Recruitment of Older Adults: Success May Be in the Details

Abstract

Purpose:

Describe recruitment strategies used in a randomized clinical trial of a behavioral prospective memory intervention to improve medication adherence for older adults taking antihypertensive medication.

Results:

Recruitment strategies represent 4 themes: accessing an appropriate population, communication and trust-building, providing comfort and security, and expressing gratitude. Recruitment activities resulted in 276 participants with a mean age of 76.32 years, and study enrollment included 207 women, 69 men, and 54 persons representing ethnic minorities. Recruitment success was linked to cultivating relationships with community-based organizations, face-to-face contact with potential study participants, and providing service (e.g., blood pressure checks) as an access point to eligible participants. Seventy-two percent of potential participants who completed a follow-up call and met eligibility criteria were enrolled in the study. The attrition rate was 14.34%.

Implications:

The projected increase in the number of older adults intensifies the need to study interventions that improve health outcomes. The challenge is to recruit sufficient numbers of participants who are also representative of older adults to test these interventions. Failing to recruit a sufficient and representative sample can compromise statistical power and the generalizability of study findings.

Insufficient recruitment is a major reason for failure in clinical trials and can contribute to reduced statistical power and difficulty producing meaningful results (Tarlow & Mahoney, 2000). Few papers report the specific procedures used to recruit older participants and recruitment outcomes (Buckwalter, 2009). Because the population of residents 65 years of age and older is expected to increase dramatically over the next 40 years (U.S. Census Bureau, 2008), there is need for interventions that address health outcomes in this growing population. Given the difficulties recruiting older persons for clinical trials (Cassidy, Baird, & Sheikh, 2001), it is important to understand the challenges and how to overcome them.

The barriers encountered when recruiting older and/or underserved individuals include mistrust and transportation obstacles (Gonzalez, Gardner, & Murasko, 2007; Ory et al., 2002; Shearer, Fleury, & Belyea, 2010), caregiver burden, medical concerns, and indifference (Saunders, Greaney, Lees, & Clark, 2003), sensory and cognitive limitations (Mody et al., 2008), and frail health (Shearer et al., 2010). There is a “point of diminishing returns” in the pursuit of potential study participants at which the likelihood of enrollment becomes very small (Ory et al., 2002, p. 8). The study design may pose additional obstacles to recruitment if there are excessive restrictions on eligibility, which limit the ability to generate an adequate sample size needed for statistical power and representation of the target population (Cassidy et al., 2001; Yancey, Ortega, & Kumanyika, 2006).

In their systematic review of interventions to improve recruitment, UyBico, Pavel, and Gross (2007) report that lack of confidentiality, fear of safety, schedule conflicts, poor access to medical care, lack of knowledge, and language/cultural differences represent additional barriers to minority recruitment. Among African Americans, these barriers have been categorized into three main areas: (a) historical context associated with past unethical research conduct; (b) lack of access and use of health care; and (c) researcher bias, including the researcher’s personal values, sense of what is just, and subsequent decisions about study design and participant inclusion (Dennis & Neese, 2000). Difficulty recruiting may be related to the unique culture of each minority group and the inability or unwillingness of researchers to understand community dynamics and establish positive relationships with community members and leaders (Stahl & Vasquez, 2004). Among some minority populations, especially Latino noncitizens, there is the fear that sharing information may lead to deportation (Levkoff & Sanchez, 2003). Challenged by unanswered or unreturned calls, disconnected telephones, and incorrect telephone numbers, Eakin and colleagues (2007) were unable to complete an initial contact for 37% of potentially eligible Latino participants and cited this as their biggest barrier to recruitment.

The aim of this article is to describe the successful recruitment plan implemented for the Multifaceted Intervention to Improve Medication Adherence (MIIMA) study, which is currently in progress (Insel, Einstein, & Morrow, unpublished). This study reports on the first 276 participants enrolled in a clinical trial designed to test the effectiveness of an in-home, tailored, multifaceted prospective memory intervention on adherence to antihypertensive medication for adults 65 years of age and older. Participants were older community-dwelling adults who were self-managing at least one prescribed antihypertensive medication on a daily basis. Medication adherence was indexed using the Medication Event Monitoring System (MEMS®; AARDEX Corporation, Zug, Switzerland) technology that records bottle openings. Participants were involved in baseline data collection and 4 weeks of baseline adherence monitoring using the MEMS® cap. If adherence to their prescribed medication was 90% or lower, participants were invited to continue with the intervention phase of the study. The intervention phase included weekly visits for 4 weeks from nurses assigned to the control or intervention group, followed by continued medication adherence monitoring for 6 months.

Development of a Recruitment Plan

As the study began, the MIIMA team identified four strategies as important for recruitment and retention. The first focused on access and identifying how to effectively reach the population appropriate for this research. A second area of focus was communication and trust-building, which affected every task from study brochure preparation and initial contact to ongoing interactions with participants throughout study completion. Recognizing that older study participants are a vulnerable population, a third strategy was to provide comfort and security whenever possible in order to reduce participant anxiety about new experiences and settings. Finally, expressing gratitude was established as a way to affirm the importance of each participant’s contribution to the study, and acknowledged that uncertainty and inconvenience may be part of the experience.

Accessing An Appropriate Population

Participants for the MIIMA study needed to be 65 years of age or older, able to speak and read English, and self-managing at least one prescribed antihypertensive medication. Prospective participants were assessed for capacity to communicate (ability to hear and process instructions), and enrolled participants were screened for symptoms of dementia and severe depression. Recruitment approaches reported in the literature include implementing principles of social marketing (Nichols et al., 2004), using a home-hosted “health party” concept adapted from Tupperware and Mary Kay pyramid marketing concepts (Sadler et al., 2006), conferring with community-based advisory boards or organizations (Levkoff & Sanchez, 2003; Mendez-Luck et al., 2011; Mody et al., 2008), hiring research team members with similar personal ethnicities as those being recruited (Areán, Alvidrez, Nery, Estes, & Linkins, 2003; Mendez-Luck et al., 2011), and the facilitation of contact with homebound older adults through trusted mobile meal delivery drivers (Shearer et al., 2010).

Recruitment for the MIIMA study was initiated using social marketing principles (Nichols et al., 2004) and included revisiting previously established community contacts at centers providing meals and activities for older adults of diverse ethnicities. These locations were visited by the team of investigators with prior consent and support of the centers’ directors and staff members, an approach known as sanctioning which promotes collaborative alliances with the community and potential participants (Dennis & Neese, 2000). The importance of creating and nurturing relationships or partnerships with key providers of community services (often viewed as gatekeepers) has been discussed widely (Gonzalez et al., 2007; Mendez-Luck et al., 2011; Mody et al., 2008; Nichols et al., 2004; Saunders et al., 2003). Reciprocity, respect for facility staff’s work, and flexibility are discussed by Mody and colleagues (2008) as being three essential issues to consider in recruitment partnerships with facility health care providers. The MIIMA staff collaborated with senior service providers by assisting with a variety of activities such as chair set-up for outreach events, monitoring and replacing refreshments as needed, and providing directions and assistance for event attendees. They also offered blood pressure checks to interested staff members even though most of these individuals were not eligible for the study.

A significant development in the first year of recruitment was the cultivation of a partnership between MIIMA researchers and a highly visible and active senior services program conducted by one of the local community hospitals. This program maintains a health information center, provides classes and lectures on health-related issues, facilitates periodic health screening events, and offers space for various healthy aging consultants and support groups to meet and interact. This liaison, which originated in a meeting including members of the research team and the senior services program manager and medical director, resulted in a number of cascading recruitment opportunities including participation in numerous health fairs and community education events for older individuals. Mendez-Luck and colleagues (2011, p. S102) also credit the “champions” within community-based organizations (CBOs), whose validation and advocacy for the research were essential for their successful recruitment of women of Mexican descent.

Communication and Trust-Building

Researchers frequently stress the importance of establishing trust during interactions with clinical sites, community-based personnel and potential participants (Curry & Jackson, 2003; Dennis & Neese, 2000; Dilworth-Anderson, 2011; Gonzalez et al., 2007; Ory et al., 2002; Shearer et al., 2010; Yancey et al., 2006). Many authors have reported the trust-building value of relying more on face-to-face contact for introducing the study to targeted vulnerable populations, often citing it as the most effective way to recruit older adults (Areán et al., 2003; Gilliss et al., 2001; Gonzalez et al., 2007; Greaney, Lees, Nigg, Saunders, & Clark, 2006; Resnick et al., 2003). The community events pursued by the MIIMA team were an appropriate setting for effectively introducing information about the study, and researchers interacted with attendees in a face-to-face situation where an attitude of “service before sales” prevailed. Researchers used the offer of blood pressure checks and education about blood pressure readings (service) as an advantageous trust-building launch for introducing more specific information about study participation (sales). The original study flyer was upgraded to a colorful, user-friendly brochure that was designed and printed cost-effectively by the research staff. Written and verbal descriptions of the study purpose and protocol were presented clearly and simply in a manner that was casual and nonthreatening. Interested attendees completed “encounter” forms which requested personal contact information and eligibility-linked data such as age and antihypertensive medication use.

During the follow-up call by the project coordinator, each prospective participant was immediately reminded about the event or circumstances for the original encounter in order to minimize uncertainty about the source and purpose of the call. When possible, this was accompanied by the coordinator’s reference to some memorable aspect of the previous conversation, such as the fact that the individual had just recently moved to town, or had been wearing a pair of unusual, distinctive earrings. The result was often a perceptible increase in relaxation and receptivity to the call once those connections were made. Each individual was also asked how she or he preferred to be addressed (i.e., use of surname, first name, or nickname), and about the best time of day and preferred telephone number for receiving future calls. Study eligibility was evaluated more fully at this time, and any enrollment obstacles that arose were explored as completely and appropriately as possible.

A mailing was sent immediately to each newly recruited person that included a signed welcome letter from the principal investigator and project coordinator. This letter, printed on university letterhead stationery, confirmed the date, time, and place for the baseline data collection appointment and included instructions about bringing blood pressure medications and a list of all medications taken, both prescribed and supplemental, to the appointment. It also included telephone numbers for reaching the project staff. The study consent form was included so that participants could review the content without pressure in the comfort of their homes. For those driving to the university, a highlighted, personalized map of the university area was provided, with special emphasis on how to find the designated parking area. Those participants scheduled for cab transport or for completion of the baseline data collection in their homes received special letters adapted to their particular circumstances. The letters included the assurance that cab costs would be paid by the study, which addressed the occasional concern that participants would be responsible for their own cab fare. The researchers confirmed appointments by telephone a day or two in advance, and clarified instructions regarding meeting location and necessary items to bring for the data collection. This call was also an opportunity for prospective participants to ask additional questions about the study.

For ongoing communication, it can be difficult to make contact with some enrollees, and reasons include work schedules, transient lifestyle, disinterest, misunderstanding regarding study protocol, and medical problems. Some participants had unexpected family emergencies requiring travel out of the area, whereas others moved from place to place for a variety of reasons including multiple residences, extended family commitments, and financial problems. In two situations, the participants’ spouses had cognitive impairment and could not be relied upon to relay phone messages left by research staff. A few participants with voicemail options included with their telephone service seemed unfamiliar or forgetful regarding how to access those messages. Dealing with similar communication obstacles, one research team recruiting Latinos implemented a protocol that required staff to make at least 10 call attempts to reach each participant at each contact point during the study, calling at various times of the day and on different days of the week before abandoning the effort (Eakin et al., 2007). The MIIMA team also made telephone calls at various times, including evenings and weekends, attempting to contact both potential and enrolled participants.

Providing for Comfort and Security

One of the most successful strategies for recruiting older adults is to balance the “benefit:burden ratio” by limiting perceived burdens such as discomfort, fatigue, time, and travel (Mody et al., 2008, p. 2342). The expectation of staff responsibility for supportive, respectful attention to participant comfort and security was established early. This resulted in a set of detailed guidelines for researchers’ interactions with participants throughout all phases of study involvement. Initially, those who described transportation challenges (unfamiliarity with the area, distance of the drive, health limitations, etc.) were assured that cab transport was available. The research staff handled all arrangements for scheduling and payment regarding local cab services. An account was established with a specific transport provider and, with few exceptions, the relationship evolved in a positive and pragmatic manner.

On the day of the scheduled appointment, MIIMA members went to the designated parking area about 15min prior to the anticipated arrival time in order to make immediate contact with the arriving participant. A university campus can be overwhelming for older individuals who may engage in limited driving and may also have some sensory, cognitive, or physical impairment. Arriving participants were initially identified by the type of vehicle driven (determined previously in the recruitment call) and were assisted to find the closest available parking space. The researchers prepaid for parking time by inserting a key device into the parking meter. Those participants who arrived by cab were met in a similar manner in front of the designated university building. MIIMA researchers discovered that cab arrivals required monitoring to ensure that participants were dropped off at the correct location, and that participants were protected from occasional misunderstandings with cab drivers regarding prearranged voucher payments.

Each participant was escorted to the specific data collection location by research staff, was offered a cold bottle of water and a box of tissues, and was informed about the availability of breaks and the location of nearby restrooms. A member of the research staff assisted each participant from the time of university arrival until departure, which included waiting with participants until requested cabs arrived for transport home. This approach minimized participants’ discomfort and allowed researchers to monitor the safety and well-being of those who arrived with poorly controlled hypertension and other health and physical limitations. During this investigation, baseline data collection was suspended for three participants who had extremely high blood pressure readings. In all three cases, after referral and/or transport for medical evaluation in accordance with study protocol, the participants returned to the university when medically stable to complete the data collection and continue with the project.

Over the course of study, all data collections were completed whenever possible by the same researcher who completed the initial baseline measures. The importance of this consistency for enhancing participation and retention was noted in a review of 95 studies focused on minority recruitment between 1999 and 2005 (Yancey et al., 2006) and was also reported by Areán and colleagues (2003). This was also a trust-building strategy in the MIIMA study because post-baseline data collections and nurse visits were completed in participants’ homes, and familiarity with the research team facilitated willingness to admit staff to their personal space. At each visit, participants were informed about what to expect next in the total sequence of events that comprised study participation.

Expressions of Gratitude

Several opportunities were identified for conveying appreciation for the effort and inconvenience that study participation may entail. During the initial follow-up call and prior to actual recruitment, the MIIMA project coordinator thanked prospective participants for their interest and curiosity to learn more about the study. The coordinator also greeted each new participant upon arrival for baseline data collection and expressed gratitude for her or his willingness to share time and information with researchers. Monetary compensation was provided for study participation at two points: initially upon completion of the baseline 4-week medication monitoring period, and later for those who completed the final 6-month data collection. The check requests for these participants were processed immediately, usually the same day that participants completed each phase of the study. Finally, upon completion of the final 6-month data collection, each participant also received a 7-day medication organizer (pill box) imprinted with the name of the study. All study participants received a hand-written note from the project coordinator, who expressed gratitude for their participation. They were also told they would receive a written report about the findings at the end of the study.

Awareness evolved over time regarding specific barriers that might preclude an individual’s enrollment in the MIIMA study, or which might be perceived by the research team as exclusionary factors for qualification. For example, although the original plan specified completion of the baseline data collection at the university, there were community residents who qualified for study eligibility but were unable or reluctant to travel to the university even with the offer of cab transportation. This prompted flexibility regarding the location for baseline data completion, and participants’ homes or other prearranged locations, such as residential clubhouses or community buildings, became acceptable alternate sites. Completing the research measures in distant settings, particularly the cognitive measures, required some care. The potential for disturbance from participants’ family members, pets, telephone calls, and visits is greater in the home or alternate setting, and preparing the participant for the nature of the experience, including a request for no interruptions, was very effective.

Another option that increased participant inclusion was the realization that the Mini-Mental State Examination ([MMSE], Folstein, Folstein, & McHugh, 1975) which is administered face to face during baseline data collection, could be repeated per study protocol during the final 6-month data collection with the telephone-delivered version (Roccaforte, Burke, Bayer, & Wengel, 1992). This procedure enabled the investigators to complete study protocol during a telephone call with participants who planned to leave the area prior to their final data collection. In these situations, the staff arranged the use of prepaid FedEx packages for the return of the medication monitoring bottle and MEMS® cap, and this allowed the researchers to retrieve the final medication adherence information. Three participants completed the study with this long-distance arrangement.

Recruitment and Retention Findingsand Discussion

Recruitment

It is unknown how many total community members were introduced to the study through various media sources (newsletters, newspapers, and television), but interactive recruitment activities for MIIMA resulted in a total of 1,394 direct contacts, of which 49.7% (n = 693) appeared initially to be eligible (Figure 1). From this group, 67.1% (n = 465) were willing to provide personal information for follow-up calls. The research staff was unable to contact 29 (6.24%) of this group, and 58 (12.5%) were found upon further exploration to be ineligible (Table 1). There were 106 (22.8%) refusals (Table 2), 31 of which occurred during the period between recruitment and scheduled enrollment appointments. A total of 276 participants were enrolled and consented, which represents 59.4% of the 465 individuals who had initially provided contact information, and 72.3% of the 382 respondents who completed follow-up calls and met eligibility criteria.

Figure 1.

Recruitment flow chart.

Table 1.

Reasons for Study Ineligibility (Total 58 of 465)

Reason	Number	Percent
Does not reside inarea	13	22.41
Not self-managingown medications	10	17.24
Health (physical/mental) limitations	9	15.52
English language limitations	7	12.07
Absence of hypertension (self-report)	6	10.34
Younger than 65years of age	5	8.62
Not taking BP medication daily	4	6.90
Screening MMSE<24 after consented	4	6.90

Table 2.

Reasons for Study Refusal (Total 106 of 465)

Reason	Number	Percent
Not interested	24	22.64
Too busy	18	16.98
Illness concerns	15	14.15
Unspecified	14	13.21
Family concerns	8	7.54
No perceived personal benefit	7	6.60
Wary about MEMS®cap use	5	4.72
Study did not meet expectations	5	4.72
Personal privacy concerns	3	2.83
Too tired/old	4	3.80
Leaving geographic area soon	1	0.94
Physician advised against study	1	0.94
Study compensation inadequate	1	0.94

A majority of study enrollees (78.62%, n = 217) were contacted face to face during interactions at a community center or event, and most of those encounters included blood pressure checks (Table 3). Response to media-related study coverage varies from study to study, with some reporting that it is not successful and can be quite costly (Areán et al., 2003; Nichols et al., 2004). A feature article about this investigation in the local newspaper resulted in 40 calls to the project coordinator, and 23 of those callers were enrolled. Of this group, 30.43% were non-Caucasian persons including six Hispanic participants. A news release about the study was featured twice in a newsletter sponsored by the county council on aging (a readership of over 20,000 older community residents), but that resulted in only 5.07% (n = 14) of MIIMA enrollees, and a local television news feature produced only two new participants. Although some authors report productive alliances with physician/clinical practice groups (Eakin et al., 2007; Gonzalez et al., 2007), others found that to be a poor source for participant leads (Greaney et al., 2006; Resnick et al., 2003). Only two participants came to MIIMA directly from physician referral in spite of the principal investigator’s presentation to a large group of cardiovascular physicians and subsequent meetings with the clinic manager, when 200 study brochures were distributed in display holders for placement in all patient examination rooms.

Table 3.

Recruitment Sources for Enrolled Participants(276 Total)

Site	Number	Percent
Community hospital seniorservices program	83a	30.07
Subsidized senior centers/meal programs	60b (28a)	21.74
Retiree residential communities	38a	13.77
Health fairs/education events	36b (31a)	13.04
Local newspaper article on study	23	8.42
County council on aging newsletter	14	5.07
Study participant referrals	9	3.30
Churches	3	1.09
University-based listserv news release	2	0.72
Television news feature	2	0.72
Clinical treatment sites	2	0.72
Study staff referral	1a	0.40
Previous study participant	1	0.40
Unspecified word-of-mouth	1	0.40
Study website	1	0.40

^aIncluded face-to-face encounters with blood pressure checks—65.57% of 276 enrolled.

^bIncluded face-to-face encounters—78.62% of 276 enrolled.

Successful recruitment of ethnic minorities is difficult to determine due to the paucity of evidence as reported by clinical researchers (Nápoles & Chadiha, 2011). Also, there are few analytical studies directly testing the effectiveness of approaches used, especially given the sociodemographic variables that could influence recruitment outcomes (Yancey et al., 2006). In the MIIMA study, the minority participation (non-Caucasian participants) was 19.57% (n = 54), and this included 10.9% (n = 30) Hispanic or Latino participants. This is lower than the U.S. Census Bureau (2005–2007) reporting which indicates that 14.6% of all residents ≥65 years of age in this area are Hispanic or Latino. However, among residents in this age group, 19.1% speak a language other than English and 8.3% speak English less than “very well,” which represents an exclusionary factor for participation in this study and was the basis for ineligibility of seven potential participants. A total of 18 (6.52%) of African Americans enrolled, which exceeds the regional population of 1.9% African Americans in this age group. Additional ethnic minorities included two Native Americans, one Pacific Islander, and three who self-identified as “other,” with a remainder of 222 (80.43%) Caucasian participants. There were a total of 207 (75%) women and 69 (25%) men enrolled.

Attentiveness to the broad issue of transportation-related challenges was critical for the successful recruitment of 32.25% of study participants, including those who agreed to cab transport (n = 58) as well as participants who required baseline data completion in the home or nearby locations (n = 31). Demographically, these participants were 89.9% women with a mean age of 80.26 years, and 28.1% of them were ethnically diverse. The reason cited most frequently for study refusal was lack of interest. Some cited employment, volunteer activities, classes, travel, and family caregiving needs as a basis for being “too busy” to participate. Others questioned what benefit, if any, was available to them personally, and weighed the “benefit:burden ratio” (Mody et al., 2008, p. 2342) when deciding whether to enroll. Some expressed concern that they would not be allowed to participate because of other health conditions, and were relieved to learn that eligibility criteria for MIIMA did not generally exclude comorbidities. The few health-related exclusionary criteria, which became apparent during baseline administration of screening measures, included cognitive limitations screened by the MMSE and depression symptoms identified by the Geriatric Depression Scale (Sheikh & Yesavage, 1986), although no participants were screened out due to indicators of severe depression.

Retention

The challenge beyond enrollment was to retain participants for the 7-month duration of the study in approximately equal proportions in the treatment and control conditions to avoid compromising the power and internal validity of the research. This investigation focused on the effectiveness of an intervention supporting prospective remembering; therefore some of the traditional approaches to prevent attrition, such as periodic postcards and birthday cards, could not be used because they may serve as memory cues. Trust-building communication strategies and expressed gratitude were employed to keep participants connected to the study. The overall attrition rate for the MIIMA study was 14.34% (n = 39), below the original projection of 15%, a figure based on the principal investigator’s preliminary work. This rate excluded those participants whose baseline MMSE scores disqualified them, but did include those who later became ineligible due to medication issues such as discontinuation of daily antihypertensive medication or an emerging inability to self-manage medications (Table 4).

Table 4.

Reasons for Participant Attrition (Total 39of 272 Enrolled)

Reason	Number	Percent
Felt might interfere withpersonal med regimen	8	18.60
Chose to withdraw duringbaseline data collection	6	14.00
Concerns about usingMEMS® cap/bottle	5	11.63
No longer taking daily BP med	4	9.30
Receiving assistance with meds—ineligible	3	7.00
Hospitalized for unrelatedmedical problems	2	4.65
Deceased (related tocomorbidities)	2	4.65
Not interested in followingstudy protocol	2	4.65
Objected to intrusion of study organizer box	1	2.33
Family problems	1	2.33
“Too busy” to continue afterinitial 4 weeks	1	2.33
Noncompliance, use of studycap/bottle	1	2.33
Perceived illness episode asrelated to study	1	2.33
Objected to overall datacollection process	1	2.33
Too tired to continue	1	2.33

The completion of study measures can cause significant anxiety for some older and minority participants due to confusion or poor comprehension of scales, fatigue, worry about what the researcher wants, or sensory and cognitive limitations (Jacelon, 2007; Levkoff & Sanchez, 2003; Mody et al., 2008). A review of longitudinal studies in the elderly reports that advancing age, less education, and lower MMSE scores all contribute to increased odds for participants to drop out of studies after enrollment (Chatfield, Brayne, & Matthews, 2005). Of the six MIIMA participants who dropped during the baseline data collection, five withdrew during completion of cognitive measures. All five participants had lower than optimal MMSE scores (ranging from 25 to 29, optimal being 30) which may also support the observation by Mody and colleagues (2008) that cognitive impairment can increase dropout rates when participants become frustrated with testing procedures. MIIMA researchers referred to data collection measures as “activities” rather than “tests” because it is known that the term “testing” can increase anxiety for participants (Austin-Wells, McDougall, & Becker, 2006).

Some participants expressed reservations about using the study-issued medication bottle with MEMS® cap, concerned that its use might interfere with their previously established system for remembering medications (the largest category of study drops), or that it was inconvenient or difficult to use. Greaney and colleagues (2006) note that men were more likely than women to withdraw from their study, and Austin-Wells and colleagues (2008) report that the majority of their dropouts occurred fairly early in the study, within the first 3 months. These findings are consistent with the experiences in the MIIMA study, in which attrition was 12.75% for women and 19.12% for men. Also, 84.62% of all drops occurred within the first 90 days of study participation. The advanced age of this sample (M = 76.32 years) increases the probability that some attrition will be related to morbidity and mortality issues, and during this investigation two participants died and another two were hospitalized due to medical issues unrelated to the study.

Summary

As other researchers have concluded, there is not one single recommended approach for successfully recruiting older participants, particularly when attempting to recruit minorities (Areán et al., 2003; Sood & Stahl 2011; Stahl & Vasquez, 2004). Recruitment strategies that were especially effective for the MIIMA study included use of social marketing principles and forming solid relationships with community-based organizations. The combination of face-to-face contact with the provision of service, such as blood pressure checks, was particularly successful. Older adults and members of minorities are considered vulnerable and additional staff time and support services seem to be necessary to boost the participation of these participants in clinical trials. This can create stress for research staff, but a strong, shared commitment to developing and maintaining respectful, positive relationships with referral sites, potential participants, and enrollees motivated the MIIMA staff to observe established recruitment principles.

The costs for recruiting and retaining individuals from older and underserved groups are high and must be anticipated (Areán et al., 2003; Katula et al., 2007; Ory et al., 2002). This includes costs for services such as transportation but also, sometimes more significantly, the considerable cost of staff time for community-based recruitment activities, telephone follow-up, participant assistance relative to parking and building access, and more intensive oversight in the event of occasional medical instability during research activities. Each research team must explore and implement multiple strategies, testing for effectiveness and revising as necessary. Each project requires thoughtful consideration of the study environment, research tasks, community resources for eligible participants, budgetary provisions, and any special needs or obstacles within the target population.

As the population of older persons continues to increase, it is imperative that research focused on their health-related issues is conducted with participant samples that are large enough to reliably test the research hypotheses and diverse enough to reflect a representative sample and thus ensure generalizability of the findings. Detailed reporting of successful recruitment methods allows researchers access to a broad array of options from which they may tailor their own best approaches for optimal recruitment and retention of eligible study participants.