CommunityRx: Optimizing a community resource referral intervention for minority dementia caregivers

Abstract

Community resource referral systems have been implemented into care settings that serve persons with dementia, but with little input from caregivers. Focus groups were conducted with African American, Hispanic and Asian caregivers to describe their preferences for community resource referral information. Caregivers discussed the significance of a community resource list for dementia caregiving and self-care and articulated strategies for effective information delivery during a medical visit. Most caregivers acknowledged that resource needs change with progression of dementia, but no patterns emerged with regard to preference for information delivered incrementally based on disease stage or all at once. Hispanic and Asian caregivers felt that resource information should specify service providers’ language and cultural capabilities. All caregivers agreed that delivery by a member of the care team with knowledge of dementia-specific resources would be most effective. Optimal delivery of community resource referrals is caregiver-centered and customizable to individual and sub-group preferences.

INTRODUCTION

The National Plan to Address Alzheimer’s Disease, developed under the National Alzheimer’s Project Act (NAPA), recognizes that caregivers of persons with dementia require support beyond the healthcare system (US Department of Health and Human Services, 2019). NAPA calls for advanced coordination of care, including clinic-based referrals to community-based supports for caregivers of and persons with dementia, especially those from racial and ethnic minority groups. In parallel, the National Academy of Medicine has called for better integration of social and medical care to address social risks that may impact health, like food insecurity or unstable housing (National Academies of Sciences, Engineering, and Medicine, 2019). With the shift to value-based payment models that promote improved care coordination and population health (Burwell, 2015), healthcare systems are rapidly adopting information systems designed to better connect patients to community resources (Gottlieb et al., 2017). However, effective and inclusive care integration strategies often lack caregiver input, especially from those of racial and ethnic minority groups (Feinberg, 2017).

Research on interventions to support caregivers substantially underreport outcomes by race or ethnicity, (Gilmore-Bykovskyi et al., 2018) and systematic and reliable assessment of specific resource needs among minority caregivers is scarce (Bangerter et al., 2019). In a prior study, the highest rated unmet need among caregivers (32% identifying as a race or ethnicity other than white) was for community resource referrals (90%) (Black et al., 2013). In an in-depth qualitative study of African American caregivers, we identified a need for places for socialization (both for themselves and for the care recipient) and for information about hospice services (Abramsohn et al., 2019). Caregivers also suggested including detailed eligibility criteria and an indication of services with providers who were trained in dementia care needs to improve delivery of clinical referrals to community resources. Studies of Hispanic caregivers note lack of support as a main factor in experiencing burden (Balbim, Magallanes, et al., 2019) and high interest in receiving information to support self-care (Xie et al., 2018). One largescale study of Asian American adults (N=2,609) found that awareness of dementia-related services increased their odds of having a plan should they develop dementia in the future (Jang et al., 2018).

CommunityRx is an evidence-based intervention, developed with community stakeholders, delivered at the point of care that connects patients to community-based supports for basic socioeconomic, wellness, disease self-management and caregiving needs. By integrating with the electronic medical record, CommunityRx algorithms generate a personalized community resource list (“HealtheRx”) specific to the individual’s age, gender, language and conditions (Lindau et al., 2016). A recent real-world trial of the impact of the HealtheRx among middle age and older adults (90% African American), demonstrated increased knowledge of and self-efficacy for finding community resources (Lindau et al., 2019; Tung et al., 2020). A HealtheRx for caregivers of persons with dementia (“HealtheRx-Dementia”) was developed with input from African American caregivers that included social and self-care resources and indication of a service’s capability to care for persons with dementia (Abramsohn et al., 2019). The current study extends this prior work and aims to evaluate the experience of caregivers from other racial and ethnic minority groups, including Hispanic and Asian caregivers, and garner their feedback on the HealtheRx-Dementia. Additionally, we begin to identify strategies to optimize delivery of community resource referrals to a diversity of caregivers.

METHODS

Study Design and Participants

This qualitative study enrolled a purposeful sample (Palinkas et al., 2015) of informal (unpaid) caregivers of persons with dementia who self-identified as African American/Black, Hispanic or Asian. Caregivers who were ≥18 years, English- or Spanish-speaking and lived in the city of Chicago were eligible. Caregivers were recruited from three healthcare systems located on Chicago’s South and West sides. Recruitment of caregivers included outreach to caregivers: 1) who had previously participated in research, 2) at support groups, 3) through recruitment flyers in geriatrics and primary care clinics, 4) through clinicians in these clinics who were informed of the study, and 5) through dementia caregiver community groups and caregivers’ social networks. Institutional Review Boards from all three recruiting sites approved this study; all participants provided written documentation of informed consent and received $50 for their participation.

Data Collection

Data were collected via focus groups for their efficiency and ability to foster dialogue and elicit similarities and differences in experiences within a racially or ethnically homogenous group (Kitzinger, 1995). Nine focus groups, stratified by race/ethnicity, were conducted on-site at the recruitment locations by one of three researchers (E.A., C.G., D.M.), with at least one other researcher present for note taking and support. Focus groups were approximately 90 minutes long and conducted in English and Spanish. Facilitation of focus group discussions used a semi-structured focus group guide (Supplemental Material, Focus Group Guide). Our prior research with African American dementia caregivers informed the domains of inquiry in the focus group guide and a theory of action, explaining how the HealtheRx-Dementia might address caregiver and care recipient needs (Abramsohn et al., 2019). Focus group facilitators included women and men with doctorate-, master’s- and bachelor-level expertise in qualitative data collection methods and focus group facilitation. No care recipients were present during the focus groups. A brief, self-administered survey collected basic caregiver sociodemographic characteristics (e.g., age, race/ethnicity, gender, insurance) (Centers for Disease Control and Prevention [CDC], 2018), self-reported health (CDC, 2018), comorbidities (Chaudhry et al., 2005) and caregiving characteristics (e.g., relationship to care recipient, years caring for person with dementia) (Avison et al., 2018). Caregivers then reviewed a list of community service types indicated for dementia caregiving and self-care developed using extant literature, expert opinion and dementia caregiver feedback (Abramsohn et al., 2019) (Figure 1). Caregivers identified resources on the list they would use for the care recipient, for themselves, or that they would not use at all. Community resources associated with these service types were used to generate the HealtheRx-Dementia (Figure 2), which caregivers reviewed. We assessed caregivers’ definitions of “dementia friendly” resources, a theme surfaced in our prior study (Abramsohn et al., 2019), and whether labeling resources as such would facilitate use. Caregivers reflected upon their preferences for resource information delivered incrementally according to the person with dementia’s disease stage, or whether they would prefer to receive this information all at once. Finally, caregivers were asked to give input about who on the healthcare team should deliver resource information, preferred mode of delivery and when in the course of a typical clinical encounter would be best to receive this information.

Figure 1. List of community resource types and services indicated in extant literature and by expert opinion for dementia caregiving and self-care.

This visualization was used to elicit responses from caregivers about the types of resources they would use for caregiving, self-care or would not use.

Figure 2.

Sample HealtheRx-Dementia

Data Analysis

Focus groups were audio-recorded and transcribed; those conducted in Spanish were translated prior to transcription. Transcription was completed verbatim; while focus group transcripts were not returned to participants for review or correction, any words or phrases that were difficult to understand or transcribe were reviewed by a researcher who was present during the focus group. We deductively applied the constant comparative method of analysis (Glaser, 1965; Onwuegbuzie et al., 2009), which is useful when there are multiple focus groups under one study to allow for comparison within and across groups. Data were analyzed in three phases with racial/ethnic group as the unit of analysis in the second and third phases.

The first phase used open coding (Glaser, 1992), wherein three analysts read transcripts from each of nine focus groups (4 with African American, 3 with Hispanic and 2 with Asian caregivers) to become situated in the data, noting key words or phrases that addressed questions or concepts aligned with the domains of inquiry. Then, analysts assigned brief descriptive words/phrases (codes) to noted passages. To ensure interrater agreeability and fidelity of research findings, one analyst coded a set of three transcripts, then provided their coded transcripts to another analyst for review; this analyst either confirmed the codes assigned or flagged codes for discussion. This process continued until all nine transcripts were coded or reviewed by all three primary analysts. Primary analysts met regularly to discuss and iterate codes and resolve interpretive differences until consensus was achieved or adjudication was needed.

The second phase of analysis employed axial coding to group thematically similar codes into broader categories, grouped by domain of inquiry. For example, one code was removed (“resource knowledge is necessary for preparation”) and incorporated into the definition of another code (“I need resources for all stages of dementia at once”) because they were conceptually similar. Final codes and categories were consolidated into a single codebook. The codebook enabled researchers to compare codes, categories and themes at multiple units of analysis including individual focus groups and across the focus groups stratified by race/ethnicity.

The third phase of analysis used selective coding (Corbin & Strauss, 2014) to identify and describe emergent themes informed by the relationships between the coded data. Conclusions drawn from the qualitative data were reviewed in the context of the entire dataset with the goal of finding discrepant information. We determined that thematic saturation was met when no new findings emerged in the domains of inquiry defined a priori, as described above. This approach, described by Saunders et al. (2018) as “a priori thematic saturation,” is used to exemplify (rather than develop) existing theory. ATLAS.ti version 7.5.18 (Scientific Software Development, Berlin, Germany) was used for coding, analysis and textual extraction. Direct quotes are noted with double quotation marks and an indicator of race or ethnicity (African American (AA); Hispanic (H); Asian (A)).

RESULTS

Seven focus groups and two, individual semi-structured interviews were conducted, stratified by race/ethnicity (Supplemental Material, Table 1). Table 1 describes caregiver sociodemographic, health and caregiving characteristics. Caregivers were predominantly women. African American caregivers (N=10) tended to be older than Hispanic (N=7) and Asian caregivers (N=5) but reported fewer years spent providing care. African American and Asian caregivers tended to self-report better health than Hispanic caregivers. Hispanic and Asian caregivers tended to be the adult child of the care recipient, whereas African American caregivers were typically spouses.

Table 1.

Caregiver sociodemographic, health and caregiving characteristics, stratified by race and ethnicity

	African American	Hispanic	Asian
	(n=10)	(n=7)	(n=5)
Sociodemographic characteristics
Age in yearsa (median, range)	64 (45-77)	51 (29-66)	57 (25-69)
Gender
Women	8	6	2
Insuranceb
Insurance through work	4	2	1
Medicare	3	2	1
Private insurance	2	1	2
Medicaid	1	1	1
Health characteristics
Self-reported health
Excellent, very good or good	7	3	5
Fair or poor	3	4	0
Common medical conditionsb
High blood pressure or hypertension	5	3	1
Osteo- or rheumatoid arthritis	3	3	0
Diabetes or high blood sugar	0	3	0
Heart condition	1	1	1
Caregiving characteristics
Relation to PWD
Adult child	3	5	5
Spouse	6	2	0
Years caring for person with dementia (median, range)	3 (1-6)	10 (5-18)	10 (5-20+)
Number of visits to the doctor with care recipient in a typical month	1-2	1-2	1

^aTwo African American caregivers did not answer this question.

^bResponses are not mutually exclusive.

Across all racial and ethnic groups, findings fell into two main domains of inquiry, outlined in the focus group guide: 1) caregivers’ community resource needs and referral information preferences and 2) caregivers’ perspectives on how best to deliver community resource information at the point of care. Within each domain, we present emergent themes and discuss findings that were similar across all focus groups, those that were different and any findings that were unique to one particular racial or ethnic caregiver group.

Caregivers’ community resource needs and referral information preferences

Theme 1: Minority caregiver’s caregiving needs are more similar across racial and ethnic groups than their self-care needs.

Caregiving needs

Caregivers across all race/ethnic groups in this study identified recreational and community-type service needs, including group meals and adult daycare, among their unmet caregiving needs (Table 3). The most common caregiving needs identified by Hispanic caregivers overlapped with both African American and Asian caregiver needs, including medical and healthcare-type services (e.g. home health or medical supply resources) and social services (e.g. financial support, legal assistance). African American and Asian caregivers identified specific needs for transportation and mental health resources, respectively.

Table 3.

Comparison of main resource needs for caregiving and self-care among caregivers

	Resource needs for caregiving:	Resource needs for self-care:	Comparison between types of needs:
African American caregivers	Medical & Healthcare Recreation & Community Services Transportation	Personal & Entertainment Physical Fitness Social Services	Caregivers identified different needs for caregiving compared to self-care.
Hispanic caregivers	Medical & Healthcare Recreation & Community Services Social Services		Caregivers identified the same needs for caregiving and self-care.
Asian caregivers	Recreation & Community Services Social Services Mental Health	Personal & Entertainment Vocational & Educational Mental Health	Caregivers identified few overlapping needs for caregiving compared to self-care needs.
Comparison between racial/ethnic groups	AA, H and A caregivers identified some overlapping needs for caregiving	AA, H and A caregivers identify few overlapping needs for self-care	OVERALL COMPARISON: Minority caregiver’s caregiving needs are more similar across racial and ethnic groups than their self-care needs.

Abbreviations: A: Asian; AA: African American; H: Hispanic

Self-care needs

Among seven commonly identified self-care needs, only two - personal and entertainment resources and social services - were identified by more than one racial/ethnic group (Table 3). African American caregivers also identified the need for physical fitness resources, a service no other group commonly noted. Asian caregivers were the only group to commonly cite vocational and educational and mental health resources. When asked about resources they would not likely use, some caregivers in each racial and ethnic group identified housing resources, indicating they did not yet see a need for them.

Caregivers across all racial and ethnic groups noted the value of indicating which resources on the HealtheRx-Dementia were “dementia friendly,” defined by caregivers in this study as services with providers who are “patient, compassionate, empathetic” [H] in their interactions with persons with dementia. Caregivers also wanted providers to be knowledgeable about how to respond to behavioral and psychological symptoms of dementia. One caregiver summarized, “[dementia friendliness] would mean that they would understand behaviors…so they may understand anxiety or hesitation or something, they would know how to interact with the person.” [AA]. Some caregivers noted that training is necessary to achieve the appropriate level of competency among dementia care providers, both in the clinical setting and in the community, “Otherwise, it’s a waste of time to any of us.” [AA]. An indicator of dementia friendliness would allow caregivers to “feel comfortable to go in a place like that” [H] and would prompt caregivers to use these services over others.

Theme 2: The need for language-specific and culturally tailored community resource referrals.

Hispanic and Asian caregivers in this study consistently expressed preference for community resources that were available in their preferred language and culturally tailored to their respective backgrounds. Resources meeting these criteria were often characterized as a better fit for these caregivers and the care recipient; for example, one caregiver stated: “To me, that's been the biggest impact. Having other people come in and help that are not family members. She is god sent because she's bilingual. She's Spanish speaking, she knows the language, the culture.” [H]. In reference to culturally tailored resources, some Asian caregivers specifically pointed to food-related resources as an example, and the care recipient’s preference for one place over another. As one caregiver explains: “… [NAME OF PLACE 1] is the only one that provides the Chinese meal. And the senior building I used to work with, they have [NAME OF PLACE 2], you know, (makes face and lowers voice) they don’t like that much… they don’t like the food. The [NAME OF PLACE 2] is good, but specific Chinese meal, this is only one facility. And they would love to go there but, you know, it’s full now.” [A]. Another caregiver in this focus group knew of this place that served culturally-specific meals that their care recipient preferred and agreed with this sentiment. Lack of knowledge or access to language-specific and culturally tailored resources was noted as a major barrier to use:

“It's very hard to find anybody that we can get counseling for him that's Spanishspeaking. We've called our insurance, and it's very hard. They have a lot of women, but he feels comfortable with a man, and he feels comfortable speaking Spanish.” [H]

“That’s my biggest issue right now, or has been, is, even to find an adult day center. I ended up having to do my own research and eventually found a place that, you know, it’s actually in Chinatown. So it’s somewhere where she would feel comfortable being at. Even though there’s a lot of services here, there is a lot of [services] that she would not feel comfortable [with] because it’s not her culture.” [A]

Theme 3: Include information about eligibility, cost and insurance.

Across all racial and ethnic groups, caregivers in this study indicated that it was important to include information about eligibility, cost and insurance types accepted by community-based service providers on the HealtheRx-Dementia. Including this information would abate some of the frustration of navigating complicated healthcare and community-based systems when searching for resources. Many caregivers noted that these details were often deciding factors for resource use. One caregiver noted, “I can’t put him in places unless I am willing to come out of pocket. Well, if we could come out of pocket, you wouldn’t have to be worrying about him because I’d be hiring somebody to come into my home.” [AA]. Another caregiver questioned the utility of including resources on the guide that were costly, especially in reference to where resources are located within the city. This caregiver stated: “Yeah, you have to look at the poverty rate. Self-pay is a very major factor on going or not going [to a community resource].” [AA].

Theme 4: Caregivers’ informational needs are dynamic, especially as they relate to dementia stage.

Caregivers across all racial and ethnic groups acknowledged that the resource needs of their care recipient will change with the progression of their dementia, and that certain resources, such as assisted living or hospice care, may be more relevant during more advanced stages of disease. As one caregiver explained, “We haven’t really gone down the path of housing or assisted living facilities yet for my dad because he’s not at that stage, uh, hopefully for a long time.” [A].

No consistent patterns emerged among racial or ethnic groups to describe caregivers’ preferences for resource information delivered incrementally based on disease stage or all at once. While not systematically assessed among Hispanic caregivers, some African American and Asian caregivers in this study characterized the benefit of receiving resource referral information for all disease stages at once in terms of logistical and emotional preparation for the future (Table 2). There were a few caregivers across racial and ethnic groups who preferred to receive information only relevant to the care recipient’s current stage of disease, citing feelings of “overwhelm” and “fear” about the future.

Table 2.

Exemplary quotes illustrating caregivers’ varied perspectives on the provision of resource referrals by dementia disease stage^a

Exemplary quote	Race/Ethnicity of Caregiver
In support of resource referral information all at once
“But at we’re getting close to where, you know, at some point she may not benefit from going to adult day center…so I would like to have that information actually kind of now so I know rather than to wait until we’re there and then we’re kind of scrambling to try to figure it out.”	Asian
“All at once. And we can prepare beforehand, even though it’s not at that stage.”	Asian
“Yeah, so, I think all that information at one time would be good because, just like, now with my wife. She can do a lot of things on her own, you know. And when I come to [support] group sometimes it’s, it’s like I listen to, my other group members and my wife is not there yet. But I could see a change coming, you know? I can see it coming. That would be helpful to know, you know, the different stages, where you can go, who is financially going to help you, what the prices are going to be.”	African American
“I’d like to know where I take my next step. I wouldn’t want to be standing there and fall off the cliff!”	African American
“And I think that you need to know what’s coming. And you don’t expect it and you don’t know until it happens. Then by the time it happens, you could have possibly minimized or prevented it, with knowing that information, it could have already been addressed. And by not knowing it, simply by the fact that it’s too much information, what you do come across is like, wow I coulda did this way back here and prevented it or minimized.”	African American
In support of resource referral information delivered incrementally by disease stage
“I think it would be best if you had beginning, middle and ending… All separate. I think that would be good. Because that might be frightening to some people. Yeah, it would be quite overwhelming and then a lot of people [caregivers] might just give up… So you don’t want to stop people if they are able to do it. Don’t wanna discourage them.”	African American
“For me, [information] as I need it. Too much information at once.”	African American
“I do know that I’m gonna have to sell the house. I think I’m gonna have to sell it at some point. Because I can keep seeing him going somewhere. You know, having somewhere to go. And we’re in a row house, so there is no, you know, there’s no protection around there. And, I don’t know, there’s just a lot of changes that I think are going to have to come about, that I don’t necessarily want to make right now.”	African American

^aThis question was not systematically assessed among Hispanic caregivers.

Caregivers’ perspectives on how best to deliver community resource information at the point of care

Caregivers from all racial and ethnic groups provided suggestions to optimize community resource information delivery in terms of who should deliver it, and how and when in the continuum of care. However, African American caregivers in this study predominantly described the physician in a role of informational authority, and that their referral recommendations carried weight with the care recipient. This feeling of physician authority was not necessarily raised by the caregivers themselves, however, but caregivers would use the clinical encounter to “reinforce what we’re already saying at home. Because when you’re not hearing us [caregivers], maybe you’ll hear him [physician]” [AA]. As one caregiver stated, “It’s the trust that you would have with the doctor. Trust, trust, trust. It would be more trustworthy to get it from a doctor.” [AA]. While this finding was consistent among African American caregivers and across their focus groups, this sentiment was not as strongly echoed among Hispanic or Asian caregivers.

Theme 1: Delivery of the resource guide by someone on the clinical staff with knowledge about resources.

While the sentiment of physician authority was specific primarily to African American caregivers, caregivers from all racial and ethnic groups noted that physicians have limited time to spend with patients and their caregivers. To address this barrier, caregivers in this study suggested resource referrals be delivered by a knowledgeable member of the clinical team, ideally with a warm hand-off from the treating physician. As one caregiver explained: “Well, it really doesn’t matter [who delivers the HealtheRx-Dementia] as long as they perhaps have some personal knowledge about specific services.” [A]. Caregivers noted that social workers, case managers and nurses often had this knowledge and spent more time with the patient, which was necessary to make quality referrals.

Theme 2: Optimal mode and timing of community resource information delivery varies.

Preferred mode of delivery of community resource information varied widely amongst caregivers in this study, including email, paper, text and social media as viable options. While caregivers generally acknowledged that the preferred mode of delivery would differ person to person, the majority of Asian caregivers indicated a preference for text or email, and the ability to access the HealtheRx-Dementia online. These caregivers cited convenience and the potential to receive automatic notifications should the information be updated. One caregiver noted that paper can be hard to keep track of and suggested that technology use among older caregivers is increasing: “These [handouts] get misplaced and our elders are more so on computers at the elderly center. So they’re learning to click on. It’s just a click and they feel really savvy at that when they use it.” [AA]. Still, some African American and Hispanic caregivers preferred the paper option:

“My fridge is all papered up with information and pictures… Something that I can put on there as an emergency and say, ‘There it is.’" [H]

In terms of timing of delivery of the resource information, caregivers from all racial and ethnic groups commonly described an ideal scenario wherein the patient and their caregiver would be given the HealtheRx-Dementia upon checking in to the clinical encounter. This would allow time to review the information and formulate questions for the care team to be discussed towards the end of the doctor’s visit:

“I think initially when you walk in, so you can read it while you’re waiting. The doctor should be the final point. It shows the seriousness of the needs. And it shows reiteration of what was already introduced. So if they have any questions, and the doctor can say, ‘Oh, I think you should try this’ and they go back to what was initially given to them. So [the patient says], ‘Oh this isn’t just a different paper. This is a doctor prescription!’” [AA]

All caregivers in this study emphasized the importance of optimizing the limited time with the physician to focus on clinical discussions, providing further support to end the visit with the discussion about community resource referrals.

DISCUSSION

African American, Hispanic and Asian caregivers in the current study differed in terms of demographic, health and caregiving characteristics, but commonly identified the need for recreational and community services and health-related resources to support caregiving. Hispanic caregivers, who tended to be younger and in poorer health, also identified a need for health-related resources to support self-care, a finding corroborated by Balbim, Marques, et al. (2019). The avoidance of housing resources such as nursing homes and assisted living facilities is not uncommon among minority caregivers and has been attributed to familial duty or fulfilling cultural norms (Balbim, Magallanes, et al., 2019; Sharlach et al., 2006). These findings, in conjunction with evidence that providing informal dementia care can negatively impact caregivers’ mental and physical health, (Schulz et al., 1995) highlight the need for healthcare providers to systematically ask their patients whether they provide care for people with dementia. (Vickrey et al., 2006). This information could facilitate delivery of caregiver-centered resource referrals to support caregiver health and continuity of care in the home.

To our knowledge, this is the first study to elicit caregivers’ definitions of “dementia friendly” resources and services. Dementia-friendly initiatives aim to raise awareness of dementia-related issues and reduce stigma (Lin, 2017; Lin & Lewis, 2015). However, definitions of the concept vary, are often focused on specific geographic communities or physical spaces and rarely incorporate the perspective of the person with dementia or their caregiver (Hollister et al., 2018; Lin, 2017; Lin & Lewis, 2015). Furthermore, evidence to suggest that these efforts promote caregivers’ actual use of dementia-friendly services is lacking. The novel finding that minority caregivers in this study would prioritize use of dementia-friendly resources, both for themselves and for their care recipients, could be considered when designing dementia care training programs for both clinical and community providers.

Hispanic and Asian caregivers’ preference for resources that are language-specific and culturally tailored has been corroborated by prior studies. Minority caregivers may face barriers to accessing resources, such as experiences of discrimination or stigma, lack of knowledge about resources or language or cultural barriers (Dilworth-Anderson et al., 2002; Epps et al., 2016; Samson et al., 2016; Sharlach et al., 2006). Additionally, information about resource eligibility, cost and insurance were important determinants of resource use for caregivers in this study, especially given the difficulty noted when trying to access services. Most dementia caregiver interventions that include referrals to community resources lack this level of detail (Feinberg, 2017). Systematic, tailored referrals with detailed information may address barriers minority caregivers face when accessing resources.

Caregivers from all racial and ethnic groups acknowledged that dementia stage was important when considering a person’s resource and informational needs, and that needs will change over time. However, there was no agreement regarding whether resource information should be delivered incrementally according to disease stage or all at once. While prior studies indicate that an incremental approach may avoid overwhelming caregivers and persons with dementia (Whitlatch & Orsulic-Jeras, 2018) (a sentiment noted by some caregivers in this study), later-stage caregivers have reported that they would have benefitted from resource information in earlier stages of the disease (Boots et al., 2015).

These findings highlight important implementation considerations. Delivery of resource information by a physician was important especially to African American caregivers in this study given the perception of the physician’s authority. Although corroborated by Peterson et al. (2016), this finding may contradict evidence of higher levels of distrust of physicians among African American people compared to other groups (Lillie-Blanton et al., 2000). Perhaps because they were caregivers – functioning as an extension of the care recipient’s medical team – African American caregivers in this study held physicians in higher regard or had developed more trusting relationships than has been seen in studies of more general populations. At the same time, the majority of all caregivers – including African Americans - preferred that physicians use their limited time with the patient to focus on medical-related needs and use a warm hand-off to a knowledgeable team member to deliver resource information. These findings suggest that endorsement of community supports for dementia by the physician, then delegation of referral-making to another clinician in the practice (e.g. a nurse or social worker) or, perhaps, to a trusted community-based dementia care organization, could facilitate delivery. Implementation of the HealtheRx-Dementia in clinical and community-based settings could be both a mutually reinforcing strategy and would increase access to resource referrals for caregiving and self-care. While mode of information delivery varied widely among all caregivers in this study, caregivers agreed that having this information at the beginning of a doctor’s visit would provide an opportunity to review and ask questions at the end of the visit. Implementation of interventions that connect caregivers to community resources for support should be designed to fit into existing clinical workflows and systematically account for caregiver preference for delivery.

These findings fill a gap in knowledge about under-studied dementia caregiver groups and are particularly informative to a longstanding, community-engaged hyperlocal effort to promote clinic to community connections (Abramsohn et al., 2019; Lindau et al., 2011, 2016, 2019). However, certain limitations should be considered. Findings may not necessarily generalize to other populations due to variation in community resources access, more heterogeneity within racial and ethnic groups than could be captured in a single region study and other factors. While resource needs and information delivery preferences may differ by caregiver gender or relationship to the care recipient, this study was not designed to assess for these differences. Due to the semi-structured nature of the focus groups, not all questions were asked of all groups, but all domains of inquiry were addressed. Although the size of some groups were limited, to our knowledge this is the first study to purposefully sample racial and ethnic minority caregivers to elicit their perspectives on a resource referral intervention to support caregiving and self-care. Additionally, the sample was diverse in terms of age and relationship of the caregiver to the care recipient. All participants were currently providing care, which should limit recall bias.

A diverse group of informal caregivers of persons with dementia in this study saw value in receiving community resource referrals during a clinic visit. Similarities were greater than differences across groups, but a caregiver-centered approach suggests that an effective resource referral intervention will be customizable to individual and sub-group preferences.