbur et al13 recruited women from employment sites located near the university testing center, that on average employed 43% to 48% African American women, such as hospitals, clinics, and a communications company. Banks-Wallace et al14 utilized frequented community centers. Both studies included sites easily accessible by bus and train.
- Publish articles in community and university newspapers;
- Lee et al12 placed articles in the Women, Infants, and Children newsletter and ran short notices for 2 weeks in the local newspaper. Sixty percent of women respondents were recruited using this strategy.
- Increase visibility by appearing on local cable stations and by public speaking at schools and community centers;
- Select an Advisory Committee of community leaders and residents;
- Fitzgibbon et al11 recruited 29 community residents including church ministers, doctors, educators, politicians, community activists, and study participants to serve on its advisory board.
- Hire research staff of the same ethnic group from the neighborhood;
- Orient research staff to day-to-day living of the community, such as competing priorities and community focus, cultural norms, family networks, and community attitudes.
- Establishing trust:
- Use mass mailings through trusted groups such as school programs;
- Fitzgibbon et al11 used a tutoring program to recruit children and their caregivers. Three separate mailings were sent to parents of children who attended the program. Twenty-three percent of children/caregiver pairs were recruited using this strategy.
- Work and be visible in the community by attending community functions;
- Use house-to-house canvassing and/or telephone recruiting with a community member;
- Fitzgibbon et al11 found telephone recruiting and neighborhood canvassing to be the best method of recruiting for their FRITAA participants. Seventy-two percent of the participants were recruited by canvassing, and 28% by telephone.
- Cultural sensitivity:
- Develop materials that are culturally and educationally sensitive so that participants envision themselves in all aspects of the research study (eg, pictures, appropriate snacks10);
- Use community consultants to ensure that research materials are culturally sensitive;
- Recognize and utilize intellectual contributions offered by participants, thus developing new knowledge.
- Strong sense of caring:
- Engage in personalized interactions face-to-face, by telephone, or personal notes, thus exhibiting a caring commitment to the participant as an individual;
- Nurture relationships by taking time to talk and hear participant's life stories;
- Use storytelling, a widely recognized means to building relationships, while imparting valuableinformation in the African American community.
- Banks-Wallace et al14 used this strategy to enroll 55% African American women in their physical activity study. Storytelling allowed the women to share their personal experiences and enhance the collaborative relationship for the study intervention.
Many of the same recruiting strategies, when carried through the research design, will enhance retaining participants, such as:
- Establish a project identity that has meaning to the target population;
- Emphasize the benefit of the study and its outcomes for the target population;
- Conduct a pilot study with the research instruments and educational materials to ensure cultural sensitivity, understanding, and applicability for the target population;
- Provide meaningful incentives suggested by the target population;
- Maintain frequent follow-up contacts;
- Provide culturally sensitive interpersonal training for research staff.
Application to Clinical Research
This section describes our experience in recruiting and retaining younger African American women while conducting an 11-week intervention study. The research design, recruiting and retaining strategies (both successful and unsuccessful), and results are described.
The purpose of this study was to examine the physiological effects of integrating lifestyle physical activity into the daily routine of hypertension-prone, sedentary African American women, aged 18 to 45. An 8-week, randomized, parallel-group, single-blind clinical trial was conducted in premenopausal, US-born African American women with blood pressure defined as high-normal (130-139/85-89 mm Hg) or untreated stage 1 hypertension (140-159/90-99 mm Hg) based on JNC6 guidelines, who consented to be randomized to 1 of 2 groups: (1) exercise or (2) no exercise. The recruiting goal was to randomize 16 participants to each group during a 12-month period. The trial consisted of 2 contiguous phases: (1) prerandomization visits (3 weeks of screening) and (2) follow-up clinic visits (8 weeks of intervention). Eligibility for the trial was determined during 3 screening visits (lasting ∼1 hour), 1-week apart.
The primary intervention was an 8-week individualized, home-based, physical activity program consisting of lifestyle physical activity. Women randomized to exercise were instructed to engage in lifestyle physical activity (eg, walking, stair climbing) for 10 minutes, 3 times/d, 5 d/wk at a prescribed heart rate corresponding to an intensity of 50-60% heart rate reserve. They were also instructed to keep a physical activity log and wear a portable heart rate monitor to verify their activity level. Women randomized to the no-exercise group were instructed to continue with their usual daily activities. All enrolled participants were seen every 2 weeks for follow-up visits (lasting ∼20-30 minutes) in the clinic to measure resting blood pressure and weight. Physical activity logs and heart rate monitors were evaluated for women randomized to the exercise group. Participants were provided a stipend to help offset travel and/or parking expenses.
A variety of recruiting strategies were used for potential candidates over a 19-month period. Based on the authors' experiences, these strategies are broadly grouped into 2 categories: active and passive recruiting methods as previously described. Active recruiting methods consisted of blood pressure screenings, health fairs, and physician referral. Passive recruiting methods consisted of flyers, advertisements, and other sources.
Five informal blood pressure screenings were conducted: 3 at the local university medical center and 2 at an inner-city YMCA. Those potentially eligible were given written information about the study, then followed-up with a telephone screening and clinic visit. Similar activities occurred at 6 community health fairs. Three health fairs specifically focused on women s health issues and 3 were nonspecific. Brochures describing the study were provided to encourage interested individuals to call the study telephone number to find out more about the research study. Several physicians were contacted directly to assist with direct referral. They were given information about the study, including inclusion and exclusion criteria and written materials to provide to potential candidates. Key physicians targeted included cardiologists, internists specializing in women's health, obstetricians/gynecologists, and pediatricians.
A number of passive recruiting strategies were implemented with flyers being the most widely used. Colorful flyers were created that contained bullet points highlighting the targeted study population, basic inclusion criteria, study duration, and telephone number to call for more information about the research study. Flyers were continuously posted at inner-city locations including 3 university/medical centers, 12 community colleges, 4 grocery stores, 2 train stations, 5 YMCAs, and 2 African American bookstores.
To broaden community involvement, flyers were mailed to 115 churches. A personalized letter was included describing the purpose of the study with a request to post flyers and brochures in the church bulletin and/or on the rectory bulletin board. Information was given emphasizing the need for increasing awareness in the African American community about high blood pressure and its complications, with an offer extended to provide information sessions at the church about high blood pressure and/or participate in health screenings. In an effort to capture locations where young African American women tended to gather, similar packets of information were mailed to 75 beauty/nail salons and 4 African American sororities encouraging interested women to call for more information about the program. Packets of information were also sent to 30 city health department clinics. Six months later, a follow-up mailing was sent to the clinics. Lastly, tabletop flyers were strategically placed in high-volume areas around the university medical center: such as cafeteria tables, waiting areas outside central registration, gift shops, main lobbies, and mammogram screening sites.
The second most widely used approach of passive recruiting included media advertisements via newspaper, radio, and Internet sources. With the hope to increase visibility to the largest numbers of women, several local newspapers were contacted with requests to run study-related advertisements. However, advertisements in 7 smaller newspapers catering to the African American community were used to specifically target the study population and because they were lest costly. Newspaper advertisements varied from daily, weekly, bimonthly, and monthly health information articles and were run multiple times over the 19-month recruiting phase. Newspaper advertisements were also placed in 2 local university newspapers, 1 park district newsletter, and the university medical center's online newsletter and quarterly "News Rounds" publications. For the most part, newspaper advertisements were designed similar to flyers describing the targeted population, basic inclusion criteria, study duration, and telephone number to call for more information, yet a few advertisements were more "newsy" and designed in a news-story fashion. Additionally, public service announcements were run on 2 African American radio stations and media advertisements were run via the Internet at the university medical center. It should be noted that a public relations specialist, who happened to be a young, African American female, was hired to help write and secure placement for 3 newspaper advertisements and 2 radio public service announcements.
Other utilized sources of passive recruiting consisted of talk radio, work-site presentations, and networking or word-of-mouth. One radio interview was conducted at a local diner catering to African Americans. The program was an informational session that provided an opportunity for listeners and participating attendants to gain knowledge about high blood pressure and ask study-related questions. Three work-site presentations catering to female workers were conducted at the university medical center and 2 at a city health department clinic. Similar to the talk radio program, these sessions provided information about high blood pressure and offered a forum to learn about the study. Another source of referral came from friends/coworkers of individuals who had been screened as part of the study or knew about the program. Continuous efforts using word-of-mouth as a source of referral were employed. Every woman screened was asked to distribute study flyers and brochures to their respective community, church, work-site, and/or school. Even women who were not necessarily study-eligible but knew someone who appeared eligible were asked to help distribute study-related materials.
Retaining study participants was a process of identifying and overcoming their personal barriers by developing strong relationships featuring unconditional regard and social support and assistance with problem solving. Among the most common barriers were distrust of research, inadequate time, and inadequate resources. The research staff was trained in options to overcome barriers such as being available by pager to answer questions and offering flexibility with scheduling clinic visits.
Recruiting and Retaining Results
The 19-month recruiting effort resulted in 136 inquiries. Thirty-five (26%) women were eligible for initial screening. Twenty-four (18%), or 72% of the originally projected sample size of 32, women were enrolled into the study. Table 1 shows the yield of eligible (ineligible) screenees obtained by recruiting strategy. Of the 35 eligible inquiries, 40% resulted from newspaper advertisements and 23% were referrals from friends or coworkers. Of the 101 ineligible inquiries, 45% resulted from posted flyers and 27% were from newspaper advertisements.
Table 2 shows the major reasons for exclusion. Ineligible screenees not meeting entry criteria included taking antihypertensive medication (29%), smoking (11%), blood pressure too low (9%), blood pressure too high (2%), beyond 45 years (9%), and male gender (9%). Other reasons for exclusion (eg, medical history, living out of state, unwilling to participate, physically active) accounted for 18% of the ineligibles. Sixteen percent of the ineligibles were excluded because they did not respond or complete the initial screening process.
Table 3 shows enrollment results by recruiting strategy. Of the 35 women eligible for initial screening, most of those enrolled were recruited by newspaper advertisements (46%) and flyers (21%), whereas newspaper advertisements (36%) and referral from friend/coworker (27%) were the most common strategies among women not enrolled. Age, educational attainment, relationship status, status of work plus other activities, and initial blood pressure are shown.
Enrolled and not-enrolled women were similar on most characteristics. There was no significant difference between groups with respect to age or educational attainment (Table 4). Comparing enrolled to not-enrolled women, most were single (50% vs 46%), working full-time (71% vs 64%), had childcare responsibilities in addition to working (42% vs 27%), and volunteered/participated in community-related service (79% vs 64%). At the initial screening visit, enrolled participants had higher blood pressures compared with those not enrolled, but only diastolic blood pressure was significantly different (134.9 vs 129.8 mm Hg; and 91.8 vs 85.2 mm Hg, P < .05). Once enrolled, the study retention rate was 96%. One participant in the exercise group dropped out due to work-related issues. Forty-four percent of the retained participants were recruited using newspaper advertisements and 22% were from flyers.
Recruiting 24 younger African American women with untreated, mildly elevated blood pressure for participation in a physical activity intervention study was challenging. Some recruiting/retaining strategies implemented were successful and may be useful for others conducting inner-city intervention trials. Other strategies were less successful and may need tailoring before meeting the needs of the target population.
Recruiting was time consuming. The initial 12-month recruiting period was extended to 19 months, thus acquiring 72% of the originally projected sample. Recruiting required continuous monitoring and reporting progress, and enormous flexibility on the part of the research staff to meet the conflicts and needs of the participants. Setting benchmarks early during the recruiting phase made it possible to adjust goals and strategies to successfully recruit. Given the limited research staff and financial resources of this study compared with other studies like the Women s Health Initiative,14 recruitment took more time but yielded comparative results with respect to projected sample size.
Newspaper advertisement was the most successful strategy capturing 40% of eligible responses and 46% randomized women. The success of this strategy may be related to the fact that newspaper advertisements were placed in community/university newspapers specifically targeting the desired population. This strategy has been used and documented in other studies.8,14 In addition, the ads were designed using culturally sensitive language and graphics, and news stories were written involving young African American women. Once this strategy was recognized as being successful, recruiting efforts (and the study budget) were adjusted to place more newspaper advertisements than originally planned.
Flyers were designed with culturally sensitive graphics and distributed throughout the community and university settings. This approach was inexpensive, easily distributed, and yielded a number of eligible responses (20%) and a greater number of randomized women (21%). Flyers were also mailed to targeted groups such as churches, beauty salons, and health clinics. The response from churches and beauty salons was exceptionally poor compared with university settings and health clinics. Multiple attempts were made to have a church pastor or parish nurse endorse the study, however, these efforts failed which may, in part, have resulted from the lack of a community consultant/recruitment coordinator visibly present at church sites. Working in the community and attending community functions, such as church events, may have increased visibility, thus establishing rapport and trust. Similarly, this may explain the lack of responses from beauty salons. Being a patron at selected beauty salons and talking with women one-on-one or in small groups may have added a personal touch, and thus, a more effective recruitment approach.
Direct referral from a friend or coworker was another successfully implemented recruiting strategy, capturing 23% of eligible responses and 17% enrolled women. Continuous efforts using word-of-mouth were utilized by asking women (and men) to distribute study-related materials in their respective communities. Women screened were asked to call a friend or family member to tell them about the study. This strategy helped overcome some of the traditional fears and reluctance of participating in research. African American women helped recruit each other by spreading "the word" in their community and worksite, thus serving to establish credibility and trust for the researchers as well as the research project.
Blood pressure screenings were time consuming with respect to logistics and resources (ie, finding the right location, proper equipment, adequate personnel), and the number of eligible responses was small (11%) compared with the amount of work/effort involved. Although blood pressure screenings can be a good method for gaining access to the community and reaching out to a large number of potential candidates, other methods (such as newspaper ads) were more efficient and less labor-intensive.
The study retention rate was high (96%), especially given the fact that most women worked and had other responsibilities such as childcare, and/or school, and/or community service. A number of strategies were implemented to accommodate these women. For example, to assist working women, early morning and some Saturday clinic visits were offered. Women were able to bring children to clinic visits and, when necessary, rescheduling clinic visits because of issues involving work or childcare was easily arranged. The research staff displayed a strong sense of caring by personalizing each interaction and showing interest in the woman as an individual. Often discussions at clinic visits would extend beyond study-related components and lead to matters about family and/or work, thus providing an opportunity for establishing trust and developing relationships. The research staff was available via pager to answer questions at any time.
Efforts were made to make women feel comfortable with the research staff and the environment. In addition, a stipend was provided at each clinic visit to defer the burden of travel and/or parking expenses. Frequent clinic visits and telephone calls may have contributed to the high retention rate in this population of young women with multiple responsibilities.
Recommendations for Future Research
Although this study was not designed using home visits for data collection, it may be considered for future studies. Home visits may provide an alternative (more convenient) for working women and/or women with small children. The incorporation of an advisory committee of community leaders and/or community consultants may be another consideration for future studies and in targeting a smaller, geographic area. Although an African American female public relations specialist assisted with some writing and placement of study advertisements, an advisory committee and/or community consultant may have been better suited for overcoming barriers at difficult recruiting venues, such as churches and beauty salons.
Recruiting and retaining young African American women to an inner-city intervention study required multiple strategies, time, and resources. Among women enrolled, the most successful strategies were newspaper advertisements and flyers. Newspaper advertisements were more expensive but required less time and effort than distributing flyers. Surprisingly, strategies targeting churches and beauty salons were not successful and may require one-on-one interaction between research staff and the community site. Referral from a friend and/or coworker was easily implemented and provided a number of eligible responses, but a small number of enrolled women. Although blood pressure screenings were a good way to reach out to the community, it provided a smaller number of enrolled participants for the amount of time expended. The retention rate for this study was high compared with other intervention studies. Flexible scheduling, frequent contact via direct visits or telephone, and a caring and nurturing relationship may have contributed to retaining these very busy women in the study.