Background
There are currently over 46 million people living with dementia worldwide [
1]. In the United States there are 5.4 million people living with Alzheimer’s Disease [
2], and more individuals are affected when other forms of dementia or cognitive impairment are included [
3,
4]. With an aging demographic, these numbers are expected to dramatically rise in coming decades. More than 80 % of people with dementia are cared for at home by one or more informal caregivers, usually unpaid family members or friends [
2,
5,
6]. Although not all identify with the term caregiver, we will use this term as shorthand in this report. More than 15 million are estimated to be informal caregivers of people with dementia in the U.S. [
2,
5,
7]. Because there is often a delay from the time of symptom onset until diagnosis [
7], additional caregivers are likely to be providing care to those who do not yet have an established diagnosis.
The role of caregiver to a person with dementia or cognitive impairment (referred to hereafter as care recipient) often levies enormous costs to time, emotional well-being, physical health, and finances [
8,
9]. Although some caregivers demonstrate resilience and coping strategies that help ameliorate a negative experience [
10,
11], for many it is an extremely challenging role, and one for which most caregivers have little background or training [
7]. More than 80 % of caregivers report needing more information on caregiving topics [
12].
Most caregivers turn to a primary care provider (PCP) for information, but the information obtained there is often limited [
5,
13‐
15]. Randomized clinical trials of more than 200 psychoeducational interventions support efficacy in improving knowledge and support of dementia caregivers [
16]. Caregiver education and support programs improve caregiver confidence; reduce caregiver depression, distress, and upsetting thoughts; improve life satisfaction and response to disruptive behaviors; reduce behavioral and psychological symptoms; and delay nursing home placement of care recipients [
17‐
21]. However, few such programs are widely implemented outside a research setting. Additionally, in the Information Age [
22] a vast amount of beneficial information is publicly available through disease-specific and caregiving organizations, but many caregivers do not access that information [
7].
Multidisciplinary specialty memory/dementia clinics provide comprehensive care that includes extensive caregiver support and education [
23]. Yet, most patients with dementia receive the bulk of their medical care in primary care settings [
24,
25], where resources of busy PCPs to offer comprehensive services and caregiver education are limited.
In this study we sought to better understand the complex determinants that lead informal caregivers of people with cognitive impairment or dementia to recognize their need for caregiving education. Further, we aimed to assess caregivers’ sources for information, barriers to seeking and receiving information, and preferences for information sources. Although the subject of information sources for dementia caregivers has been examined previously in large quantitative surveys [
7,
12], we aimed to explore in depth using qualitative semi-structured interviews caregivers’ perspectives, impressions, and opinions, in order to develop a more rich and nuanced view of this subject.
Discussion
Our study used SSI with content analysis to reveal several prominent themes related to CG perceptions and perspectives about information sources, barriers and preferences, as summarized in Table
6. While several of these themes have previously emerged in broad quantitative surveys, an important novel theme emerged: despite endorsing not having received adequate information from PCPs, caregivers continue to place critical value on their relationship with care recipient PCPs to receive recommendations, guidance and endorsement to sources of caregiving information in the early stages of the caregiving journey.
A first theme that emerged was that
while CGs would not choose for their loved ones to be ill,
many caregivers choose to take on the role of caregiving when needed. On the surface, this finding appears to be at odds with the lack of choice reported by about half of caregivers in an AARP-NAC survey [
12]. However, this may reflect a difference in semantics arising from the different methodologies used to assess choice of caregiving; compared to survey methodology, the SSI methodology enabled us to clarify the CGs’ interpretation of the word choice. Our findings do not support a hypothesis that most caregivers do not seek information because they are resistant to being in the role. An important implication of our findings is that many caregivers could be enlisted by health care providers as active care partners in patient care.
A second set of themes that emerged was that
there is often a long delay between the insidious onset of symptoms and the caregiver’s acknowledgment of disease,
and thus a corresponding delay in a caregiver identifying with the role as a caregiver. Our findings also support those of others [
27‐
30] that
while memory loss is commonly the initial symptom noticed,
behavioral and personality changes are more commonly the most troublesome and difficult symptoms to manage. Importantly, our results reveal that while many caregivers recognize the potential implications of progressive memory loss for a possible dementia diagnosis, they initially may not recognize that behavioral symptoms could indicate a medical disease, and, furthermore, appear unaware of the existence of non-pharmacological strategies available to help with these symptoms.
A potential confound to recognizing, and measuring, clinically significant cognitive decline due to dementing conditions is the problem of appropriately considering baseline level of premorbid intellectual function, or cognitive reserve: the individual differences in how people process tasks allow some to cope better than others with brain pathology [
31,
32]. Conceptually, individuals with higher cognitive reserve show more resilience to, and can better functionally compensate for, multiple forms of neurologic insult, as well as depression [
33]; they also manifest slower rates of decline in normal cognitive aging, and can have delayed onset, but faster apparent decline in AD/dementia [
31]. Our study did not consider the potential confound of cognitive reserve on the dynamics of caregiver recognition of clinical effects of disease and their own need for education; it is plausible that these would both be delayed in cases when care recipients have higher reserve.
We also found that caregivers often lack the vocabulary to describe symptoms to the PCP or to search for information themselves, especially if a diagnosis has not yet been established or discussed. However, CGs in our study reported that with progression of symptoms, they eventually alerted the PCP, rather than the PCP first detecting and assessing the problem.
Important implications of these findings are that improving caregivers’ ability to recognize and describe both cognitive and behavioral symptoms, as well as acknowledging their own role as caregiver, is likely to prompt earlier diagnosis and earlier provision of caregiver information. Public education campaigns could give caregivers a more effective vocabulary to report symptoms to PCPs. Furthermore, effective screening by the PCP for cognitive symptoms as part of the Medicare Annual Wellness Visit [
34], and importantly, also effectively screening for behavioral symptoms, could prompt earlier diagnosis and earlier provision of vital information.
Another important theme that emerged was that caregivers hold a relatively holistic view of medical diagnosis/treatment and behavioral strategies/caregiver support. While health care providers may conceptualize and organize information to distinguish medical diagnosis and treatment from symptom management and supportive care, this may not be a meaningful or helpful distinction to most caregivers. Though physicians may focus on medical diagnosis and treatment, leaving the latter to other allied health providers (e.g. nursing, social work, palliative care), our results reveal that this was not a paradigm that makes sense to caregivers, who expect guidance about both a disease and caregiving strategies to come from the same sources, preferably initiated by the PCP. This implies that effective caregiver education should include early access to both medical and caregiving information.
The final crucial theme that emerged is that despite CGs reporting that they have received little information from PCPs or from other publically available sources, they endorse a strong trust in, and reliance on the PCP as the first source of health and caregiving information. CGs were sympathetic in their assessment that PCPs may not have the knowledge or time to provide all of this information themselves, but still expected the PCP to be their port of entry into reliable health information, and to make specific recommendations to direct them to the information they need. While CGs were open to learning caregiving information from different sources in a variety of settings and technologies, no dominant source, setting, or technology emerged as being crucial for all CG, and it may be that a variety of sources is necessary to reach different caregivers. The most important factor to validate use of available information was for their PCP to make a specific recommendation, referral or endorsement. The implication of this finding is that there is vital necessity for implementing better methods to facilitate detection and management of clinically relevant cognitive and behavioral changes in the PCP office (e.g. by providing and incentivizing specific dementia-related education, remuneration and resources) such that PCPs can either directly provide accurate diagnostic, treatment, education and caregiver support themselves, or can refer patients and caregivers to other reputable sources. Optimally, both would occur, with the PCP beginning the process, and then reinforcing and managing it in collaboration with other referrals and resources.
Our study has several important strengths, in addition to some limitations. A primary strength of the study is the advantageous and novel application of rigorous SSI and content analysis techniques to assess our specific study aims. Several previous qualitative studies focusing on a wide range of CG experiences have been reported [
13‐
15,
35‐
40]. But to our knowledge, this is the first study using SSI methods to specifically focus on the perceptions and issues of information needs, barriers, and preferences for dementia caregivers. Unlike large population surveys containing data on caregiver information sources reported previously, SSI allowed in-depth and nuanced explorations that gained a rich narrative of data from a relatively modest sample of caregiver subjects regarding their perceptions and impressions, thus enriching the understanding of the rationale and motivations underlying caregiver responses. Going beyond the general pre-determined topic questions, follow-up interview questions tailored to individual subject responses allowed us to obtain both a broader (and more sensitive) and deeper (more specific and precise) understanding of the topics. Although we quantified and reported the incidence of emerging themes among the interviews, a limitation of using SSI for data collection is that the quantification may not be as precise as would be expected with a structured survey using specific pre-determined questions. Nonetheless, we considered the frequency of emerging themes arising in the interviews to be an indication of their relative importance, and therefore included this data in our report.
A possible limitation of our study is that the interviews were not recorded, although the interviewer took extensive written notes during the interview, including representative verbatim quotes. However, this approach did ensure anonymity of the participants and confidentiality of the interview content, since the investigators who would later see the CR and CG for clinical purposes would not be able to recognize the voices from a taped interview or potential specific identifying content from a verbatim transcript. Although we acknowledge that this method of data collection may have resulted in incomplete data retention, in designing the study we balanced the potential limitation in data rigor and precision in favor of the strength of ensuring full anonymity of the participants and confidentiality of their responses.
Another study strength is that it mirrored caregiver demographic characteristics of previous large-population surveys of caregivers in the United States [
2,
5,
7,
12], and is likely to be representative of caregiver perspectives, particularly those caregivers living in urban areas of the U.S. While, unknown factors that prompted referral of these CRs to our urban specialty memory clinic might potentially distinguish our group of CGs from those in the general U.S. population, the CGs in our study include representative subjects of both genders, a wide range of ages, educational backgrounds, and durations of CR symptoms. A potential limitation of our study, that may underestimate the relative magnitude of barriers to caregivers receiving information, is that we did not specifically query the ethnic and racial characteristics of our subjects. All CGs were fluent in spoken English (and reported being fluent in written English), but for minority English-non-fluent CGs not included in the study the identified barriers are predicted to be even more pronounced. Similar to prior population surveys, our CGs were relatively well-educated, but the most salient barriers are likely to have an even greater impact on less educated caregivers. For example, the top three barriers to seeking information (see Table
5) were responsible for over 75 % of the identified barriers – all these barriers are knowledge/education-based barriers.
While CGs in our study were, on average, college-educated and about 59 years old, 42 % of the major causes for information source barriers identified were due to Internet searches and sources being frustrating or unreliable, (45 % were due to PCPs not providing sufficient information), and 82 % of CGs reported wanting the PCP to be the primary source of information and referral. These issues may be even more present in caregivers with less education and facility with the Internet, and in those who place greater reliance on their PCP.
Another potential limitation, that may limit generalization to a much broader population of silent caregivers who are at earlier stages of the dementia evaluation process or caregiving journey, is possible bias in CG selection. The CGs in our study were already down a path of referral to our single specialty memory clinic; as such their perspectives may not fully represent those of caregivers not referred to our clinic or caregivers even earlier in the process, or even their own perceptions during a previous stage in their caregiving journey – though they were asked about this, their responses required retrospective thought about their prior state. Also, while CGs were interviewed in a systematic way for this study, the CRs they described were not directly examined by study staff for the purposes of this study, so the determination of functional severity reported was an estimate based on the description by the CG; however, even in more formal classification schemes, functional severity is typically determined via caregiver report. Finally, of 77 potential participants identified by screening, we completed interviews for only 27, thus raising the possibility that the participants we interviewed and their responses were not fully representative of the entire group.
Despite wide availability of information on a plethora of caregiving topics, most CG in our study still expected the PCP to be the first source of medical and caregiving information and referral. However, there remains a disparity between expectation and actual experience; our results support findings from prior surveys in which only a third of caregivers report that a health care provider had inquired what was needed to care for the care recipient; even fewer caregivers are reportedly asked about self-care [
7,
12].
Our results further illuminate the critical value caregivers place on their relationship with care recipient PCPs. CGs were open to a range of information sources including referral to allied health professionals or medical specialists, written materials, support groups, classes, Internet sites, or a variety of other technological solutions, but they want the PCP to recommend and endorse specific sources. These findings provide further support that caregivers want and could benefit from receiving education and information about dementia diagnosis and treatments (including prognosis and expectations for treatments), and for learning strategies to address issues that commonly arise when providing care; issues such as behavioral changes, communication, daily activities, advance care planning, respite, and self-care. Providing multifactorial approaches to early detection, management, psychoeducation, care and support for patient-caregiver dyads is associated with better long-term outcomes in Alzheimer’s disease [
41‐
45].
Current dementia caregivers are predominantly older; thus it is possible that the ways caregivers seek and receive information will evolve in coming years as a younger generation, who has grown up using technology even more extensively, become caregivers. Future caregivers may prove to be more facile and reliant on technological information sources than is the current generation. However, it is our contention that the important human element of medicine is unlikely to change, and caregivers’ and patients’ trust and reliance on the PCP as the first authority and arbitrator on health issues will endure. Furthermore, with an increasing number of dementia patients and a shortage of dementia specialists available to diagnose and treat them, PCPs will continue to have a prominent role in their diagnosis, treatment, and providing education and support for their caregivers [
7,
24,
46].