Participation in clinical research by minority group members is essential for eliminating health and health care disparities in the United States (National Institute of Nursing Research, 2011). Researchers have published their challenges and solutions for recruiting vulnerable populations (e.g., Fischer, Burgener, Kavanaugh, Ryan, & Keenan, 2012), including several articles appearing in a special issue of Research in Gerontological Nursing (described in Buckwalter, 2009). Many articles have included a special focus on recruiting elder adults who are Latino or of Mexican descent (Berger, Begun, & Otto-Salaj, 2009; Calderón et al., 2006; Gonzalez, Gardner, & Murasko, 2007; Jackson et al., 2009; Loue & Sajatovic, 2008). Focus has been on barriers and motivators for participation in research. However, recruitment of minority group members continues to be a challenge. This article demonstrates how we faced challenges for recruitment using evidence-based strategies, under especially challenging circumstances, recruiting Mexican American elder adult-family caregiver dyads for an intervention during elder adults’ brief hospitalizations.
The purpose of the randomized control trial (RCT) was to test the efficacy of a telenovela intervention designed to increase use of home health care services (HHCS) by Mexican American elder adults. Eligibility criteria for elder adults were: age 55 or older, of Mexican descent, able to read or speak Spanish or English, score 6 or more on the Short Portable Mental Status Questionnaire (Pfeiffer, 1975), hospitalized, referred for HHCS, receiving daily assistance with activities of living (ADLs) or instrumental ADLs, having a medical or surgical diagnosis, and having a family caregiver who either resides with the elder adult or lives within a 30-minute drive of the elder adult’s home. Criteria for caregivers were: a primary caregiver self-identified or identified by the elder adult, of Mexican descent, age 18 or older, able to speak or read Spanish or English, and reside with or reside within 30 minutes of the elder adult.
To generate initial contacts, the principal investigator (PI) and bilingual, bicultural project director (PD) introduced the study to various hospital staff, including case managers, nursing staff, physicians, and social workers. The PD followed protocols approved by the university’s institutional review board (IRB) to identify qualifying individuals on adult inpatient units. For elder adults who qualified, the PD asked staff (e.g., nurses, unit technicians, other hospital personnel) to request permission to talk with potential participants about the study by reading the recruitment script/flyer. If potential participants gave permission, the PD described the study and asked the elder adult, and the caregiver if present, if he or she were interested in participating. If the family caregiver was not present, the PD followed up with the elder adult or dyad later the same day or the next day. As soon as the elder adult and caregiver verbally agreed to participate, the PD conducted the consent process.
After 1 month of daily attempts to recruit participants (December 2009 to January 2010), only one eligible dyad had been successfully recruited. We made certain immediate changes that helped us start to recruit more dyads. Although we continued to meet challenges, we successfully recruited 68 dyads during the following 15 months. The goal was to recruit at least 60 dyads (72 to allow for attrition) for the 2-year study. The 5,818 dyads that were assessed are depicted in Figure 1.
Figure 1. Flow diagram of elder adult-caregiver dyad recruitment for the study.Note. SPMSQ = Short Portable Mental Status Questionnaire.
Challenges and Strategies at The Hospital and Individual Levels
HHCS Referrals Written Late During Hospitalization. Initially, the major and immediate challenge was that the actual referral to prescribe HHCS was written at the last minute, just before discharge, precluding our chance to administer the two intervention doses. Each intervention consisted of showing a 12-minute telenovela (a dramatized televised story), coupled with a 50-minute guided dialogue. Also, per protocol, we needed time to provide both doses of the intervention. Current hospital stays are typically very short (2 to 3 days). Thus, some dyads who met all criteria could not be recruited in time for us to provide two intervention doses (reported within the “discharged too soon” category of Figure 1, but not counted separately). In addition, hospital staff rarely took the time to alert the PD if an apparently eligible elder adult who was likely to receive a HHCS referral was admitted to their unit. Others have reported, as we found, that nurses usually did not have the time to initiate study recruitment (Bishop, 2009).
Hospital Staff Reluctance to Assist With Entrée. Recruiters found staff members uncomfortable with or too busy to ask patients for permission to enter the room to explain the study. Resistance by staff to approach patients for recruiters to talk with them has been acknowledged by other researchers (Weierbach, Glick, Fletcher, Rowlands, & Lyder, 2010).
Other Challenges. Interruptions included routine nursing care, primary care practitioner visits, meals, procedures, and tests, often not previously scheduled. Another challenge was the number of patients who were unable to participate in the study because they were in isolation status. The risk of non-nursing research staff’s inadvertently spreading infection during various recruitment and research procedures precluded participation by these patients. It is possible the sample was skewed by excluding these patients, but they were not tracked.
Change of Study Protocols. We changed original study protocols after the first month of attempting to recruit and having recruited only one dyad. First, we deleted the inclusion criterion of elder adults’ having been referred for HHCS, although we knew this would restrict our sample size for addressing the long-term effects of the intervention. Second, we deleted the requirement to recruit participants only if there would be time for two doses of the intervention (Figure 1 shows both totals.)
We also increased our team staff dedicated to the recruitment and consent process. We trained our interventionist graduate research assistant (a 25% employee whose original responsibility was to provide the intervention) to recruit and consent. We trained other non-intervention team members to conduct the same processes, as well as baseline and posttests, to decrease the waiting time for participation. We also hired two additional bicultural recruitment personnel, using indirect funds, spending approximately $4,000 above what was budgeted.
In the instances when staff members were reluctant to request permission for entrée to speak with patients, we explained this IRB requirement. We also reminded hospital staff to use the IRB-approved recruitment script. Staff became more cooperative as we increased our visibility and branded the study, as described below.
Increasing Visibility of the Study. The PI and PD continued to attend staff meetings and present updates on the study. The PI attended monthly staff meetings on each unit. She brought snacks to these meetings to increase goodwill and show appreciation to staff for helping the team. Sometimes the PI and recruiter role-played requesting assistance and a staff member assisting us with gaining entrée into the patient’s room. These presentations were kept very brief (i.e., 3 to 5 minutes), just long enough to stimulate interest and promote goodwill.
The PI frequently accompanied the recruiters to increase visibility and further legitimize their activities. The nurse chief executive officer of the onsite hospital distributed flyers at all-hospital nurse retreats to show support. The directors of case management at both hospitals similarly endorsed the study to their hospital and management staff members. Other researchers’ methods have supported our findings that in response to organizational challenges, education and support by administration were critical (Mackin et al., 2009). We presented certificates of appreciation with formal letters, as well as copies to their managers, to staff members who had been helpful. This continued to keep our study visible to both staff members and administration. We met with nursing students and nursing instructors who had clinical assignments at the onsite hospital, explained the participant criteria, and requested they remind their preceptors to contact the PD if they identified qualifying participants. Additionally, we screened out elder adults in isolation as soon as we knew this status, to prevent staff’s and our team’s spending time on dyads we should not recruit.
Branding the Study. Branding was a way to both increase visibility and legitimize our efforts. For example, we added ENCASA (Elder aNd Caregiver Assistance and Support At-home) team members’ photographs to all flyers and also adhered the flyers to the boxes of treats we brought to staff. After an addendum approval by the IRB, we added a brief invitation to learn more about the study to the onsite hospital’s electronic sign, located both at the hospital entrance and near the cafeteria (Figure 2). We printed “business cards” with study information to distribute to nurses, social workers, case managers, students, and faculty and placed copies of the consent forms and the PD’s and PI’s business cards, which we gave to participants, in attractive school-color College of Nursing packets. We purchased “College of Nursing” polo shirts in the university’s color for each team member involved in recruitment to wear. Other researchers have also reported success through increased visibility and branding, for example, bringing staff snacks with attached messages about the study (Weierbach et al., 2010) and recognizing helpful staff with acknowledgement to their supervisors (Segre, Buckwalter, & Friedmann, 2011).
Figure 2. A brief invitation to learn more about the study was posted on the onsite hospital’s electronic sign, located both at the hospital entrance and near the cafeteria.
Ill Elder Adults. Ill elder adults understandably experienced low energy capacity, endurance, pain, and/or fatigue when returning from treatments (e.g., dialysis) or were heavily medicated to control pain. In addition, normal changes of aging (e.g., hearing impairment) and comorbid conditions made it difficult for them to stay alert or even awake and for us to start or continue the screening and consent process. These challenges would require the recruiter to postpone the recruitment session and return later. Also, sometimes physicians, staff members, and caregivers believed the research procedures might tire the patient and would discourage participation. Our observation that family members sometimes acted as protective gatekeepers has been reported by others (Mackin et al., 2009; Sisk et al., 2008).
Minority Elder Adults. Potential participants being of a minority posed challenges to recruitment in response to a general distrust of research, given today’s political context related to immigration, especially in the southwestern United States where Arizona Senate Bill 1070, which requires law enforcement agencies to check immigration status of individuals they encounter (Immigration Policy Center, n.d.), caused some patients to be afraid of being recruited into the study. Anti-immigrant sentiments had been expressed fervently in the public/political discourse (Johnson, 2010). At least six staff members and three patients or caregivers overtly expressed concerns about prejudice, racial profiling, or immigration status issues as reasons for not wanting to participate. There were likely other potential participants who chose not to divulge these concerns. Fear of discrimination has been reported as a major reason for minority individuals not participating in studies (Lopez, Morin, & Taylor, 2010), as well as a major reason for not using health services such as HHCS (Barr & Wanat, 2005; Ennis, Rios-Vargas, & Albert, 2011).
Elder Adult-Caregiver Dyads. Recruitment was time consuming partly because both individuals needed to agree to participate in the study before we could recruit either one. This required additional planning and coordination through telephone messages and waiting at the hospital for family caregivers to arrive between work and other obligations.
Pacing, Including Caregivers, and Reiterating the Study. In consideration for ill elder adults, we paced the recruitment presentation and the intervention according to their need for rest and recuperation. For example, we interrupted the recruitment or consent process if the elder adult needed to rest, eat, or use the telephone. When possible, we included and enlisted other family members to encourage participation. We also found that we frequently needed to respond to questions about whether we provided direct services.
Employing Personalismo and Respeto. We consulted with the ENCASA Community Advisory Council as an initial strategy in response to the pervasive political climate. We were consistent in our approach in clarifying our purpose to both staff members and elder adult-caregiver dyads, modeling respect rather than discrimination in the politically edgy context of our border state. Language was an important factor in recruiting monolingual Spanish-speaking elder adults. Others have also found that being able to present the study in Spanish was very important (Berger et al., 2009; Jackson et al., 2009).
The recruitment team used culturally appropriate methods to establish rapport and build a trusting relationship with potential elder adult and caregiver participants (Murphy et al., 2007). We employed personalismo, showing human-to-human regard and friendliness, taking the time to establish trust by giving personal attention (Espino & Oakes, 2007), during every contact. This also includes demonstrating respeto, respectful deportment, including, but not limited to, polite manners; and cultural humility (constant self-reflection to acknowledge, respect, and accommodate diversity; National Alliance for Hispanic Health, 2012).
Adjusting Our Timeline. We took the time and effort to contact family caregivers if they were not at the hospital during our regular recruitment hours. Team members stayed late to accommodate caregivers’ work schedules.
Proactive Rather than Passive Recruitment
We recruited 100% of our dyads using face-to-face (proactive) recruitment, as opposed to any potential participants’ responding to flyers (reactive recruitment). This was the norm as reported by other researchers with similar populations (Calderón et al., 2006; Larson, Ferng, Wong-McLoughlin, & Wang, 2009; Mackin et al., 2009). None of our dyads were recruited by anyone other than research team members. A few hospital staff (i.e., one case manager, two residents, and one physician) attempted to initiate recruitment in five cases, by telling us about potentially eligible participants. However, each of these potential participants either did not agree for us to approach them and explain the study or did not meet criteria. Other studies also found that potential participants did not self-initiate participation. Toobert et al. (2010) noted that this may be a positive occurrence because self-selected samples could be biased.
Recruitment was facilitated by bilingual, bicultural research team members’ pointing out to potential recruits that their participation would benefit the shared community by helping educate the community about HHCS as a resource for keeping elder adults at home. Participants volunteered comments to the PD that helping the community was more of an incentive than the monetary gift. Other studies with Hispanic individuals also reported that the purpose of the study (to improve the health of the Mexican American community) was rated as more important than other aspects (Kneipp, Lutz, & Means, 2009; Larson et al., 2009). Also, the necessity to reiterate that the study did not provide direct services occurred in previous studies (Crist, Parsons, Warner Robbins, Mullins, & Espinosa, 2009).
Although combined strategies provided overall recruitment success, we have no measure for identifying which specific strategies were more successful. For example, we cannot determine whether the electronic sign improved recruitment.
Eliminating the recruitment criterion that the elder adult be referred for HHCS was a trade-off, enabling us to recruit more participants but limiting the sample size for evaluating the intervention. Relaxing this criterion and including elder adults not referred for HHCS decreased our final subsample for assessing whether the intervention increased HHCS use. This subsample totaled 16 of 68 dyads (24%), making it difficult to assess the distal outcome of the intervention (increased use of HHCS), although we were able to analyze pre- and posttest data on other variables. However, the other protocol change, deleting the second dosage, did not show differential effects. Buckwalter (2009) indicated that one third of all clinical trials fail to recruit projected sample numbers, sometimes because of overly restrictive recruitment criteria. Our future studies will now be able to benefit from the established partnerships with key personnel in the recruitment setting, in order to facilitate earlier identification and recruitment of specific samples (Weierbach et al., 2010), in our case, elder adults who will receive HHCS referrals.
As we have done in past studies, we used community-based participatory research, asking for input and collaborating with the “community” (Crist & Escandón-Domínguez, 2003), in this case, in the hospital setting. We viewed inpatients, their families, and staff at the hospital as the community with which to collaborate (Crist et al., 2009).
We enriched our partnerships by inviting elder adults and caregivers who expressed opinions about the telenovela or concerns about health care disparities to join the ENCASA Community Advisory Council after they completed their 6-month participation in the study. We also invited interested nurse managers and unit technicians to join the ENCASA council. We have found their combined input to be a very important addition to our partnership. Other research supports the success of emphasizing the benefits to the community of scientific research in building trust and enhancing recruitment (Barr & Wanat, 2005; Larkey, Gonzalez, Mar, & Glantz, 2009).
Budgets for recruitment are rarely reported (Berger et al., 2009; Buckwalter, 2009), although Toobert et al. (2010) costed out their recruitment expenses by participant. Some of our unexpected costs were not funded. One example is that recruiters added minutes used on their own personal cell phones to contact each other to ensure timely and smooth transitions between the recruitment, data collection, and intervention steps. Costs can be better anticipated in future studies. Also, planning for time needed to establish initial partnerships and support of the study at all levels will be necessary. Additionally, we needed to take a no-cost extension for the study because of our slow start in recruiting for the longitudinal study.
Our feasibility study was instructive, reminding us that strategies should include accurately estimating numbers of eligible participants and timelines, consistently branding, using personalismo in every elder adult-caregiver encounter, continuing a community-based participatory research approach, and having an adequate recruitment budget. Such strategies are essential to meet recruitment goals for RCTs with vulnerable populations. In this way, interventions to eliminate disparities can be supported by adequate samples, which will provide significant and applicable data.
- Barr, D.A. & Wanat, S.F. (2005). Listening to patients: Cultural and linguistic barriers to health care access. Family Medicine, 37, 199–204.
- Berger, L.K., Begun, A.L. & Otto-Salaj, L.L. (2009). Participant recruitment in intervention research: Scientific integrity and cost-effective strategies. International Journal of Social Research Methodology, 12, 79–92. doi:10.1080/13645570701606077 [CrossRef]
- Bishop, V. (2009). Comment: A qualitative study exploring practice nurses’ experience of participating in a primary care-based randomized controlled trial. Journal of Research in Nursing, 14, 449–450. doi:10.1177/1744987108098229 [CrossRef]
- Buckwalter, K.C. (2009). Recruitment of older adults: An ongoing challenge. Research in Gerontological Nursing, 2, 265–266. doi:10.3928/19404921-20090816-01 [CrossRef]
- Calderón, J.L., Baker, R.S., Fabrega, H., Conde, J.G., Hays, R.D., Fleming, E. & Norris, K. (2006). An ethno-medical perspective on research participation: A qualitative pilot study. Medscape General Medicine, 8(2), 23. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1785211/
- Crist, J.D. & Escandón-Domínguez, S. (2003). Identifying, recruiting and sustaining Mexican American community partnerships. Journal of Transcultural Nursing, 14, 266–271. doi:10.1177/1043659603014003013 [CrossRef]
- Crist, J.D., Parsons, M.L., Warner Robbins, C., Mullins, M.V. & Espinosa, Y.M. (2009). Pragmatic action research with two vulnerable populations: Mexican American elders and formerly incarcerated women. Family and Community Health, 32, 320–329. doi:10.1097/FCH.0b013e3181b91f72 [CrossRef]
- Ennis, S.R., Rios-Vargas, M. & Albert, N.G. (2011). The Hispanic population: 2010. 2010 Census briefs. Retrieved from http://www.census.gov/prod/cen2010/briefs/c2010br-04.pdf
- Espino, D.V. & Oakes, L. (2007). Access issues in the care of Mexican-origin elders: A clinical perspective. In Angel, J.L. & Whitfield, K.E. (Eds.), The health of aging Hispanics (pp. 195–201). New York: Springer. doi:10.1007/978-0-387-47208-9_14 [CrossRef]
- Fischer, D.J., Burgener, S.C., Kavanaugh, K., Ryan, C. & Keenan, G. (2012). Conducting research with end-of-life populations: Overcoming recruitment challenges when working with clinical agencies. Applied Nursing Research, 25, 258–263. doi:10.1016/j.apnr.2011.04.002 [CrossRef]
- Gonzalez, E.W., Gardner, E.M. & Murasko, D. (2007). Recruitment and retention of older adults in influenza immunization study. Journal of Cultural Diversity, 14, 81–87.
- Immigration Policy Center. (n.d.). Arizona SB 1070, legal challenges and economic realities. Retrieved from http://www.immigrationpolicy.org/clearinghouse/litigation-issue-pages/arizona-sb-1070%E2%80%8E-legal-challenges-and-economic-realities
- Jackson, J., Mandel, D., Blanchard, J., Carlson, M., Cherry, B., Azen, S. & Clark, F. (2009). Confronting challenges in intervention research with ethnically diverse older adults: The USC Well Elderly II Trial. Clinical Trials, 6, 90–101. doi:10.1177/1740774508101191 [CrossRef]
- Johnson, J. (2010, September2). Justice Department sues Arizona’s Sheriff Joe. Newser. Retrieved from http://www.newser.com/story/99611/justice-department-sues-arizonas-sheriff-joe.html
- Kneipp, S.M., Lutz, B.L. & Means, D. (2009). Reasons for enrollment, the informed consent process, and trust among low-income women participating in a community-based participatory research study. Public Health Nursing, 26, 362–369. doi:10.1111/j.1525-1446.2009.00791.x [CrossRef]
- Larkey, L.K., Gonzalez, J.A., Mar, L.E. & Glantz, N. (2009). Latina recruitment for cancer prevention education via community based participatory research strategies. Contemporary Clinical Trials, 30, 47–54. doi:10.1016/j.cct.2008.08.003 [CrossRef]
- Larson, E.L., Ferng, Y., Wong-McLoughlin, J. & Wang, S. (2009). Retention and protocol adherence of Hispanic volunteers in a longitudinal trial. American Journal of Health Behavior, 33, 435–444. doi:10.5993/AJHB.33.4.9 [CrossRef]
- Lopez, M.H., Morin, R. & Taylor, R. (2010). Illegal immigration backlash worries, divides Latinos. Retrieved from the Pew Hispanic Center website: http://www.pewhispanic.org/2010/10/28/illegal-immigration-backlash-worries-divides-latinos/
- Loue, S. & Sajatovic, M. (2008). Research with severely mentally ill Latinas: Successful recruitment and retention strategies. Journal of Immigrant Minor Health, 10, 145–153. doi:10.1007/s10903-007-9063-9 [CrossRef]
- Mackin, M.L., Herr, K., Bergen-Jackson, K., Fine, P., Forcucci, C. & Sanders, S. (2009). Research participation by older adults at end of life: Barriers and solutions. Research in Gerontological Nursing, 2, 162–171. doi:10.3928/19404921-20090421-05 [CrossRef]
- Murphy, M.R., Escamilla, M.I., Blackwell, P.H., Lucke, K.T., Miner-Williams, D., Shaw, V. & Lewis, S.L. (2007). Assessment of caregivers’ willingness to participate in an intervention research study. Research in Nursing & Health, 30, 347–355. doi:10.1002/nur.20186 [CrossRef]
- National Alliance for Hispanic Health. (2012). Delivering health care to Hispanics manual and workbook (4th ed.). Retrieved from http://www.hispanichealth.org/resource/publications.aspx
- National Institute of Nursing Research. (2011). Bringing science to life: NINR strategic plan (NIH Publication No. 11-7783). Retrieved from http://www.ninr.nih.gov/NR/rdonlyres/8BE21801-0C52-44C2-9EEA-142483657FB1/0/NINR_StratPlan_F2_508.pdf
- Pfeiffer, E. (1975). A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. Journal of the American Geriatrics Society, 23, 433–439.
- Segre, L.S., Buckwalter, K.C. & Friedmann, M. (2011). Strategies to engage clinical staff in subject recruitment. Journal of Research in Nursing, 16, 321–332. doi:10.1177/1744987110387475 [CrossRef]
- Sisk, J.E., Horowitz, C.R., Wang, J.J., McLaughlin, M.A., Hebert, P.I. & Tuzzio, L. (2008). The success of recruiting minorities, women, and elderly into a randomized controlled effectiveness trial. Mt. Sinai Journal of Medicine, 75(1), 37–43. doi:10.1002/msj.20014 [CrossRef]
- Toobert, D.J., Strycker, L.A., Glasgow, R.E., Osuna, D., Doty, A.T., Barrera, M. & Ritzwoller, D.P. (2010). ¡Viva bien!: Overcoming recruitment challenges in a multiple-risk-factor diabetes trial. American Journal of Health Behavior, 34, 432–441. doi:10.5993/AJHB.34.4.5 [CrossRef]
- Weierbach, F.M., Glick, D.F., Fletcher, K., Rowlands, A. & Lyder, C.H. (2010). Nursing research and participant recruitment: Organizational challenges and strategies. Journal of Nursing Administration, 40, 43–48. doi:10.1097/NNA.0b013e3181c97afb [CrossRef]