Family Caregivers’ Experiences With Health Care Workers in the Care of Older Adults With Activity Limitations

IMPORTANCE Family and unpaid caregivers often play an active role in managing the care of older adults with activity limitations. OBJECTIVE To examine caregivers’ experiences with older adults’ health care workers. DESIGN, SETTING, AND PARTICIPANTS This survey study constitutes a secondary analysis of a sample of 1916 family and unpaid caregivers to 1203 community-living older adults with activity limitations who participated in the 2017 National Health and Aging Trends Study. Data analysis was performed January to August 2019. EXPOSURES Caregiver sociodemographic characteristics, caregiving intensity, and frequency speaking with or emailing older adults’ health care workers. MAIN OUTCOMES AND MEASURES Caregiver-reported experiences when interacting with older adults’ health care workers in the prior year, including being listened to, being asked about understanding of treatments, and being asked about help needed in managing older adults’ care. RESULTS Caregivers (mean [SE] age, 59.4 [0.5] years; 63.7% women) assisting community-living older adults with activity limitations reported that they never (56.3%), sometimes or rarely (33.0%), or often (10.7%) spoke with or emailed older adults’ health care workers in the prior year. Most caregivers who interacted with older adults’ health care workers reported being always (70.6%) or usually (18.2%) listened to and always (54.4%) or usually (17.7%) being asked about their understanding of older adults’ treatments. Fewer caregivers reported being always (21.3%) or usually (6.9%) asked whether they needed help managing older adults’ care, and nearly one-half (45.0%) were never asked. Caregivers who interacted with older adults’ health care workers often (vs sometimes or rarely) were more likely to report being always or usually listened to (94.8% vs 86.9%; P = .004), being asked about understanding treatments (80.1% vs 69.5%; P = .02), and being asked about needing help (40.8% vs 24.1%; P < .001). No other exposures were consistently associated with caregiver experiences. Measures of caregiving intensity, including caring for an older adult with dementia, were not associated with being listened to or asked about understanding, but were associated with being asked about needed help. Although caregivers of persons with dementia were more likely than caregivers of persons without dementia to report always being asked about needed help (26.9% vs 19.0%), a high percentage in both groups were never asked (41.2% vs 46.5%) (P = .007). CONCLUSIONS AND RELEVANCE These findings reinforce the need for strategies to better support family and unpaid caregivers, who are the main source of assistance to older adults with physical and/or cognitive limitations. JAMA Network Open. 2020;3(1):e1919866. doi:10.1001/jamanetworkopen.2019.19866 Key Points Question What are family and unpaid caregivers’ experiences with health care workers in the care of older adults with activity limitations? Findings In this national survey study, most caregivers reported that older adults’ health care workers always (70.6%) or usually (18.2%) listened to them and always (54.4%) or usually (17.7%) asked about their understanding of the older adult’s treatments, but fewer caregivers reported being always (21.3%) or usually (6.9%) asked whether they need help managing older adults’ care. Meaning These findings reinforce the need for health system strategies to support family and unpaid caregivers, who are the main source of assistance to older adults with physical and/or cognitive limitations. Author affiliations and article information are listed at the end of this article. Open Access. This is an open access article distributed under the terms of the CC-BY License. JAMA Network Open. 2020;3(1):e1919866. doi:10.1001/jamanetworkopen.2019.19866 (Reprinted) January 24, 2020 1/13 Downloaded From: https://jamanetwork.com/ on 01/27/2020


Introduction
Family and other unpaid caregivers provide most of the assistance to community-living older adults with disability. 1 In addition to helping with household and self-care activities, nearly one-half of caregivers assist with health care activities, such as managing medications, coordinating care, or attending medical encounters. 2 Although caregivers often participate in the exchange of patients' health information and in medical decision-making, [3][4][5] supporting their involvement in health care interactions is not straightforward. Patients vary in their preferences for communication assistance, 6,7 and caregivers vary in their knowledge of patient health conditions and priorities. 8,9 Interacting with caregivers poses challenges for health care workers (eg, physicians, nurses, other clinicians, and social workers), who are responsible for ensuring patient privacy and promoting patient autonomy but are not reimbursed for additional time spent educating or counseling family and unpaid caregivers or nurturing productive partnerships. [10][11][12] Available evidence-largely anecdotal and setting specific-has described interactions between family and unpaid caregivers and health care workers as being tense or adversarial. 11,13 Although the communication challenges of surrogate decision-makers are well documented, [14][15][16][17] little is known about the experiences of family and unpaid caregivers during routine interactions with older adults' health care workers. Having a better understanding of caregivers' experiences during interactions with older adults' health care workers is especially timely given the growing interest in the contribution of social risk to outcomes of care. 18 Recent changes to Medicare conditions of participation and reimbursement models that set forth an expectation and the possibility of more explicit caregiver support heighten the importance of better understanding critical elements of person-centered and family-centered care. 19,20 This study draws on 2 linked nationally representative, population-based studies 21,22 that are uniquely able to provide insight regarding the experiences of family and unpaid caregivers to a wellcharacterized sample of community-living older adults with physical or cognitive impairment. Our study has 2 broad objectives. First, we assess the frequency and nature of family and unpaid caregivers' interactions with health care workers, including whether these interactions are associated with characteristics of older adults, family and unpaid caregivers, or caregiving circumstances and appraisal. In doing so, we examine caregivers' perceptions of aspects of communication that have been identified as important in person-centered and family-centered care, including being listened to, being asked about understanding of treatments, and being asked about the need for help in managing older adults' care. [23][24][25] Second, we comparatively examine family caregivers' experiences by whether they assist an older adult with or without dementia, recognizing that family caregivers' interactions with health care workers are likely to be both more frequent and important in the context of impaired memory and judgment. 26 21 and the National Study of Caregiving (NSOC). 22 The initial NHATS sample was drawn from the Medicare enrollment file in 2011 and replenished in 2015; written informed consent was obtained from participants. 21 In 2017, the study was conducted with surviving beneficiaries aged 67 years and older. In-person interviews were conducted with NHATS study participants or with proxy respondents if the participant was unable to respond. Study participants were asked whether and how they performed daily activities in the month before the interview. Among those receiving assistance, a detailed roster was created that lists the relationship of each helper and the specific activities with which they provided help.
The NSOC 22 is a nationally representative survey of family and other unpaid caregivers to older persons with activity limitations. Participants in the NSOC were eligible if they were family or other unpaid helpers to NHATS study participants receiving help with mobility, self-care, or household activities for health and functioning reasons or living in a residential care facility with supportive services. Upon obtaining oral consent, a telephone interview was conducted with up to 5 eligible helpers for each older adult. For older adults with more than 5 eligible helpers, 5 helpers were selected at random for interview.
Of 6312 older adults who participated in the 2017 NHATS, 2214 were included in the NSOC cross-sectional sampling frame, and 4676 of their helpers were eligible for the NSOC. Nonresponse to the NSOC can arise from the NHATS participant (who may refuse to provide contact information for helpers) or caregivers (who may refuse to participate). Participants in the NHATS did not provide contact information for 359 eligible family caregivers, and 1665 of the remaining 4317 eligible family caregivers could not be located or refused to respond. In total, 2652 family caregivers to 1697 older adults responded to the NSOC, 22 yielding 92.4% and 61.9% first-stage and second-stage response rates, respectively. We excluded caregivers to 291 older adults who died and caregivers who were helping 323 older adults living in residential care or 101 adults living in nursing facilities because the nature of caregiving is likely to vary according to availability of services.

Measures
Our main exposure measure is the frequency of caregivers' interactions with older adults' health care workers. Caregivers were asked to respond yes or no to the question "In the last year, did you ever speak to or email any of [older adult's] medical providers about [his/her] care?" Those responding "no" were categorized as never interacting with older adults' health care workers. Those responding "yes" were asked, "In the last year, how often did you speak to or email [older adult's] medical providers about [his/her] care (often, sometimes, rarely)?" We grouped caregivers reporting they interacted with health care workers sometimes or rarely together and examined mutually exclusive categories of never, sometimes or rarely, and often.
Measures of caregiver-reported experiences when interacting with older adults' health care workers were derived from 3 questions that were fielded to caregivers who reported interacting with older adults' health care workers in the last year. These caregivers were told to think about the medical professional with whom they communicated with most often and were then asked, "In the last year, how often did that provider listen to what you had to say, ask if you understood [older adult's] health treatments, and ask if you needed help managing [older adult's] health treatments?" For these 3 dimensions of communication, we contrast caregivers responding "always" or "usually" with caregivers responding "sometimes" or "never." Additional measures included caregivers' sociodemographic characteristics, caregiving intensity, and caregiving-related appraisal. Caregivers' sociodemographic characteristics included age, gender, educational attainment, and relationship to the older adult (spouse, adult child, or other). Caregiving intensity included hours of care provided in the previous week, caregiving for 4 or more years, and types of assistance. We examined provision of help with types of assistance that have special relevance for interactions with health care workers, including medical tasks (tracking medications, administering injections, providing ostomy care or intravenous line care, and providing skin care, such as for wounds or sores) and health system logistics of care coordination.
We examined 2 measures of older adults' physical and cognitive functioning, which have distinctive outcomes on the need for care: self-care and mobility limitations and dementia. We recognize that self-care and mobility limitations may impose the need for transportation or logistical assistance to access medical care, whereas the outcomes of dementia on memory, understanding, and reasoning may impose the need for help in the exchange of medical information and decisionmaking. We constructed a composite measure of self-care or mobility disability that reflects whether the older adult received help in the prior month with 1 or more self-care (eg, eating, dressing, bathing, and toileting) or mobility (eg, getting outside, getting around inside, and getting out of bed) activities. We used a composite measure of probable dementia constructed from self-reported physician diagnosis of Alzheimer disease or dementia, the AD8 dementia screening interview (administered to proxy respondents), 28 and cognitive tests to evaluate memory, orientation, and executive function. 29 Caregiving appraisal refers to a composite measure of caregiving strain (range, 0-9), constructed from 6 items that encompass appraised emotional, physical, and financial difficulty; having no time for oneself; being overwhelmed; and exhaustion. To assess caregiving-related difficulty, caregivers were first asked "Is helping emotionally/physically/financially difficult?" Those responding "yes" were then asked to rate the difficulty of helping in each domain on a scale from 1 (a little difficult) to 5 (very difficult). These items were scored as follows: no difficulty, 0 points; 1 to 3 (some difficulty), 1 point; and 4 to 5 (a lot of difficulty), 2 points. Affirmative responses to questions about having no time for oneself, being overwhelmed, and being exhausted were coded as 1, and negative responses were coded as 0. The composite measure has previously been found to have clinical relevance. 30,31

Statistical Analysis
We first assessed the frequency with which family and unpaid caregivers interacted with older adults' health care workers in the prior year. We estimated the number and characteristics of family and unpaid caregivers reporting they never, sometimes or rarely, or often interacted with older adults' health care workers. We then described family caregivers' experiences during interactions with older adults' health care workers. We compared family and unpaid caregiver sociodemographic characteristics and their caregiving intensity and appraisal by whether they reported that older adults' health care workers always or usually listen to what they have to say, ask about understanding of older adults' treatments, and ask about help needed by caregivers to manage older adults' treatments, in comparison with those reporting that such interactions occurred sometimes or never.
Finally, because family caregivers' interactions with health care workers are particularly important in the context of impaired memory and judgment, 26,27 and the care needs of persons with dementia are particularly demanding, 32 we comparatively examined family caregivers' experiences by whether they assist an older adult with or without dementia.
All analyses were conducted in SAS statistical software version 9.4 (SAS Institute) and Stata statistical software version 14 (StataCorp). Between-group differences were examined from P values associated with the Rao-Scott χ 2 test for categorical measures and the adjusted Wald test for continuous measures, with 2-sided P < .05 considered statistically significant. We weighted our analytic sample to account for caregivers having different probabilities of selection and different probabilities of responding to the NSOC. To make accurate statements about the variance of those estimates, we applied design variables to account for NSOC's complex sample design. Additional details about NSOC survey weights and sampling procedures have been published elsewhere. 33,34 This secondary analysis was conducted between January and August 2019. sometimes or rarely; P = .02) and more likely to be female (81.0% vs 59.2% for never and 65.9% for sometimes or rarely; P < .001), better educated (education beyond high school, 74.2% vs 53.0% for never and 69.6% for sometimes or rarely; P < .001), an adult child (65.7% vs 35.0% for never and 45.2% for sometimes or rarely; P < .001), and to have been providing care for 4 or more years (66.4% vs 55.7% for never and 64.2% for sometimes or rarely; P = .01). Caregivers who interacted with health care workers often were providing care of greater intensity as measured by providing care 20 or more hours per week (44.7% vs 31.0% for those interacting sometimes or rarely and 23.3% for those with no contact; P < .001) and to be assisting with medically oriented tasks and care coordination. Caregivers who interacted with health care workers often were more likely to be caring for an older adult with dementia (36.6% of those interacting often vs 25.9% and 22.5% interacting sometimes or rarely and never, respectively; P = .001) and to report moderate or high caregivingrelated strain (48.7% of those interacting often compared with those interacting with older adults' health care workers less frequently vs 31.0% and 20.7% interacting sometimes or rarely and never, respectively; P < .001).

Results
Family and unpaid caregivers' experiences with older adults' health care workers varied markedly by type of communication support (Figure 1). Most caregivers reported that older adults' We compared the experiences of caregivers for persons with and without dementia during interactions with health care workers (Figure 2). Caregivers for persons with dementia were no more likely than caregivers for persons without dementia to report being listened to and being asked about understanding of treatments by the older adults' health care workers, but were more likely to report they were always asked about needed help managing older adults' treatments (26.9% vs 19.0%) and were less likely to report being never asked (41.2% vs 46.5%) (P = .007).

Discussion
To our knowledge, this study provides the first national information about the experiences of family and unpaid caregivers during routine interactions with older adults' health care workers. We found   health care workers about whether they needed help in managing older adults' treatments, nearly one-half (45.0%) were never asked. Our finding that caregivers providing more hours of care to older adults in worse health were more likely to be asked about needed help suggests that health care workers generally recognized that higher-intensity caregivers may especially need additional support. Although consensus guidelines regarding the circumstances under which an assessment d Composite measure of strain from 6 items that assess physical, emotional, and financial difficulty, exhaustion, role overload, and lack of time for oneself due to caregiving (range, 0-9), with a score of 0 or 1 denoting little or none (570 participants) and a score of 2 to 9 denoting some or high strain (343 participants). Given the important consequences of interpersonal communication between patients and clinicians in the delivery of high-quality care, appropriate services use, and health outcomes, 50 it is encouraging that elements of information exchange were rated especially highly in the subgroup of caregivers with particularly demanding responsibilities who interacted with older adults' health care professionals most often. Although family caregivers commonly accompany older adults to medical encounters, 3,4 actively participate in face-to-face discussion, 3 and are influential in adherence and care coordination, 1,3,51 surprisingly little is known about the frequency, nature, and consequences of these interactions. There is a growing appreciation that understanding what people view to be important is foundational to the measurement and delivery of person-centered and family-centered care. 23 Findings from this study and others 7,25 suggest a benefit to assessing family perspectives in the measurement of care quality. Because family caregivers are often heavily involved in the care of persons with diminished capacity to self-report on experiences with care, 2,32 incorporating family caregiver-reported information may be especially important to ensuring that assessments of care quality incorporate all perspectives. Newly developing methods and instruments are encouraging in this regard. 25,52

Strengths and Limitations
Several strengths and limitations of our study merit comment. Strengths of our study include drawing on nationally representative population-based data for a well-characterized sample of older adults and their family and unpaid caregivers. Although our study appropriately draws on caregiver-reported measures when reflecting on their experiences of care, self-reported information is subject to potential misclassification for measures such as the frequency of interactions. As a secondary analysis of a national survey, our study was constrained to available information regarding interactions with medical professionals, and we are unable to comment on factors such as the type or specialty of health care workers, their gender, their years in practice, the duration of their relationship with the older adult, and their communication style, all of which are known to be associated with interpersonal rapport and trust. 50,53 We are unable to differentiate caregivers' expectations or preferences for actively engaging with health care workers, or how patient expectations or preferences about the involvement of their caregiver are associated with these interactions. Although we examined specific elements of interpersonal communication, as opposed to a psychometrically validated multi-item instrument, the measures we assess are directly pertinent to aspects of care that have been identified as important to patients and families, 25,54 as well as clinicians and policy makers who seek to improve care for those with chronic and disabling conditions. 39

Conclusions
Well-coordinated, team-based care is an important element in the delivery of care that is safe, efficient, and high quality. 55 Our study affords a novel perspective on such care by contributing insight regarding the experiences of family and unpaid caregivers when interacting with older adults' health care workers. The high percentage of caregivers who reported being well supported when communicating with health care workers about older adults is indicative of their visible and better recognized role in maintaining the health and well-being of older adults with physical and/or cognitive limitations; that caregivers were less likely to report being asked about needed help reflects the greater ambiguity in clinician responsibility toward supporting family caregivers' needs. These findings are directly relevant to the movement to recognize family as contributing members of an interdisciplinary care team 35,56 and contribute benchmark information that may helpful in guiding and monitoring progress toward the advancement of person-centered and family-centered care. 23