Abstract
Improving caregivers’ quality of life begins with conducting research that aims to understand caregiver’s needs. However, caregivers may be reluctant to participate in research studies, adding to the importance of developing, tracking, and evaluating recruitment strategies. Here, we review the nationwide community outreach practices we implemented as our lab embarked on a program of research to develop an online intervention tailored for those caring for a spouse with Alzheimer’s Disease or a Related Dementia (ADRD). We made hundreds of “cold” calls and e-mails to organizations across the US who may have access to caregivers, but whom we had no prior relationship with. It took 36.5 contact attempts to possible recruitment sites for every 1 enrolled caregiver. While time consuming, this low-cost recruitment method may offer a route for research teams to recruit caregivers into research studies without access to a medical center or collaborating physician(s).
• Reports recruitment data using a grassroots approach to recruiting ADRD spousal caregivers for an online intervention development study. • These data demonstrate that ADRD caregiving research may be possible for research teams without connections to a medical center or collaborating physician(s).
• Other research teams can apply these methods to their own studies to improve their access to caregivers. • Utilizing these methods may broaden researchers’ ability to recruit caregivers living in more rural settings or those who have less access to care.What this paper adds
Applications of study findings
By 2060, an estimated 13.8 million Americans will be living with Alzheimer’s or a related dementia (ADRD; Alzheimer’s Association, 2022), with the majority cared for by their spouses (i.e., about 60%; Harris et al., 2021). Caregiving for a spouse living with ADRD is extremely stressful, often prolonged (e.g., 7–10 years; Liang et al., 2021; Todd et al., 2013), and the stress associated with caregiving may put people at risk for health issues themselves (Wu-Chung et al., 2022; Schulz & Beach, 1999). Thus, there is a need for interventions that may also help “care for the caregiver” (Harris et al., 2021). Yet, recruiting spousal caregivers into research studies is difficult (e.g., Baker et al., 2023).
The primary means of recruiting spousal caregivers into ADRD-related research is through formal healthcare centers (e.g., Reed et al., 2014; Tarlow & Mahoney, 2000). This, however, requires that research teams are embedded into or collaborate closely with a medical facility, and may limit researcher’s ability to recruit caregivers living in more rural settings or those who have less access to care.
In this study, we discuss our methods to and outcomes of recruit(ing) spousal caregivers, nationwide, into an online intervention development study, conducted in the west south central region of the U.S., without direct access to a medical center or collaborating physician(s). We also suggest future directions for making research with ADRD spousal caregivers more attainable to research teams just beginning to conduct research with this population.
Methods
IRB-approved recruitment activities occurred between June 2023 and April 2024, with most active recruitment efforts (i.e., recruitment calls) happening during the latter five months. The larger project aimed to assess feasibility and acceptability of an online expressive writing intervention for spousal caregivers via participation in at least one online 90-min focus group. Independent from the recruitment efforts reported in this brief report, we enrolled thirteen spousal caregivers in our study, who were referred from a collaborator’s National Institutes on Aging (NIA) R01-funded study conducted in Houston, TX (Wu-Chung et al., 2024). In contrast, the sample reported here was an additional sample recruited independently by our research team without prior relationships with the participants or community partners. These additional recruitment strategies were utilized to diversify our overall study sample to include caregivers who had no former relationships with our research team (or the research team(s) of our collaborator(s)) in an effort to reduce potential bias in participant’s feedback on our intervention materials during the focus groups. To maintain consistency across the study sub-samples, we adopted the same eligibility criteria as the R01 study the n = 13 sub-sample of spousal caregivers had formerly participated in. Below, we outline our methods used to independently recruit the additional subsample of n = 14 caregivers, organized within three broader phases: (1) Recruitment Site Database Construction, during which our team built a database of potential recruitment sites / avenues of recruiting caregivers; (2) Community Outreach, which entailed contacting these community sites and communicating a recruitment call-to-action; (3) Participant Outreach, during which we made direct contact with potential participants. To manage phases 2 and 3, our team was divided into community outreach and participant outreach teams, with a small number of team members crossed-trained on tasks associated with both phases.
Recruitment Site Database Construction
To begin, a team of 6 research assistants created a community outreach database, to serve as a repository for potential community recruitment sites. The team used coordinated Google searches to identify key contacts (e.g., leaders, site coordinators) at groups and/or organizations (e.g., support groups, resource centers) across the U.S. who may have access to spousal caregivers of people living with ADRD; each team member used unique search phrases to target different geographical areas of interest (e.g., “ADRD support groups Washington”). Targeted areas of interest were chosen to create geographic diversity, encompassing 48 states within the US. Overall, our team identified over 1000 possible sites, then recorded site information in the community outreach database. Information recorded in the community outreach database included organization title, location, point of contact, and whether they appeared to serve minority populations. We also targeted areas that we anticipated would be higher in population density (e.g., metropolitan areas) with the assumption being that organizations stationed in these areas may have access to a larger number of caregivers. Organizations were then sorted to prioritize those outreach sites that were most likely to include historically underrepresented caregivers.
Community Outreach
We focused on about 25 cases in the community outreach database at a time, opening about 10–20 new cases on a rolling basis, as previous cases were exhausted. We created standard operating procedures to ensure effective, professional, and consistent communication during (“cold”) phone calls and e-mails; these included phone scripts, email templates, and a study flyer. We made a minimum of three attempts before discontinuing contact with an unresponsive site. We e-mailed a flyer to sites that requested study information or to those unresponsive by phone. We made no more than three contact attempts unless a site had explicitly expressed interest. We asked interested community sites to distribute the focus group flyer to their stakeholders and worked with them to determine the most appropriate route of distribution (e.g., e-mailing flyer via listserv, sharing with a support group, and posting study on webpages frequented by caregivers). We conducted follow-up calls to thank the contact for their commitment, assess their progress, and address challenges. We documented all call details and action items in the community outreach database.
Participant Outreach
Potential participants contacted the research team using information (phone or e-mail) posted on a recruitment flyer; once in contact by phone, a team member communicated the details of the study to the participant, and then if the participant was still interested and identified as a spousal caregiver, we conducted an eligibility screening. Eligible participants must have self-identified as the principal caregiver for a spouse with a physician based ADRD diagnosis, devoted at least four hours of daily care to their spouse for at least the last three months, and have been married or committed to their spouse for at least three years. At the beginning of our study, we also required that participants be aged 60 years or more, however, we later relaxed our age requirement to be more inclusive of all spousal caregivers. If eligible, we then informed participants of the medical release form (MRF) process, which included first, asking the caregiver to review, complete, and sign the MRF. Once received, the research team faxed the completed MRF to the spouse’s medical provider the participant listed on the MRF to formally request ADRD diagnosis disclosure. We contacted medical providers’ offices as much as required to confirm ADRD diagnosis. Most often, the medical provider’s office faxed medical records; upon receipt, our team confirmed ADRD diagnosis within the medical records, and then deidentified, redacted, and filed the minimum necessary for our records (typically one page), destroying the remaining pages, for confidentiality. Then, we contacted participants to schedule a virtual introductory session, during which participants provided written informed consent, and subsequently participated in a focus group.
Results
We contacted 222 sites in the community outreach phase (see Figure 1 for outcomes). It took approximately 3–4 weeks of active community outreach calls to find a community site willing to share our study materials. Of the sites we successfully reached, 61.6% were open to distributing study information and 25 (22.3%) later confirmed that they shared our study information with stakeholders. These organizations included religious groups, senior centers, national Alzheimer’s Disease associations, and community support and resource organizations for individuals with ADRD and their caregivers. In contrast, 38.4% of sites were unable or unwilling to share. Notably, these proportions do not include 22 sites we were actively approaching when we met our target sample size and halted recruitment; thus, it is possible that a larger portion of sites may have helped had we continued to ask. Depicts community and participant recruitment flow.
Forty-six percent of the potential participants who contacted us completed an eligibility screening, while others were deemed ineligible before completing the formal eligibility questionnaire (28.6%) or lost via discontinued contact (17.9%). Eighty-five percent of those screened were eligible and sent an MRF; 81.8% of those individuals completed & returned MRFs, which we faxed to medical offices. Of the eligible participants, we lost 28% for MRF-related reasons (e.g., participants not returning MRFs or the research team losing contact with their medical office). It took an average of 2.61 (SD = 2.56) contact attempts to medical offices before receiving diagnosis disclosure via fax. Of the 18 potential participants who confirmed a spousal ADRD diagnosis, we lost 3 (16.6%) to scheduling conflicts and 1 (5.6%) enrolled in a conflicting study. Of note, we developed an evidenced-based system to identify bots/fraudulent participants after being infiltrated with suspicious e-mails (based on Pozzar et al., 2020), and discontinued contact with 4 (7.1%).
Discussion
We conducted nationwide community outreach to successfully recruit ADRD spousal caregivers into virtual focus groups. Several novel community partnerships were created through this effort, as none of the participants reported here were recruited from previous partners. Maintaining these partnerships can form a recruitment pipeline to sustain a larger research program with spousal caregivers. However, working with organizations that have existing ties to the research group, if available, may also be effective for subject recruitment. For our upcoming randomized clinical trial, we will first call upon previously successful sites to distribute study information, then will reopen community outreach with the 700 remaining sites in our community outreach database, focusing on 20 sites at a time. The research team worked with sites to identify the most appropriate and efficient way for each contact to distribute our study information based on the organization’s structure, function, and/or resources. This flexibility allowed organizations the autonomy to share the information in ways that would best fit their messaging, mission, and audience. This flexibility, along with personalized messaging, helped build positive relationships with community sites.
We lost a significant proportion of potential participants during the diagnosis conformation step. Many caregiving studies do not specify whether diagnosis was confirmed (e.g., Travis et al., 2024; Williams et al., 2010); thus, requiring only that people self-identify as caregivers may be sufficient. Moreover, this method enhances inclusivity. However, eliminating diagnosis-confirming steps may also create a loophole for fraudulent research engagement (e.g., “bot” infiltration). Fraud can be combated using evidence-based recommendations (e.g., Pozzar et al., 2020), but diagnosis confirmation adds another layer of confidence in the substance of our data.
Recruiting ADRD spousal caregivers may be more difficult than recruiting other familial caregivers (Baker et al., 2023). More research is needed to better understand which methods are most efficient at recruiting people who care for a spouse living with ADRD into research studies. Nevertheless, these data demonstrate that a grassroots approach to recruiting ADRD spousal caregivers can work, even for research teams without connections to a medical center or collaborating physician(s). The nature of our data collection (i.e., focus groups) required that participants not only be eligible for the study but also be available to participate synchronously with three to five other participants. Researchers using more individualized data collection methods may see better yields than those reported here. Further, for inclusivity, we relaxed our caregivers’ age requirement as the study continued; other studies may not have this same flexibility. Further, because we were hoping to find new sources of caregivers rather quickly, we targeted geographical areas that we anticipated may have access to a higher number of caregivers (e.g., larger metropolitan areas); however, it should be noted that this approach excludes caregivers in more rural areas who may face unique challenges. Future research should make efforts to utilize the methods reported here to reach caregivers living in more rural geographical areas.
It took about three months of active community outreach to enroll our first participant. However, a fax number error on our MRF template may have unintentionally increased this lead time. Nine community sites were ultimately the source of all enrolled participants reported here. Personal referrals seemed more effective than posting recruitment announcements on websites. Maintaining relationships with enthusiastic community partners, including personalized messages (e.g., thank you notes) may be key to recruitment success, over time.
Footnotes
Acknowledgements
We thank our research team and collaborators for their commitment to and support of the planning and data collection for this project—especially Dr. Chris Fagundes, Ann Wang, Kristi Parker, Alex Wherrey, Grant Shrable, Shreyas Srinivasan, Emily Kolb, Dylan Puepke, Ashton Richards, Madelyn Edlund, Victoria Norton, Caroline Wilkerson, as well as Nola Majoros. We would also like to thank Dr. Jonathan Singer, whose enthusiasm for our project and recruitment strategy motivated us to write this recruitment article. Most importantly, we thank the caregivers who participated in our study and so generously and selflessly shared their insight and experiences with us, for the greater good of those whose families are impacted by Alzheimer’s or a related Dementia.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was funded by an NIH K01 Career Development Award from the National Institute on Aging (K01AG073824-02, PI: Angie LeRoy).
