A survey exploring caregiver burden and health-related quality of life in hereditary transthyretin amyloidosis

Background Hereditary transthyretin amyloidosis (ATTRv) is an ultra-rare, life-shortening disease with a high unmet need. This study examined ATTRv caregiver health-related quality of life (HRQoL) and productivity. Methods A cross-sectional online survey, including EQ-5D-3L, Hospital anxiety and depression scale (HADS), and caregiver and patient characteristics questions, was developed to assess ATTRv caregiver burden. A companion general population survey collected EQ-5D-3L, HADS and chronic health conditions data. Caregiver-control group differences in HRQoL were assessed using t-tests and chi-square tests. Ordinary Least Squares regression was used to estimate the disutility of being a caregiver compared to controls stratified by patient ambulatory status. Results Thirty-six caregivers and matched controls completed the survey (n = 72). The disease severity of patients they cared for was varied: 33% required no assistance walking, 58% required assistance with walking and 9% required a wheelchair/were bedridden. On average, caregivers spent 6 h daily on practical care and 4 h daily on emotional support. Fifty-six percent indicated that they had changed their employment due to providing ATTRv care. Caregivers reported lower HRQoL, as indicated by lower EQ-5D 3L utility scores (M = 0.772, SD = 0.178 vs. M = 0.849, SD = 0.218) and higher HADS anxiety (9.3 vs. 6.1, p < 0.01) and depression (7.6 vs. 4.4, p < 0.01) scores, compared with matched controls. Caregivers were also more likely to report sleep problems (33% vs. 8%, p < 0.01) and stress (42% vs. 0%, p < 0.001) as chronic conditions than controls. Conclusions The study results indicate that caring for a person with ATTRv can have a considerable negative impact on caregivers’ HRQoL and productivity. The study findings provide important information for economic evaluations of ATTRv treatments.


Background
Hereditary transthyretin amyloidosis (ATTRv) is an ultra-rare genetic disorder characterised by deposits of misfolded transthyretin (TTR) protein in the body's organs and tissues. Build-up of TTR protein deposits lead to a diverse clinical presentation of neurological and/or cardiac symptoms with varying degrees of progressive disease, typically resulting in premature death. Neurological symptoms include peripheral neuropathy (sensory abnormalities, motor weakness, loss of ambulation) and autonomic dysfunction (severe gastrointestinal [GI] symptoms, cardiac arrythmias, orthostatic hypotension, bladder dysfunction), with malnutrition and wasting causing death. Cardiac complications include cardiomyopathy (typically thickening of the ventricular walls) resulting in heart failure. Recent qualitative and quantitative research has described the profound impact of ATTRv on patient's health-related quality of life and the considerable caregiver burden of ATTRv [1][2][3]. Relatedly, previous patient and caregiver surveys have also demonstrated that the caregiver burden of ATTRv was also high [4]. One study found a high number of hours spent on care. Furthermore, caregiving was associated with a considerable impact on work productivity as well as poor physical and mental health. Policy makers and health technology appraisal (HTA) agencies, such as the National Institute for Health and Care Excellence (NICE) in the UK, have recognised the need to consider caregiver burden, where relevant, when reviewing cost effectiveness of new treatments [5]. To quantify treatment benefits for caregivers in cost-effectiveness analyses, preference-based measures, such as the EQ-5D, are required to calculate utility (or disutility) values associated with caregiving (e.g. Acaster et al. [6]; Landfelt et al. [7]). As an example of the growing recognition of the importance of caregiver burden, reviews of the use of caregiver utilities in NICE technology appraisals (TAs) show an increase over time from zero in 2005 [8], rising to two in 2008 [9], six in 2012 [10] and 16 as of January 2019 [11]. The most recent review of NICE appraisals also highlights the growing importance of the application of carer utilities as it was commissioned by NICE to review and critique the methods being applied, shortly before they announced their HTA evaluation methods review.
To date, no caregiver (dis)utility data have been reported for ATTRv. The caregiver burden has also not been isolated from that of being a patient, as a high proportion (40%) of caregivers in the Stewart et al. [4] study were also diagnosed with ATTRv themselves. To obtain robust caregiver utilities and broader caregiver burden data, the caregiver impact needs to be disentangled from that of being a patient. Estimating the burden associated with caregiving for an ATTRv patient would also benefit from a matched population norm comparator to account for demographic variations in health unrelated to caregiving. Finally, the impact of patient disease severity on caregiver burden remains unexplored in ATTRv. Pennington and Wong's [11] review of NICE appraisals emphasised the need to explore caregiver health-related quality of life (HRQoL) changes over time, including when the patient's health improves or worsens, to increase accuracy of health economic models. Thus, understanding how disease severity impacts caregiver burden is particularly relevant in the context of a progressive disease such as ATTRv, with patients able to walk unassisted at disease onset but needing a wheelchair or being bedridden in the final stages of disease.
The present study aimed to extend and refine the caregiver impacts reported by Stewart et al. (2018) to provide a more robust and detailed understanding of the ATTRv caregiver burden, including caregiver utility data suitable for inclusion in economic analyses. More specifically, this study aimed to capture caregiver burden using validated measures of health-related quality of life (EQ-5D), anxiety and depression (Hospital anxiety and depression scale) and bespoke assessments of time spent providing care, career impacts and chronic health conditions. Caregiver health impacts were estimated by including a matched general population sample and isolating caregiver impact from any patient impact by accounting for the ATTRv status of caregivers. Finally, the impact of ATTRv disease severity, as classified by Coutinho stages of ambulatory status [12], on caregiver burden was also explored.

Study design
A cross-sectional online survey was developed to assess the caregiver burden of ATTRv. The survey was reviewed by Amyloidosis Research Consortium (ARC) UK and piloted using feedback from one caregiver in the ARC UK network. Pilot data were included in the analysis. A companion general population survey was designed to collect data to compare with the caregiver data where relevant.

Sample and participant recruitment
Thirty-six ATTRv English-speaking caregivers were recruited via ARC UK and a healthcare panel agency in the UK, US, Canada, Australia and New Zealand. Caregiver inclusion criteria were: (i) currently providing regular (at least weekly) informal (unpaid) care for someone with ATTRv; (ii) residing in the UK, US, Canada, Australia, or New Zealand; and (iii) being aged eighteen or over. Data was collected across a number of Englishspeaking countries to achieve the largest possible sample across populations that are comparable. Country differences in the caregiver data were not examined, as the small sample size precluded meaningful group comparisons between countries.
A matched general population sample (n = 36) was recruited in the UK via the same healthcare panel agency. The general population sample inclusion criteria were: (i) not providing informal (unpaid) care for someone with a disease or disorder on a regular (at least weekly) basis; (ii) residing in the UK; and iii) being aged eighteen or over. The general population sample was matched with the caregiver sample on the group-level in terms of gender, age group (18-24/25-34/35-44/45-54/55-64/65 or over), living situation (living with partner or spouse / other living arrangements) employment status (employed full-time/employed part-time/unable to work due to own health/unemployed or homemaker or retired or other).

Measures
The survey included items related to caregiver burden, caregiver proxy-reported patient characteristics, socio-demographics and chronic health conditions. The survey also included two validated questionnaires: the EQ-5D-3L and the Hospital anxiety and depression scale (HADS). Study-specific measures and the validated questionnaires are described below.

Caregiver burden and characteristics
The caregiver sample completed caregiver burden measures including (i) hours of informal care, broken down by time spent on practical care and emotional support; and (ii) employment impact, defined as change in employment status due to ATTRv care (stopped working/ reduced hours/changed jobs/other, please specify).
Caregivers were also asked about (iii) the number of ATTRv patients they were caring for; (iv) the relationship between the caregiver and the ATTRv patient(s); and (vi) how long they had been a caregiver for.

Patient characteristics
Caregivers were asked to proxy-report patient characteristics including (i) time since diagnosis; (ii) type of ATTRv (cardiomyopathy/polyneuropathy/both/other or not sure); (iii) V-30 M status (yes/no/not sure); (iv) ambulatory status using Coutinho staging (no assistance with walking/assistance walking/wheelchair/bedridden); (v) symptoms, including a list of 7 neuropathic, 2 cardiac, 4 gastrointestinal and 8 other symptoms; (vi) patient demographics.

Socio-demographic and chronic conditions
All (i.e. caregiver, caregiver-patient and general population) participants completed background questions on (i) socio-demographics; and (ii) chronic health conditions, which included a list of common chronic conditions (arthritis, anxiety, cancer, chronic kidney disease, depression, diabetes, digestive pain/discomfort, heart disease, hypertension [high blood pressure], respiratory condition [e.g. asthma, COPD], sleep problems, stress) and an open-ended 'other' response option.

EQ-5D-3L
All (i.e. caregiver, caregiver-patient and general population) participants completed the EQ-5D-3L questionnaire. The EQ-5D-3L is a generic health status measure [13]. Participants indicate their current health status on five domains (mobility; self-care; usual activity; pain/ discomfort; anxiety/depression) as either no problems, some problems or severe/extreme problems. Participants also indicate their current health on a visual analogue scale. Health utilities were derived from the EQ-5D using UK general population preference weights [14], which provide a potential range of scores from − 0.59 to 1.0 (where a score of 1 represents full health and a score of 0 represents dead).

HADS
All (i.e. caregiver, caregiver-patient and general population) participants completed the HADS questionnaire. The HADS is a valid and reliable self-rating scale that measures anxiety and depression in both hospital and community settings [15]. The questionnaire gives clinically meaningful results as a psychological screening tool and can assess the symptom severity and "caseness" of anxiety disorders and depression in patients with illness and the general population. The HADS consists of fourteen items; seven questions for anxiety and seven for depression. The responses being scored on a scale of 0-3 (3 indicating higher symptom frequencies). Scores for each subscale (anxiety and depression) range from 0 to 21 with scores categorised as normal (0-7), mild (8-10), moderate (11)(12)(13)(14) and severe (15)(16)(17)(18)(19)(20)(21). Moderate and severe scores are equated with 'probable clinical caseness'; UK population norms have been reported for such levels of anxiety and depression [16].

Statistical analysis
General population and caregiver socio-demographics, caregiver burden, caregiver characteristics, and patient characteristics were tabulated using descriptive statistics.
Caregiver burden and characteristics were also analysed by patient ambulatory status (Coutinho stages, with 'requires wheelchair' and 'bedridden' collapsed due to small cell sizes) using independent t-tests and chi-square tests as appropriate. For caregivers who cared for multiple ATTRv patients (n = 2), the worst (more severe), or if equal severity the first, reported patient's ambulatory status was used in all analyses.
Caregiver HADS anxiety and depression scores were compared with matched general population scores and by patient ambulatory status using independent t-tests. Group differences in the proportion of moderate/severe anxiety and depression ('probable clinical cases'; score range:11-21) were tested for using chi-square tests.
Frequencies and distributions of the EQ-5D profile scores were described using bar charts for the general population and caregiver samples. Utility scores were calculated from the EQ-5D 3L data based on the UK valuation weights [14]. Ordinary least squares regression was used to estimate the disutility associated with being a caregiver, stratified by Coutinho stage, versus the matched general population sample. Due to small sample sizes the 'requires wheelchair' and 'bedridden' stages were collapsed, giving four groups in total: matched control (general population), requires no assistance walking, requires assistance walking, and requires wheelchair/ bedridden, with controls as the reference category. The model was run with and without socio-demographic variables as covariates, as the caregivers and controls were only matched at the overall group level. The caregiver sample included only one patient-caregiver (n = 1), i.e. a caregiver who was also an ATTRv patient, who was the only participant unable to work due to ill health in the caregiver sample. All descriptive statistics are provided for the overall caregiver sample, with patient-caregiver responses described in table footnotes where relevant. All statistical tests were run with and without the patient-caregiver and one general population matched control respondent (n = 1) who, like the patientcaregiver, was unable to work due to ill health. Only test results using the overall sample were reported, unless results with and without patient-caregiver and matched control differed.
All statistical tests were two-sided and p-values below 0.05 were considered statistically significant.

Sample characteristics
A total of 36 English-speaking caregivers and 36 matched controls completed the online survey. Of the total caregiver sample, 6 were recruited via ARC-UK and 30 via a healthcare panel agency. One participant dropped out after completing the informed consent form, resulting in a dropout rate of 3%. No general population participants dropped out. Of the 36 caregivers, 23 (64%) resided in the United States, 8 (22%) in the United Kingdom, 3 (8%) in Canada, 1 (3%) in Australia and 1 (3%) in New Zealand.
Over three-quarters (78%) of caregivers in the sample were caring for a partner or spouse, with the remainder caring for a sibling (14%), parent (3%) or child (3%). Only one (3%) caregiver had indicated she had ATTRv herself.
Caregiver and general population sample characteristics are described in Table 1. Most caregivers were female (81%) and had a mean age of 56.3 years (SD = 12.5). The vast majority of caregivers were living with a partner or spouse (94%). The caregiver sample was varied in terms of employment status, with 33% retired, 28% part-time employed and only 14% full-time employed. As expected due to the sampling methodology, the caregiver and general population sample characteristics were comparable in terms of age, gender, living situation and employment status.
Compared with the general population sample, caregivers were more likely to report sleep problems (caregivers: 33% vs. gen. pop.: 8%, χ 2 = 6.8, p < 0.01) and stress (caregivers: 42% vs. gen. pop.: 0%, χ 2 = 19.0, p < 0.001) as chronic health conditions. There were no statistically significant differences between caregiver and general population samples in the proportion who reported anxiety, depression and common physical health diseases as chronic health conditions.

Patient characteristics
Patient characteristics were proxy-reported by the caregiver sample and are shown in Table 1. The ATTRv patients the caregiver sample cared for were on average diagnosed 5.8 years (SD = 4.8) ago. Most were male (78%) and had a mean age of 59.0 (SD = 13.4). Only 17% of patients were part-time or full-time employed. Onethird (33%) of patients could walk unassisted, over half (58%) required assistance walking, 6% required a wheelchair and 3% were bedridden. Over half (56%) reported that the patient had cardiomyopathy and polyneuropathy, around one-third (31%) had reported polyneuropathy and 6% had reported cardiomyopathy. Approximately one-fifth (22%) reported a V-30 M variant. When asked to report specific symptoms the patients experienced, all (100%) reported one or more neuropathic symptom(s) and most also reported gastrointestinal symptom(s) (83%) and cardiac symptom(s) (72%).

Caregiver burden Informal care and impact on caregiver employment
Overall, caregivers spent an average of 5.8 h on practical care (median = 3.0; SD = 6.2) and 4.1 h on emotional support (median = 2.0; SD = 6.1) on a daily basis ( Table 2). Hours spent on practical care was significantly associated with patient ambulatory status, with those with more severe impairment requiring more practical support (p < 0.05). In contrast, hours spent on emotional support was not associated with patient ambulatory status.
The majority of caregivers were not employed, with only 42% in either full-time or part-time employment. Over half (56%) of caregivers had indicated that they had changed their employment as a result of ATTRv care; Fig. 1 shows a breakdown of caregiver change in employment. Caregiver change in employment was significantly associated with patient ambulatory status, with a higher proportion of caregivers who supported patients requiring assistance with walking reporting a change in employment compared with caregivers of patients who did not require assistance walking (71% vs. 42%, χ 2 = 6.83, p = 0.03).

Impact on caregiver anxiety and depression
As illustrated in Fig. 2, caregivers had significantly higher HADS anxiety scores than the matched general population (M = 9.3 vs. M = 6.1, t = -3.25, p < 0.01). Furthermore, a significantly higher proportion of caregivers had HADS anxiety scores qualifying as 'probable clinical anxiety' compared with the matched general population, which was comparable to the UK population norm data (36% vs. 14%, χ 2 = 4.74, p = 0.03). Within the caregiver sample, patient ambulatory status was not significantly associated with caregiver HADS anxiety scores or the proportion with 'probable clinical anxiety' .
Caregivers also had significantly higher depression scores than the matched general population (M = 7.6 vs. M = 4.4, t = − 3.39, p < 0.01; Fig. 2). While the proportion of caregivers with HADS depression scores qualifying as 'probable clinical depression' was higher than that in the general population matched controls, these group differences were not significantly different (22% vs. 11%, χ 2 = 1.6, p = 0.21). Among caregivers, patient ambulatory status was not significantly associated with caregiver HADS depression scores or the proportion with 'probable clinical depression' . Table 1 Caregiver sample characteristics and proxy-reported patient characteristics a Characteristics of the patient-caregiver aged 38 (n=1) and the patient aged 46 who the patient-caregiver was caring for b Characteristics of the matched control aged 38 (n=1) *Requires no assistance walking, requires assistance with walking and requires wheelchair correspond to Coutinho stages 1, 2 and 3 respectively Living situation: with relative 2 (5%) a 0 (0%)

Caregiver health-related quality of life
Overall, caregivers reported lower HRQoL than sociodemographically matched members of the general population as measured using the EQ-5D-3L health utility values. Utilities values for the caregiver and general population groups and by patient ambulatory status for caregivers are shown in Table 3. All utility and estimated disutility values are given with and without data from the patient-caregiver and corresponding matched control. For the overall sample, the average caregiver utility score was 0.772 (SD = 0.178) compared with a matched general population utility of 0.849 (SD = 0.218). Excluding the patient-caregiver and corresponding matched control, average caregiver utility was 0.778 (SD = 0.219) compared with a general population utility of 0.871 (SD = 0.122). The caregiver-general population difference in utilities was statistically significant when excluding the patient-caregiver (b = − 0.094, 95% CI − 0.178-− 0.009, t = − 2.21, p = 0.03) but was not significant for the total sample. Further examination of the EQ-5D-3L profiles suggests that lower caregiver utility values were predominantly driven by lower caregiver scores on the EQ-5D-3L anxiety/depression domain ( χ 2 = 11.9, p < 0.01; Fig. 3).
Estimated disutility values associated with providing care stratified by ambulatory status, using the general population as a reference category, are also provided in Table 3. As the general population were matched at the group level (rather than at the individual caregiver level), the regression model was repeated including sociodemographic covariates.
In line with the EQ-5D-3L utility values, caregivers also had lower EQ-5D VAS scores than the matched general population, although this did not reach statistical significance (

Discussion and conclusions
The mental health impact of providing care was considerable. Anxiety levels as measured by the HADS showed that caregivers had significantly higher anxiety scores, reporting two and a half times higher levels of 'probable clinical anxiety' than matched controls. Notably, the levels of anxiety among controls in this sample were in line with previously published UK population norms for the HADS [16]. Caregivers also reported significantly higher HADS depression scores than controls. Caregivers were twice as likely to report 'probable clinical depression' , although this was not statistically significant. In addition to higher levels of HADS anxiety and depression scores, caregivers were also significantly more likely to report sleep problems and stress as chronic health conditions than the matched general population sample. These impacts were reflected in lower HRQoL, as indicated by lower EQ-5D-3L utility values among caregivers Table 3 EQ-5D-3L Utility and Disutility: general population, caregivers and caregivers by ambulatory status a Sample excluding patient-caregiver (n=1) and corresponding matched control (n=1). b The caregiver utility value for the bedridden patient is that of the patient-caregiver (n=1). c Estimated using Ordinary Least Squares (OLS) regression *p<0.05