Recruitment of Community-Based Samples: Experiences and Recommendations for Optimizing Success

Abstract

Background

Recruitment in health and social science research is a critically important but often overlooked step in conducting successful research. The challenges associated with recruitment pertain to multiple factors such as enrolling groups with vulnerabilities, obtaining geographic, cultural, and ethnic representation within study samples, supporting the participation of less accessible populations such as older adults, and developing networks to support recruitment.

Purpose

This paper presents the experiences of two early career researchers in recruiting community-based samples of older adults, their caregivers, and associated health providers.

Methods

Challenges and facilitators in recruiting two community-based qualitative research samples are identified and discussed in relation to the literature.

Results

Challenges included: identifying potential participants, engaging referral partners, implementing multi-methods, and achieving study sample diversity. Facilitators included: making connections in the community, building relationships, and drawing on existing networks.

Conclusions

Findings suggest the need for greater recognition of the importance of having clear frameworks and strategies to address recruitment prior to study commencement as well as the need to have clear outreach strategies to optimize inclusion of marginalized groups. Recommendations and a guide are provided to inform the development of recruitment approaches of early career researchers in health and social science research.

Background

Challenges pertaining to participant recruitment are often underappreciated and limit the successful conduct of community-based research within a variety of contexts. Description of recruitment methods tends to be underreported in the literature thus highlighting the gap in the knowledge base regarding effective strategies to overcome recruitment challenges (Foster et al., 2011; Rockliffe et al., 2018). Given the complexity of the recruitment challenges described in the literature, there would be benefits from generating recommendations to consider when developing and implementing research with community-based populations. The aim of this paper is to describe the experiences of two early career researchers in recruiting community-based samples, to position these findings within the literature, and to provide recommendations to inform the recruitment of community-based samples.

Participant recruitment is a critical step in the conduct of health and social science research. In quantitative research, participant recruitment must meet sample size requirements that are large enough to be representative of the population of interest and have sufficient statistical power (Gul & Ali, 2009). In qualitative research, sufficient recruitment is required to ensure data adequacy, namely a sample that reflects suitable size and composition to address the research question (Vasileiou et al., 2018). The process of participant recruitment is often challenging, complex, and impacted by multiple factors including the availability of resources to fund targeted recruitment, the type of research being conducted, the potential participant demands associated with the research, and individual characteristics of recruiters and participants (McDonald et al., 2006; Newington & Metcalfe, 2014; Treweek et al., 2010). For example, recruitment of groups who speak different languages or those who may not be actively engaged with the health care system may be more difficult to reach using standardized approaches such as databases and large-scale mail outs (Newington & Metcalfe, 2014). The geographic location of potential participants can also impact recruitment in a variety of ways. For example, participant populations in rural settings may not be reached by common recruitment strategies such as flyers or in-person engagement via community hubs due to their remote locales (Ellard-Gray et al., 2015). Furthermore, findings by a Canadian team of researchers suggest that it is important to “know” the desired sample within the greater community and then employ targeted recruitment strategies that are directed to the geographic locations or facilities frequented by the sample (Williams et al., 2017).

Certain populations may be challenging to recruit including individuals from ethnic or racial minority groups, persons with disabilities, children, older adults, or persons of a two-spirit, lesbian, gay, bisexual, transgender, queer, inquiring (2SLGBTQI+) orientation (Flanagan & Hancock, 2010; Rockliffe et al., 2018). Factors impacting their recruitment include language barriers, ability to engage community leaders, mistrust rooted in historical oppression, fear, and abuse or prior negative experiences in accessing services (Flanagan & Hancock, 2010; Rockliffe et al., 2018). As such, these populations are underrepresented in research leading to gaps in knowledge and potentially poorer health outcomes (Rogers & Meek Lange, 2013; Winter et al., 2018). The effects of this underrepresentation can be multi-faceted, with some segments of the population never being represented in research and others who are more accessible being oversampled (Kroll et al., 2007; Szabo et al., 2018).

Increasing representation of underrepresented groups is a common challenge experienced in recruitment but in some contexts, segments of the population may be oversampled, such as in major urban locations with teaching hospitals and abundant research programs (Gilbert et al., 2012). Approaches to overcome these potential biases in recruitment and sampling include employing flexibility and willingness to meet participants within their own communities as pivotal for success (Foster et al., 2011).

Community-based older adults represent a segment of the population that frequently experience chronic health conditions potentially resulting in physical limitations and a reduction in their social participation in research (Griffith et al., 2017). While some older adults with chronic conditions may be high users of health care, limitations in mobility, cognition, and sensory function place them at risk of becoming homebound and therefore less likely to participate in research (Crawford Shearer et al., 2010; Latham & Ackroyd-Stolarz, 2014). For example, visual disabilities, while not uncommon in older adults, may render traditional recruitment approaches using advertisements and letters of information which tend to be “text heavy” as unsuitable for reaching this population (Banas et al., 2019). Older adults who are socially isolated or homebound may also be unreachable by other commonly used recruitment techniques such as snowball sampling or community engagement activities. Furthermore, older adults who are homebound may be reluctant to invite strangers into their home and therefore decline to participate in research when they are contacted (Crawford Shearer et al., 2010).

Other factors that can adversely impact the recruitment of older adults include the presence of chronic conditions and their perceived meaning for older adults’ health. For example, common conditions that affect older adults, such as dementia, incontinence, falls, and depression, can be socially stigmatizing and thus adversely affect older adults’ willingness to disclose the conditions and discuss their experiences openly in a research context (Flanagan & Hancock, 2010; Richeson & Shelton, 2006; Wyman et al., 2018). Moreover, older adults with chronic conditions may deem their declining health an inevitable and a “normal” part of aging and therefore be unwilling to engage in health care services, such as primary care providers, which serve as potential avenues for research recruitment (Walters et al., 2001).

In addition to experiencing potential health challenges, older adults may be caregivers themselves, a role that can further impact their recruitment into research studies. Caregiving often involves an ongoing and constant commitment by caregivers, leaving them at heightened risk of emotional and financial stress, as well as declining health (Camak, 2015; Garlo et al., 2010). Researchers must be sensitive to the demands on their time and energy yet also provide opportunities for their voices to be heard. This may entail recruiting caregivers during their daily activities such as meeting them at appointments and facilitating their research participation through convenient in-home interviews to reduce the likelihood of additional burden being placed on the caregivers. Furthermore, researchers must recognize that health providers, who can often facilitate participant recruitment, may deem caregivers as unsuitable for research due to fears of adverse events that could result from their participation (Townsley et al., 2005). This may necessitate that additional time and relationship building must precede the recruitment process to optimize the opportunity for study success.

Recruitment challenges are well-described in the literature, but thus far there are limited resources to inform the development and implementation of research with community-based populations. The purpose of this study is to describe the experiences of two early career researchers in recruiting samples for their doctoral theses of community-based older adults, their caregivers, and associated health providers in southern Ontario, Canada; to situate our experiences within the context of the literature; and to outline a set of recommendations to guide the recruitment of participants within a variety of community-based settings and contexts.

Recruitment experiences: Two community-based studies

Study one synopsis

Evidence has shown that family and friend caregivers of stroke survivors are significantly and negatively impacted by caregiving. The negative effects of caregiving may persist over time suggesting the need for ongoing care and support. However, little is known about caregivers’ use of formally funded health and social services or the factors influencing their access to and use of these services (Garnett et al., 2021). Study one used a qualitative research design with an interpretive description approach to explore caregivers’ experiences accessing and using formal health and social services in southern Ontario. This was done to increase understanding of the context of stroke caregiving, service access, and factors impacting access to and use of formal health and social services. To capture the experience of caregiving over time, the inclusion criteria stipulated those caregivers who were providing support to a stroke survivor who experienced a stroke at least 6 months and up to 5 years ago were invited to participate. Recruitment took place over 6 months and a total of 22 stroke caregivers at an average of 30 months post-stroke and 18 health providers were recruited.

Study two synopsis

Urinary incontinence is a common complication of type 2 diabetes mellitus for older adults yet is often unrecognized and thereby untreated. In older adults, urinary incontinence is associated with reduced quality of life and risk of institutionalization. Study two used a convergent, mixed methods research design to better understand the complexity of living with type 2 diabetes and urinary incontinence in older adults who receive home-care services in southern Ontario, Canada to inform the development of home-care supportive interventions (Northwood et al., 2019). In the qualitative strand, an interpretive descriptive approach was used, and one-on-one interviews were conducted with older home-care clients to understand their perspectives on managing incontinence and diabetes. As well, home-care nurses were interviewed to provide the “thoughtful practitioner” perspective and contribute their expert experiences formulated from caring for many older adults with diabetes and incontinence over time (Thorne, 2016, p. 92). Eighteen older adults and 15 home-care nurses were recruited over a 6-month period (Northwood et al., 2021a, 2021b).

Studies’ commonalities

Studies one and two had commonalities that facilitated our shared reflection on recruitment approaches. The first commonality was that older adult caregivers and older home-care clients are both community-dwelling populations that are not immediately or easily accessible as a group, for example, older hospital patients. A second commonality was that both studies employed an interpretive description design, and as such, recruited health care providers to triangulate what we learned from speaking directly to older adults and caregivers about their experiences with observations from health care providers in their care and support of these populations. Lastly, in both studies, we employed a variety of recruitment strategies and had similar rates of effectiveness. Refer to Table 1 for a summary of recruitment strategies by a number of participants across both studies.

Table 1.

Recruitment strategy by number of participants in studies 1 and 2.

Recruitment strategy	Number of older adult and caregiver participants	Number of health care professional participants
Professional network	8	18
Adult day program	13	3
Community organizations	4	9
Congregate exercise class	8	0
Planned referral partners	6	n/a
Advertisement	1	0
Other	0	3
Total	40	33

Recruitment challenges of studies one and two

Identifying potential participants

An initial challenge experienced in both studies was engaging participants in research while lacking a centralized location or system to identify these potential participants. Research conducted in controlled settings such as hospitals and publicly funded medical clinics often lends itself to participant identification using databases and clear alliances between universities and teaching hospitals. In contrast, community-based research often necessitates strategies to first identify the population of interest in the community before recruitment can commence. For example, caregivers may not be receiving services themselves and must be located using multi-pronged strategies such as community outreach, social media, and patient health providers acting as arm's length access points. Or older adults may be receiving home services from more than one organization which can complicate the process of accessing client information for research purposes.

To maximize recruitment success in study two, 10 study information sessions were held across a variety of settings, such as congregate exercise programs, retirement homes, diabetes education programs, and adult day programs to reach potential participants. In addition, informational postcards and posters were provided to assisted living buildings, municipal housing, home-care nurses, and diabetes educators. Of note, the caregivers and older adults receiving home-care services in study two were often homebound due to their health and caregiving roles, which meant that partnering with community health clinics was ineffective for identifying these potential study participants.

Engaging referral partners

Engaging community organizations as partners in recruitment strategies in both studies proved complicated and difficult to implement despite the agencies’ expressed enthusiasm to support the research. Communicating research information to potential participants via a variety of in-home clinicians ultimately proved untenable due to the clinicians’ already burdened caseload. Other logistical challenges hampered recruitment such as a restructure of the home-care sector resulted in a concomitant change in home-care coordinators’ roles, responsibilities, and reporting. This meant that these key potential facilitators of the recruitment process were now inaccessible.

Additional hurdles presented when lower than expected recruitment numbers necessitated expanding the geographic boundaries of one of the studies. This led to the requirement of further ethics approval from a new partnering hospital and added senior-level authorization from community-level partnerships. Ultimately these unanticipated hurdles were overcome, and successful recruitment was achieved but considerably more time and resources were required. Furthermore, as early career researchers, limited funding constrained the use of dedicated recruiters.

Implementing multi-methods

Using a multiple-pronged approach to recruitment was ultimately successful for our respective studies, however, our experiences highlighted the challenges potentially facing student and early career researchers. For example, as an independent researcher, publishing flyers, completing mail outs, and offering incentives were costly and expenses frequently ran higher than predicted to reach sample size goals. In the case of study one, some participants felt that an honorarium was not appropriate; they wanted their voices to be heard and felt compensation was counter to their motive to participate. Sampling that initially focused on one region was ultimately expanded to encompass broader geographic catchments due to low rates of recruitment. Changing the recruitment strategy necessitated greater time to build new relationships with key stakeholders, present the study information at multiple venues, and significantly increased study costs due to impact on time and travel. While each of these factors alone may not significantly hamper recruitment, experienced together, they can substantially delay the research process and must be anticipated to set realistic guidelines to meet study and granting agency expectations.

Study sample diversity

Attaining a racially, ethnically, and gender heterogeneous study sample was challenging in both studies. We acknowledge the contributing shortcomings of our respective study methodologies including inclusion criteria that stipulated only English speakers were eligible for participation and moderately sized geographic recruitment regions. Study one's sample included older adult participants from a range of socioeconomic backgrounds; however, there was limited racial variability across the sample. Study two included the use of geographic stratification in the recruitment strategy thereby obtaining a sample with good socioeconomic variability. However, in the case of both studies, recruitment strategies (e.g., stratifying sampling by racial, ethnic identity, gender, or using interpreters) that would have potentially improved diverse representation were not used. These strategies would also have required significant time and resources that were beyond the scope of the studies. We acknowledge that these limiting factors may have potentially excluded participants of more diverse racial, gender, and ethnic identities.

Both studies utilized focused strategies such as snowball and purposeful sampling in the later stages of participant recruitment to maximize the likelihood of increasing other gender representations in the studies. However, neither study sample had participants who identified as non-binary. We acknowledge that the attainment of a diverse sample should have received greater attention during the initial development of our recruitment strategies. In addressing sample diversity, researchers could proactively consult with community organizations that represent a range of underrepresented groups (e.g., Egale or culturally specific community organizations) to obtain advice, build relationships, and foster trust with these underrepresented groups.

Recruitment facilitators

Making connections in the community

A key facilitator of successful recruitment within the context of these two health research studies was the assistance provided by a variety of community service organizations. These organizations were typically outside of the scope of mainstream health services, such as medical clinics and community hospitals, and included adult day programs, community organizations such as the YMCA, and caregiver support groups and associations. As researchers, it was important to understand that although the research was focused on understanding health service delivery and use, not all care needs of this population were met by health care services alone. By building diverse relationships with these other community partners, we founded strong alliances with organizations that also supported older adults and their caregivers. Furthermore, they had the interest and capacity to support recruitment when other organizations could not.

Building relationships

Building professional relationships with community partners was integral to establishing strong community connections. As early career researchers, we recognized that we did not yet have established and trusted relationships with community partners. Thus, senior researchers were key in this process of relationship development drawing upon the depth and breadth of their professional networks to facilitate introductions and help build and develop the working relationships we sought to develop as junior researchers. Often, they had pre-existing relationships with key stakeholders such as community diabetes programs or provincial stroke networks. Importantly, establishing relationships within the community should also focus on increasing opportunities to connect with marginalized populations such as those from racial or ethnic minorities and those who identify as 2SLGBTQI+.

These important introductions were just the initial step and we had to invest significant time being visible and engaged in the activities of community partners and this also enabled us as junior researchers to share our research expertise, our concern and role as advocates for change. These actions helped foster a trusting, reciprocal relationship with the community partners and potential participants. We determined that our visibility within the community was key to this process, ultimately being the most successful recruitment strategy, we used, in contrast to flyers or posters that resulted in little to no study participants.

Drawing on existing networks

An invaluable and initially underappreciated facilitator of recruitment turned out to be the networks we had developed during our previous and current clinical practice positions. As experienced gerontological nurses, we had many colleagues that supported the aims of our research and likewise were committed to improving integrated care for older adults and their caregivers. These professional networks included specialized geriatric and neurology services, nurse continence advisors, nursing association networks, and community support services. The generosity of these professionals in connecting us to their partner professions in other geographic locales effectively doubled and sometimes even tripled the number of individuals who were able to support our study recruitment. These experiences highlight the importance of exploring connections even if at first they seem distal to the study goal and purpose. For more junior researchers, this could mean connecting with student health associations or through professional organizations.

Discussion

The purpose of this article was threefold: (1) to describe the experiences of two early career researchers in recruiting community-based samples of older adults and associated health providers in southern Ontario, Canada; (2) to situate the findings within the literature; and (3) to develop guidelines to facilitate recruitment of community-based samples. Study one used a qualitative approach to explore caregivers’ experiences accessing and using formal health and social services. Study two was a mixed-methods study to explore the complexity of living with type 2 diabetes and urinary incontinence in older adults who receive home-care services. Our recruitment experiences highlight some of the barriers we encountered but they also highlight the facilitators that ultimately led to the successful recruitment of our study samples.

One of the main challenges experienced in both studies was the difficulty in identifying potential study participants without the benefit of clinical databases commonly maintained in settings such as teaching hospitals. Overcoming this barrier necessitated using multi-pronged approaches such as combining community outreach with posters and health providers acting as direct access points to potential participants. Ironically the use of multi-pronged approaches presented new barriers, those of additional time and cost, findings echoed by Williams et al. (2017) who successfully implemented the door-to-door distribution of flyers, newspapers, and large-scale organizational engagement to meet recruitment goals but also with substantial additional cost and time requirements. However, our initial challenge in identifying where best to engage in recruitment also spoke to our need to familiarize ourselves with the dynamics of our populations and the factors that may impede their willingness to participate in research (Stahl & Vasquez, 2004).

In situations when research involves populations who are hard to reach such as older adults, successful participant recruitment must also take care to abide by ethical principles such as being transparent, respectful, and using non-coercive recruitment strategies (Gledhill et al., 2008). Research suggests that including older adults in the earliest stages of study development can help overcome issues such as participant identification and participation (Newington & Metcalfe, 2014). Our readiness to build trust and learn about our populations of interest in addition to our willingness to accommodate the timing and location of study interviews was an integral aspect of achieving recruitment success (Crawford Shearer et al., 2010).

There is widespread acknowledgment that successful recruitment in clinical research does not occur in isolation (Newington & Metcalfe, 2014). As junior researchers, the need to involve partners in the recruitment process was paramount; in part due to constraints on our funding to support recruitment but also because we were in the preliminary stages of establishing research networks. Research suggests that active approaches (e.g., person-to-person engagement) to recruitment are more successful than passive approaches (e.g., poster or social media post), a finding our research corroborated (Gul & Ali, 2009). As such, clinicians across a variety of health professions served as important entry points for connecting with potential study participants in both studies. This was especially important in study one where caregivers may not have been accessing services, but the person they were supporting was.

The frontline clinicians also served as connection points between the researchers and other recruitment locales such as day programs and specialized community services. Clinician workload is acknowledged in the literature as impacting their participation in the recruitment process and was documented by our experience, particularly in the context of study two (Sullivan-Bolyai et al., 2007). However, many clinicians were enthusiastic about the research studies and sought to facilitate the recruitment process as they were able, a finding also corroborated by the literature (Graffy et al., 2009). Importantly, findings from our research studies emphasize the value of engaging referral partners early in the recruitment process but also highlight the need to have multiple partners engaged in this process. Lack of partner engagement was frequently due to systemic factors and researchers should explore these potential confounders to recruitment as they develop their study methodology.

The use of multiple approaches to achieve our recruitment goals was both necessary and challenging in the context of our respective studies. Recruiting community-based populations requires a carefully considered approach that takes into account the behaviors and sociocultural norms of the population and is known to be taxing (Williams et al., 2017). In our studies, as cited by Rockliffe et al. (2018) and Ellard-Gray et al. (2015), community and individual engagement was more successful than using posters, advertisements, and remote engagement strategies. This may have been due to participants’ constraints on mobility but also likely influenced by the older adults’ vulnerability and their improved ability to develop trust when in situations of direct face-to-face contact (McHenry et al., 2015).

At different stages of recruitment, we tailored our approach to meet recruitment goals by increasing the geographical reach and increasing attendance at community events, but this impacted expenses through increased time and travel costs. All participants received honorariums declared at the time of recruitment, but some participants felt this was unnecessary and undermined the pure intent of their participation. Using study incentives continues to be controversial but is increasingly common, calling attention to the need for clear guidelines and recommendations for researchers (Head, 2009). These experiences also highlight the lack of available frameworks or conceptual maps that could be used to guide the recruitment process and help researchers plan cost-effective recruitment strategies (Foster et al., 2011; Gilmore-Bykovskyi et al., 2019). Moreover, greater emphasis on publishing the details of recruiting strategies in the literature would aid study transparency and be a great asset for junior and early career researchers (Buckwalter, 2009; Rockliffe et al., 2018).

Recruiting a diverse sample of participants who met the clinically relevant inclusion criteria presented a challenge for both studies. Although qualitative data collection methods offer multiple approaches for participation, they can still be exclusionary, for example, through the use of unilingual language (Banas et al., 2019). The difficulties associated with obtaining a sample of participants that includes both racial and ethnic diversity are well-documented in the literature (Gledhill et al., 2008). Issues of trust, language, and cultural barriers contribute to their underrepresentation in research, and present substantial challenges for recruitment (Corbie-Smith et al., 2002; Levkoff & Sanchez, 2003). In the case of a trust, researchers may face recruitment challenges because the population of interest may have historically had experiences that have resulted in mistrust in relationships, for example, migrant workers who have precarious status and may fear deportation (Fete et al., 2019). Or older adults and ethnic minorities who may be wary of research and institutions based on historical events such as abuse or provision of misinformation and therefore avoid consenting for study participation (Corbie-Smith et al., 2002; Moreno-John et al., 2004).

Roots of mistrust can also be grounded in perceived power differences and vulnerabilities between study participants and researchers (Banas et al., 2019). Members of 2SLGBTQI+ communities have often experienced extensive marginalization and discrimination over their life course both within and outside of health systems (Boule et al., 2020; Fredriksen Goldsen et al., 2019). Therefore, as older adults, their perceptions and trust of the health professions may be impacted by their fears of increased vulnerability as a member of a non-gender-conforming group (Pang et al., 2019). Increasing efforts must be made to increase the confidence and trust within these communities using strategies such as engaging them at the stages of research planning and development, engaging in 2SLGBTQI+ outreach, and staff education on cultural competency (Wood et al., 2016).

The sample obtained in study one represented a range of socioeconomic backgrounds but lacked both racial and ethnic diversity. To overcome this obstacle, a geographically stratified recruitment approach provides one of the best opportunities for increasing the sample diversity (Fete et al., 2019). This strategy was used in study two whereby recruitment was stratified by geography resulting in a sample, which was socioeconomically heterogeneous. However, there was not a concomitant increase in the racial or ethnic representation of the obtained study sample. Strategies such as facilitating data collection in other languages or using strategies such as photovoice for data collection could help increase participation from a variety of racial and ethnic groups (Bonevski et al., 2014).

Historically, older adults are underrepresented in the research literature (Herrera et al., 2010; Kuchel, 2019). These two studies help fill this well-known gap in the knowledge base, but our recruitment experiences also highlight the importance of addressing racial and ethnic identity in research with older adults. Moving forward, future recruitment strategies should ideally address the parameters of ethnicity, race, and socioeconomic a priori to realize recruitment goals and attain study samples that are represented across a range of factors. Failure to do so may impact the ability of a study to provide results that are relevant and adequately representative of the populations that were sampled (Tarlow & Mahoney, 2000). However, overly specific eligibility criteria can also limit the ability to reach a suitable sample size so the cost and benefits should be carefully assessed prior to study commencement (Yancey et al., 2006).

The importance of making connections and building relationships in the community was a facilitator to recruitment but challenging to achieve. The process highlighted the need to understand the context including the behavior and relational patterns of the target study population to plan recruitment (Gilmore-Bykovskyi et al., 2019). Importantly, building relationships can take time and may depend upon the extent of pre-existing networks, which could be a limiting factor for early career researchers. Time and cost must be considered in the development of recruitment strategies to engage community-based samples of participants. For example, a systematic review of walking intervention studies conducted primarily in the United States (US), Australia, and the United Kingdom noted that the overall study schedule did not allow sufficient time to recruit participants, monitor the effectiveness of recruitment strategies, and alter plans if required (Foster et al., 2011). Time is also needed to determine the most successful recruitment strategy for the population of interest. For example, a US intervention study for family caregivers of older adults with early stage dementia used multiple recruitment methods, including radio advertisements, phone contacts, and community seminars, yet small numbers of participants were successfully recruited from each strategy (Szabo et al., 2018).

As demonstrated in the literature, there is potential for differential recruitment success depending on the population. For example, in study one, the investigator had great success in recruitment from meeting with community groups, but other researchers have described this method as the least successful (Gledhill et al., 2008; Rockliffe et al., 2018; UyBico et al., 2007). This could be in part related to the delivery of the meetings as Gledhill et al. (2008) presented to groups as large as 90 and on average 40 or the lack of trust toward researchers among community leaders cited by Rockliffe et al. (2018). In both studies, we met with much smaller groups, typically no more than 10 attendees. These smaller, more informal sessions likely ensured that potential participants felt a sense of connection and trust toward the researchers, a known contributor to recruitment success of vulnerable populations (Yancey et al., 2006). This type of face-to-face approach was found to be very successful in recruiting racial and ethnic minority populations in the systematic review of recruitment strategies by Ibrahim and Sidani (2014) and is widely recognized as the best approach for recruiting older adults (Gonzalez et al., 2007; Saunders et al., 2003).

Both studies achieved the greatest recruitment success through community engagement initiatives, but we acknowledge that the samples attained may reflect biases as those who were more involved in community programs may have been more likely to be recruited to the studies (Meyer & Wilson, 2009). Obtaining community-based samples are inherently challenging and it is important to maximize the success of recruitment strategies that work well. However, researchers must also strive to provide clear and transparent accounts of their recruitment processes as an integral part of study reporting.

Professional networks previously established through clinical and academic activities were key facilitators to successful recruitment in both studies. We approached these clinical service agencies, individuals, or stakeholders prior to study recruitment, and they facilitated access to participants or informed us of potential recruitment contacts. Effectively, our professional standing among our peers included a strong element of trust which they extended to us in providing support for our research activities. The trust potential participants placed in these contacts were extended to us as researchers culminating in a successful recruitment strategy (Crawford Shearer et al., 2010; Gilmore-Bykovskyi et al., 2019; Szabo et al., 2018). Our experience highlights the potential value of involving would-be participants and community partners in the earliest stages of research development akin to participatory research (Bonevski et al., 2014; Schmittdiel et al., 2010). Furthermore, our experience also highlighted how valuable our professional networks were even if they initially appeared tangential from the primary goals of our studies. Our recruitment experience demonstrated that even junior researchers with limited budgets and resources can still achieve substantial recruitment reach if they are invested in exploring strategies for maximizing recruitment success.

Using our experiences in recruiting community-based samples, in conjunction with contributions from the literature, we developed a guide to inform and optimize opportunities for success when recruiting community-based samples. This guide can be used during study design and study implementation phases to assist researchers in considering factors that will impact recruitment success such as costs (honorariums, travel, payment for professional recruiters), time, and the importance of relationships and networks. In addition, we include a list of strategies that can assist researchers in successfully meeting their recruitment targets (e.g., clarifying rules around honorariums at the outset, tracking and regularly evaluating recruitment strategies, and engaging representative members of target populations in recruitment processes) (Table 2).

Table 2.

Guide for optimizing recruitment success of community-based samples.

Influencing factors	Potential strategies	Optimizing success
Identifying potential participants	□ Identify geographic locales for potential participants meeting inclusion criteria (including cultural/ethnic/gender representation) □ Identify access points for health- and social-care services—consult with community partners for suggestions/advice □ Review potential barriers (cost, travel, time) to participants study participation □ Plan to mitigate known barriers in recruitment strategy □ Understand potential participants’ expectations regarding honorariums □ Budget for participant honorariums/incentives □ Engage patients and the public at all stages of research design and implementation	□ Consult with community partners during the study design phase □ Enquire about existing community engagement networks such as patient and family councils □ Clarify rules and regulations around the use of honorariums
Engaging referral partners	□ Identify community or volunteer organizations that may facilitate access to a population of interest □ Identify KEP within the community □ Plan time for KEP engagement prior to study commencement □ Identify health providers engaged with a population of interest □ Test networks to determine feasibility and fidelity of recruitment strategy	□ Consider applying for funding to support recruitment □ Engage participants in knowledge mobilization/translation □ Foster relationships with community partners beyond focal research projects
Implementing multi-methods	□ Include multi-methods recruitment in proposal development □ Determine feasibility of implementing recruitment strategy in different geographies (time, cost, resources) □ Conduct a rapid pilot phase of testing and monitoring recruitment strategy □ Consider match of a recruitment strategy to potential participants (e.g., social media use, radio, print advertisement, recruitment partner, snowballing)	□ Track and evaluate recruitment strategies early and frequently □ Consider whether recruitment strategy scalable to other jurisdictions/regions □ Consider potential arms-length recruitment (e.g., friends/family)
Study sample diversity	□ Ensure the study sample is reflective of a population that the research aims to inform □ Plan explicitly for study sample diversity (e.g., racial and ethnic identity, sexual orientation, socioeconomic status) □ Consult resources such as Egale Canada (https://egale.ca/about/) to inform recruitment approaches that are inclusive of members of 2SLGBTQI+ communities	□ Foster equity, diversity, and inclusion at all stages of research by maintaining awareness of issues such as unconscious bias □ Engage representative members of the target sample in the recruitment process
Making connections in the community	□ Visibility in the community is important to engage KEP and potential participants □ Attend events/activities with opportunities for potential participant engagement □ Actively engage in knowledge exchange throughout the study (e.g., deliver caregiver wellness presentation)	□ Liaise with professional organizations □ Liaise with patient and caregiver support organizations □ Liaise with community organizations (e.g., non-profit organizations, neighborhood organizations, Egale Canada) to help build relationships between researchers and community members
	□ Actively seek out KEP in the community □ Remain actively engaged with KEP throughout the study and provide explicit deliverables upon study completion
Drawing on existing networks	□ Explore within professional and academic networks for recruitment support/contacts □ Engage with senior researchers’ networks to build and foster KEP and recruitment	□ Make use of social media where appropriate □ Consider using community databases

2SLGBTQI+: two-spirit, lesbian, gay, bisexual, transgender, queer, inquiring; KEP: knowledge exchange partners.

Limitations

While this commentary provides perspective and recommendations on the recruitment of community-based samples, it is not without limitations. While the authors drew upon a strong body of recruitment research to inform the manuscript, the reflective component included only two experiences. Some of the identified strategies and recommendations to enhance recruitment may be context-specific and may require context-specific adaptation. Future research on the methodological aspects of recruitment is needed to better inform guidelines that can support the engagement of study samples that respect principles of equity, diversity, and inclusion.

Conclusion

Our experience as early career researchers seeking to recruit representative samples of community-based older adults, caregivers, and associated health providers proved to be challenging. However, this experience provided a great opportunity for us to identify important steps and strategies to consider when planning and implementing the recruitment of community-based populations for research studies. Through our work, we have identified some key challenges and facilitators to recruitment. Challenges included: identifying potential participants, engaging referral partners, implementing multi-methods, and achieving study sample diversity. Facilitators included: making connections in the community, building relationships, and drawing on existing networks. Our findings, in conjunction with the literature, informed the development of a guide for optimizing recruitment success that may be useful for others conducting community-based research.

Author Biographies

Anna Garnett, RN, PhD is an Assistant Professor in the Arthur Labatt Family School of Nursing at Western University in London, Ontario.

Melissa Northwood is an Assistant Professor in the School of Nursing at McMaster University in Hamilton, Ontario.