Support Care Cancer DOI 10.1007/s00520-016-3120-7
ORIGINAL ARTICLE
Psychological distress, health, and socio-economic factors in caregivers of terminally ill patients: a nationwide population-based cohort study Mette Kjaergaard Nielsen 1 & Mette Asbjoern Neergaard 2 & Anders Bonde Jensen 3 & Flemming Bro 1 & Mai-Britt Guldin 1,2
Received: 10 September 2015 / Accepted: 9 February 2016 # Springer-Verlag Berlin Heidelberg 2016
Abstract Purpose At some point in life, most people become caregivers to a terminally ill relative. Previous studies have shown that many caregivers experience psychological distress and declining physical health, but these studies have predominantly been conducted in specialized palliative care settings. Therefore, caregiver studies with a population-based approach are needed. We aimed to describe socio-economic characteristics, situational factors, pre-loss grief symptoms, depressive symptoms, caregiver burden, and health status in a general population of caregivers to terminally ill patients. Method We conducted a nationwide population-based cohort study. Caregivers were systematically recruited through patients registered with drug reimbursement for terminal illness in 2012. Data on socio-economic characteristics was mainly obtained from Danish registries, whereas data on situational factors, distress, and health was measured in questionnaires. Results Of patients to responding caregivers (n = 3635), 89 % suffered from cancer, predominantly lung cancer (23 %). Of responding caregivers, 62 % were partners and 29 % were adult children. In total, one third of caregivers reported severe outcome, 15 % reported severe pre-loss grief symptoms, 16.1 % had moderate to severe depressive symptoms, and
* Mette Kjaergaard Nielsen [email protected]
1
Research Unit for General Practice, Aarhus University, Bartholins Allé 2, 8000 Aarhus C, Denmark
2
The Palliative Team, Department of Oncology, Aarhus University Hospital, Noerrebrogade 44, 8000 Aarhus C, Denmark
3
Department of Oncology, Aarhus University Hospital, Noerrebrogade 44, 8000 Aarhus C, Denmark
12 % experienced high caregiver burden. Partners had the highest levels of pre-loss grief and depressive symptoms, while adult children reported the highest levels of caregiver burden. Conclusions From this cohort, which was estimated to be representative of caregivers to terminally ill relatives in the general population, we found high levels of pre-loss grief, depressive symptoms, and/or caregiver burden in one third of all caregivers. These findings call for increased focus on caregivers’ need of support. Keywords Informal caregivers . Terminal illness . Pre-loss grief . Depression . Caregiverburden . Socio-economicfactors
Background Severe illness in a close relative is a big challenge to face. Cancer, COPD, heart failure, and other severe diseases cause a large number of deaths in modern society and inevitably place relatives in the role as caregivers. The need for informal caregiving is expected to increase due to an aging population in Western countries; fortunately, most caregivers wish to take part in giving care to their close relatives [1, 2]. However, the need for caregiving may increase as the disease progresses [3, 4], and caregivers may be at risk of developing illness themselves [4] if the demands for care exceed their resources [5]. Informal caregivers are at risk of experiencing severe preloss grief symptoms [6, 7]. This implies that the grief symptoms in caregivers can be very severe early in the terminal illness trajectory and affect their daily life functioning. Furthermore, severe pre-loss grief symptoms have been found to strongly predict complicated grief (i.e., severe grief impairing caregiver functioning for more than 6 months after the loss) [6, 7]. The few existing studies on pre-loss grief [6–9]
Support Care Cancer
are mainly based on data from specialized palliative care settings, and severe pre-loss grief was found in about 15 % of caregivers to these patients [8]. Caregiving may cause other adverse outcomes. Former studies have shown that depressive symptoms are present in up to half of caregivers to dying patients [10, 11], that the physical functioning of caregivers is impaired during illness [12, 13], and that caregiver burden is reported in up to 32 % of caregivers [14]. These adverse outcomes in caregivers seem to be associated with situational factors during the patient’s illness trajectory; these include low level of communication between the patient and the caregiver about illness and death [15], perceived low preparedness in the caregiver for the impending death [16], and a high number of caregiving hours [14]. In a recent review, females and young caregivers were found to be at particular risk of developing psychological distress, while inconclusive results were reported regarding the caregiver’s relations with the patient [3]. Large-scale population-based studies of informal caregivers to adult patients with terminal illness are needed [17, 18]. Specifically, we need more knowledge on adverse outcomes and socio-economic characteristics of caregivers, including caregivers to terminally ill patients who do not receive specialized palliative care. Such knowledge may provide a solid foundation for more targeted and effective support to caregivers [18]. Therefore, in a large-scale nationwide caregiver cohort, we aimed to: 1) describe socio-economic characteristics of caregivers and situational factors including number of caregiving hours and level of preparedness, and 2) investigate levels of pre-loss grief symptoms, depressive symptoms, caregiver burden, and health status in caregivers Furthermore, we wanted to investigate differences between caregivers in the abovementioned factors on the basis of the caregiver’s relation to the patient.
Methods Setting The Danish health-care system is tax-funded, and welfare services provided at hospitals and in primary care are free of charge. However, medication expenses are only partly covered, but citizens assessed by a physician as incurably ill are entitled to full drug reimbursement due to terminal illness after formal registration at the Danish Medicines Agency [19]. Furthermore, a relative of the terminally ill patient may take
compassionate leave and has a statutory right to receive a special allowance during such leave [20]. For the present study, we systematically approached caregivers of all terminally ill patients in both a general and a specialized palliative care setting through letters to patients receiving drug reimbursement for terminal illness in Denmark in 2012. Data from caregiver questionnaires and Danish national registers were linked at an individual level using the Danish personal identification number (CPR number). All data were anonymized and accessed at Statistics Denmark through an authorized connection [21] and reported according to the STROBE Statement [22]. Sampling procedure On a weekly basis throughout 2012, the first author received a list from the Danish Health and Medicines Authority of all patients newly registered for drug reimbursement. The patients received a letter within 2 days and were asked to forward an enclosed questionnaire to their closest relative. The letter contained a participant information leaflet and a questionnaire with written consent form. Variables and data sources Age and gender were calculated from CPR numbers of patients and caregivers, and patients were followed-up for survival until 1 June 2014. Questionnaire-based scales and variables A pre-loss version of the Prolonged Grief-13 scale (PG-13) measured grief symptoms presented as a sum score and categorical criteria scores (yes, no) based on the criteria for the PG-13 scale [8]. The 21-item Beck’s Depression Inventory-II (BDI-II) measured depressive symptoms [23]. An additional category (does not apply) was added to item 21 on sexuality as the question was considered offensive in a previous caregiver study [24]. Sum scores and categorizations were handled in accordance with the manual [23, 25]. The 36-item Short Form Health Survey (SF-36) measured health status on a physical component score (PCS) and a mental component score (MCS); these ranged from 0 (worst) to 100 (best) [26–28]. The Burden Scale for Family Caregivers (BSFC) was reported as a sum score and a categorical score (severe, moderate, mild, no symptoms) according to the manual [29, 30]. Ad hoc items addressed the caregiver’s relation to the patient (partner, child, other), the amount of hours spent on both practical and total care (0–2, 3–8, 9–16, 17–24 h/day), and whether the caregiver had compassionate leave (yes, no, not on labor market). Caregiver’s preparedness for death was measured on a single-item question and reported as a dichotomized value (high, low) [31]. For the five-item Couples’
Support Care Cancer
Communication about Illness and Death scale (CCID) [15, 32], a sum score and a dichotomized value (high, low) was reported. Register-based data The socio-economic variables retrieved from Statistics Denmark [33] for both patients and relatives were cohabitation status (married/cohabiting, living alone), educational level (≤10 years, >10 and ≤15 years, >15 years of education), ethnicity (immigrant/descendant, non-immigrant/descendant), and urbanicity (50,000 inhabitants in the community) per 1 January 2012. For caregivers, data were also retrieved on children living at home (yes, no), residency (owned, rented), and gross income in 2011 (50,000 euros/year). Diagnoses for cancer patients were obtained from the ICD-10 codes recorded in the Danish Cancer Registry (DCR) [34]; these were categorized into colorectal, lung, breast, prostate, hematological, and other cancers. Patients without a cancer diagnosis were categorized with a Bnon-cancer^ diagnosis. Data on diagnoses from hospital admissions for both patients and caregivers were retrieved from the Danish National Patient Registry (NPR) [35]; these were used to identify the most frequent noncancer diagnosis among patients and the number of chronic diseases on the Charlson Comorbidity Index (CCI) [36, 37] for caregivers. Statistical analyses Variables were reported using a mean value with 95 % confidence intervals (CI) or a median with interquartile range or proportion, whichever appropriate. Patients of responding and non-responding relatives were compared by t tests or chi square tests to identify any differences in the socioeconomic characteristics between the two patient groups. In accordance with previous studies, continuous summary score variables for BSFC, SF-36, BDI-II, and pre-loss PG-13 were presented as a mean with 95 % CI [8, 23, 30, 38]. Only respondents with complete responses on the CCID, SF-36, BDIII, and pre-loss PG-13 or less than two missing values on the BSFC [29] were included in the analysis. Data were analyzed using the statistical software Stata 13.0 (StataCorp LP, TX).
Results Study population We obtained 11,628 CPR numbers of patients registered with drug reimbursement in 2012, and 9512 eligible patients were contacted (Fig. 1). A total of 3635 caregivers responded positively (38 %), and 3488 caregivers provided their CPR
number for inclusion in the register-based analysis (Fig. 1). The patients of responding caregivers were significantly more likely to be male, be young, have a partner, have a higher education, be non-immigrants, have a cancer diagnosis, and have a longer survival time after receiving drug reimbursement than patients of non-responding caregivers (Table 1). The median time to death of patients of non-participating relatives was 64 days after drug reimbursement and 93 days for patients with a participating caregiver (Table 1). Socio-economic and situational factors The majority of caregivers were partners (62 %), and 29 % were adult children. The vast majority (89 %) were caregivers to cancer patients, predominantly lung cancer (23 %) (Table 2). Most caregivers (66 %) spent more than 3 h per day providing care for the patient, and most caregivers reported high preparedness for the patient’s impending death, although 17 % of all caregivers stated that they were not prepared (Table 3). Pre-loss grief, depressive symptoms, health, and caregiver burden The criteria for severe pre-loss grief symptoms were fulfilled for 15 % of caregivers, and moderate to severe depressive symptoms was found in 16 %. The SF-36 mean PCS sum score was 50.2 (95 % CI 50.4–51.1), and the mean MCS sum score was 41.0 (95 % CI 40.6–41.4) (Table 3). Moderate to very severe caregiver burden was experienced by 12 % of caregivers. In total, 30 % of the caregivers reported symptoms of pre-loss grief, depression, or caregiver burden, and 13 % had more than one of the mentioned conditions (data not shown). Severe pre-loss grief symptoms were reported by 17 % of partners and 11 % of adult children (Table 3). In total, 17 % of partners and 15 % of adult children had moderate to severe depressive symptoms, 11 % of partners reported moderate to very severe caregiver burden, and 14 % of adult children experienced caregiver burden.
Discussion In this nationwide large-scale caregiver cohort, two third of the caregivers was partners of the patient, and 17 % reported low preparedness for the impending death. More than two third of the caregivers spent more than 3 h per day providing care for the patient. The criteria for severe pre-loss grief symptoms were met for 15 % of participants, moderate to severe depressive symptoms were found in 16 %, and high caregiver burden was experienced by 12 %. Adult children reported a higher level of caregiver burden than partners. However,
Support Care Cancer Fig. 1 Flow chart of study population of patients (pt.) granted drug reimbursement and their caregivers
Population source: Patients granted drug reimbursement in 2012 11,628 Pt. dead prior to contact: 1,348 (11.6%) Pt. protected from contact*: 757 (6.5%) Pt. 15 years of education Missing data Gross incomeb 50,000 euros/year Missing data Residency Owned Rented Missing Ethnicity Not immigrant/descendant Immigrant/descendant Missing data Urbanicity 50,000 inhabitants Missing data Number of chronic diseasesc 0 1–2 >2 Missing data
All caregivers
Partners n = 2254 (62.0 %)
Children n = 1037 (28.5 %)
Other relation n = 269 (7.4 %)
61.2 (60.8–61.6)
66.5 (66.1–66.9)
49.9 (49.2–50.5)
58.8 (57.2–60.3)
659 (18.1) 855 (23.5) 1067 (29.4) 907 (25.0) 147 (4.0)
152 (6.8) 377 (16.7) 848 (37.6) 848 (37.6) 29 (1.3)
450 (43.4) 400 (38.6) 134 (12.9) 9 (0.9) 44 (4.2)
57 (21.2) 77 (28.6) 83 (30.9) 47 (17.5) 5 (1.9)
2420 (66.6) 1068 (29.4) 147 (4.0)
1419 (63.0) 806 (35.8) 29 (1.3)
787 (75.9) 206 (19.9) 44 (4.2)
214 (79.5) 50 (18.6) 5 (1.9)
3026 (83.3) 451 (12.4) 158 (4.4)
2128 (94.4) 92 (4.1) 34 (1.5)
731 (70.5) 258 (24.9) 48 (4.6)
163 (60.6) 99 (36.8) 7 (2.6)
717 (19.7) 2760 (75.9) 158 (4.4)
198 (8.8) 2022 (89.7) 34 (1.5)
452 (43.6) 537 (51.8) 48 (4.6)
67 (24.9) 195 (72.5) 7 (2.6)
980 (27.0) 1651 (45.4) 802 (22.1) 202 (5.6)
741 (32.9) 1028 (45.6) 416 (18.5) 69 (3.0)
152 (14.7) 511 (49.3) 320 (30.8) 54 (5.2)
85 (31.6) 111 (41.3) 63 (23.4) 10 (3.7)
921 (25.3) 922 (25.4) 880 (24.2) 760 (20.9) 152 (4.2)
786 (34.9) 610 (27.1) 460 (20.4) 366 (16.2) 32 (1.4)
86 (8.3) 216 (20.8) 347 (33.5) 343 (33.1) 45 (4.3)
48 (17.8) 92 (34.2) 73 (27.2) 50 (18.6) 6 (2.2)
2512 (69.1) 942 (25.9) 181 (5.0)
1644 (72.9) 565 (25.1) 45 (2.0)
690 (66.5) 291 (28.1) 56 (5.4)
175 (65.1) 83 (30.9) 11 (4.1)
3389 (93.2) 88 (2.4) 158 (4.4)
2161 (95.9) 59 (2.6) 34 (1.5)
966 (93.2) 23 (2.2) 48 (4.6)
256 (95.2) 6 (2.2) 7 (2.6)
1391 (38.3) 1041 (28.6) 1038 (28.6) 165 (4.5)
924 (41.0) 662 (29.4) 629 (27.9) 39 (1.7)
369 (35.6) 303 (29.2) 317 (30.6) 48 (4.6)
96 (35.7) 76 (28.3) 88 (32.7) 9 (3.3)
2448 (67.4) 488 (13.4) 91 (2.5) 608 (16.7)
1483 (65.8) 372 (16.5) 72 (3.2) 327 (14.5)
761 (73.4) 85 (8.2) 14 (1.4) 177 (17.1)
200 (74.4) 30 (11.2) 5 (1.9) 34 (12.6)
All variables as of 1 January 2012, unless otherwise stated a
Information on relation to the patient was missing for 75 caregivers (2.1 %); these account for missing values in the relation variable
b
For the year 2011
c
Charlson Comorbidity Index (CCI) based on hospital registered diagnoses during 2001–2011. Therefore, missing values include persons unregistered at the hospital
Support Care Cancer Table 3
Self-reported caregiver data on situational factors, distress and health Relation to patient (n = 3560)a All caregivers (n = 3635)
Partners 2254 (62.0 %)
Adult children 1037(28.5 %)
Other relation 269 (7.4 %)
Situational factors Employment Working
1252 (34.5)
462 (20.5)
661 (63.7)
129 (48.0)
Compassionate/other leave
557 (15.3)
302 (13.4)
226 (21.8)
29 (10.8)
Not working (retired/unemployed)
1687 (46.4)
1442 (64.0)
139 (13.4)
104 (38.7)
Missing data
139 (3.8)
48 (2.1)
11 (1.1)
7 (2.6)
Caregiving timeb 0–2 h/day
1121 (30.8)
406 (18.0)
551 (53.1)
145 (53.9)
3–8 h/day
890 (24.5)
521 (23.1)
294 (28.4)
65 (24.2)
9–16 h/day
535 (14.7)
426 (18.9)
81 (7.8)
18 (6.7)
17–24 h/day
964 (26.5)
816 (36.2)
93 (9.0)
34 (12.6)
Missing data
125 (3.4)
85 (3.8)
18 (1.7)
7 (2.6) 195 (72.5)
Practical care timec 0–2 h/day
2167 (59.6)
1120 (49.7)
820 (79.1)
3–8 h/day
762 (21.0)
594 (26.4)
124 (12.0)
31 (11.5)
9–16 h/day
203 (5.6)
156 (6.9)
36 (3.5)
8 (3.0)
17–24 h/day
292 (8.0)
248 (11.0)
27 (2.6)
10 (3.7)
Missing data
211 (5.8)
136 (6.0)
30 (2.9)
25 (9.3)
Preparedness Low level
598 (16.5)
402 (17.8)
150 (14.5)
34 (12.6)
High level
2851 (78.4)
1702 (75.5)
871 (84.0)
227 (84.4)
Not relevant
66 (1.8)
55 (2.5)
5 (0.5)
5 (1.9)
Missing data
120 (3.3)
95 (4.2)
11 (1.1)
3 (1.1)
Communication Low level
295 (8.1)
216 (9.6)
58 (5.6)
14 (5.2)
High level
3108 (85.5)
1860 (82.5)
951 (91.7)
243 (90.3)
Missing data
232 (6.4)
178 (7.9)
28 (2.7)
12 (4.5)
29.1 (28.8–29.4)
30.7 (30.3–31.1)
26.2 (25.6–26.8)
26.6 (25.3–27.8)
No
2941 (80.9)
1765 (78.3)
901 (86.9)
228 (84.8)
Yes
544 (15.0)
386 (17.1)
117 (11.3)
28 (10.4)
150 (4.1)
103 (4.6)
19 (1.8)
13 (4.8)
12.2 (11.9–12.5)
12.9 (12.5–13.3)
11.3 (10.8–11.9)
9.7 (8.7–10.7) 182 (67.7)
Pre-loss grief symptomsd Sum score mean(95 % CI) Categorized:
Missing data Depressive symptomse Sum score mean (95 % CI) Categorized: No
2141 (58.9)
1269 (56.3)
659 (63.6)
Mild
692 (19.1)
452 (20.0)
188 (18.1)
42 (15.6)
Moderate
405 (11.1)
261 (11.6)
102 (9.8)
29 (10.8)
Severe
181 (5.0)
115 (5.1)
54 (5.2)
7 (2.6)
Missing data
216 (5.9)
157 (7.0)
34 (3.3)
9 (3.3)
25.6 (25.2–26.0)
25.5 (25.0–26.1)
26.6 (25.7–27.4)
22.5 (20.9–24.1)
Caregiver burden Sum score mean (95 % CI) Categorized None or mild
3050 (83.9)
1901 (84.3)
867 (83.6)
227 (84.4)
Moderate
377 (10.4)
212 (9.4)
134 (12.9)
24 (8.9)
Severe
40 (1.1)
26 (1.2)
13 (1.3)
1 (0.4)
Missing data
168 (4.6)
115 (5.1)
23 (2.2)
17 (6.3)
Support Care Cancer Table 3 (continued) Relation to patient (n = 3560)a All caregivers (n = 3635)
Partners 2254 (62.0 %)
Adult children 1037(28.5 %)
Other relation 269 (7.4 %)
Physical Component Score (PCS), mean (95% CI)
50.2 (50.4–51.1)
49.8 (49.3–50.2)
53.0 (52.5–53.6)
50.5 (49.2–51.7)
Mental Component Score (MCS), mean (95% CI)
41.0 (40.6–41.4)
40.4 (39.6–40.8)
41.5 (40.6–42.3)
44.8 (43.3–46.4)
Missing data
168 (4.6)
126 (5.6)
19 (1.8)
12 (4.5)
Healthf
a
Information on relation to the patient was not obtained from the questionnaire for 75 caregivers (2.1 %)
b
The question was: BHow many hours did you spend per day (24 h) on providing care for your relative?^
c
The question was: BHow many hours did you spend per day (24 h) on practical care for your relative, e.g. bath, food, toilet visits, medication?^
d
Pre-loss version of the Prolonged Grief-13 scale (PG-13)
e
Beck’s Depression Inventory-II (BDI-II)
f
The 36-item Short Form Health Survey-36 (SF-36): Bcomponent scores^ are summary scores for physical and mental health, respectively
measured with the same scale [29, 30]. Caregiver burden was higher among adult children compared to partners. This association has previously been sparsely reported, and our finding was consistent with a prior study [42]. The large subgroup of adult children in the present study (n = 1037) were predominantly female (76 %), many had children living at home (44 %), most attended a full-time job (64 %), and 22 % had compassionate leave. Almost half of the adult children spent at least 3 h per day providing care for their ill parent although only one fifth had compassionate leave. Taking care of an ill parent while working full-time and attending one’s own family may be a complex and stressful situation. More adult children than partners reported low preparedness and a low level of communication about illness and death, which may also contribute to the increase in caregiver distress. Caregiver burden has previously been associated with younger age, being a wife (compared to being a daughter) [43], female gender, living with the patient, a high number of caregiving hours, and inability to continue regular employment [14]. The present study differs partly from these findings as adult children of terminally ill patients were found to have the highest level of caregiver burden. On the other hand, adult children were younger than partners, and it seems likely that the level of caregiver burden would be strongly affected by having young children and full-time work. Our findings add to the understanding of the caregivers’ situation and indicate that lack of time and concurrent commitments may play a central role for caregiver burden. The main advantage of this study was the systematic population-based sampling procedure and nationwide recruitment. This provided a high generalizability and a large number of participating caregivers of terminally ill patients, which gave high precision in the estimates and allowed us to perform subgroup analyses, which are scarce in the existing literature [3]. Further strengths of this study were the completeness of
the questionnaire data and the precise classification in the used registries [44]. Thus, the risk of information bias is estimated to be very low. The response rate was 38 %, which is comparable to other studies of vulnerable populations [6]. Higher participation rates have been reported in studies recruiting caregivers directly through health professionals, but these studies had considerably fewer participants and were not population-based [29, 45]. The sampling procedure allowed recruitment of a large sample of patients without the influence from health professionals; this approach reduced the risk of selection bias and permitted non-response analysis. Regarding the generalizability of the findings, the largest group of non-respondents (n = 4468) stated no reason for nonparticipation. The reported main reasons for non-participation were that the patient was imminently dying (n = 284), lack of caregiver energy (n = 122), and irrelevance as no symptoms of terminal illness were present (n = 139) (Fig. 1). The first two reported reasons may cause underestimation of distress to the results reported, while the last may cause overestimation. This information point to diverging reasons for non-participation, and no systematic selection of participating caregivers seems to have occurred in our caregiver cohort. However, patients of responding caregivers had a higher educational level and were younger than patients with no responding caregiver, which might influence the generalizability of our results and cause overestimation of psychological distress in the caregiver due to younger patient age. The median survival time from formal registration of drug reimbursement to death was 93 days for patients with responding caregivers compared to 47 days for patients with non-responding caregivers. This indicates that only a minority of the patients of responding caregivers are in the late terminal phase, which is the period when most caregivers tend to report depressive symptoms [11]. This would potentially lead caregivers to report lower levels of
Support Care Cancer
psychological problems compared to caregivers to patients in the late terminal phase. The nationwide cohort of approximately 11,000 terminally ill patients with drug reimbursement status during 2012 may comprise a selected group of terminally ill patients as nearly 50,000 patients die from illness-related death each year in Denmark [46]. Due to multiple causes of death, these patients may comprise a heterogeneous group. In comparison, patients registered with drug reimbursement may be more likely to have a low socioeconomic position due to their need for financial remuneration. Higher levels of psychological distress in caregivers may be seen in those with low socio-economic position, which might have caused higher levels of distress in our study. However, formal registration with drug reimbursement due to terminal illness may also provide an opportunity for communication with health professionals regarding the patient’s illness stage, which may reduce caregiver distress. In conclusion, the caregivers in this study comprise a systematically sampled large-scale population-based cohort, which appears to be representative of caregivers of patients with life-threatening illness in the terminal phase.
Implications for clinical practice and future research Providing care to a severely ill loved one is an important human activity; it helps the patient, benefits the caregiver, and relieves the health-care system. Increased focus is required to ensure that the demands do not exhaust the resources of the caregivers, reduce the risk of adverse outcomes, and improve the terminal phase for both patient and caregivers. A strategy for systematic assessment of caregiver resources and needs during caregiving in both primary care and hospital settings is warranted along with guidelines for interdisciplinary support. For caregivers on the labor market, additional practical arrangements such as compassionate leave and/or flexible working hours should be an option to ensure sufficient time and energy for supportive caregiving and ultimately a decent end-oflife period for the patient. Assessment of symptoms of grief and depression in caregivers before the patient’s death may prove useful to reduce distress during caregiving and prevent adverse outcome during bereavement. Ongoing contact between family and health professionals is essential for accurate assessment of caregivers’ distress and need for targeted support or treatment for caregivers during the patient’s illness trajectory. Interdisciplinary collaboration between health professionals such as primary care physicians, community nurses, hospital staff, and specialized palliative care teams may ensure smooth transitions between
different health-care units and continuity in the care for both patients and caregivers. Adverse outcomes may be prevented if health professionals have the opportunity to prepare caregivers of terminally ill patients for the caregiving role through information on the practical aspects of end-of-life care and possibility to get assistance from their general practitioner, community nurses, hospital staff, or the family network. Future research may be directed at in-depth analysis of specific risk factors (e.g., low preparedness) of adverse outcomes in caregivers and also during the patient’s illness trajectory.
Conclusion This nationwide large-scale cohort comprised 3635 caregivers who were recruited through terminally ill patients, regardless of needs for specialized palliative care. The analyses revealed substantial levels of psychological distress in this group. In total, 15 % of all caregivers reported severe pre-loss grief symptoms, 16 % suffered from moderate to severe depressive symptoms, and 12 % experienced caregiver burden. Caregiver burden seemed more prevalent in adult children, while partners reported significantly higher levels of pre-loss grief and depressive symptoms. These findings implicate a need for directing attention towards the individual caregivers and their specific needs for support, e.g., in terms of preparation for the patient’s end-oflife period, practical arrangements regarding home care, and regular contact with health-care professionals. Caregivers are placed in a vulnerable position and face increased risk of distress; there seems to be a need to place this at the center of attention in future end-of-life care initiatives. Acknowledgments We wish to extend our profound gratitude towards patients and caregivers who participated in this study. Thanks to the staff at the Research Unit for General Practice, in particular IT specialist Hanne Beyer for setting up the database, professor Peter Vedsted and IT specialist Kaare Rud Flarup for providing us with access to the registry data from the Cancer in Primary and Secondary Sector (CAPS) database, statistician Anders Helles Carlsen for statistical assistance, and language editor Lone Niedziella for proof reading the manuscript. This work was supported by the Danish Cancer Society, the Danish foundation TrygFonden, and the Danish Health Foundation (Helsefonden). Compliance with ethical standards According to the Committee on Health Research Ethics of the Central Denmark Region, the Danish Act on Research Ethics Review of Health Research Projects did not apply to this study as questionnaire surveys generally do not require ethical clearance. The study was approved by the Danish Data Protection Agency (File no. 2013–41-2603). Disclosures None
Support Care Cancer
References 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Weibull A, Olesen F, Neergaard MA (2008) Caregivers’ active role in palliative home care - to encourage or to dissuade? A qualitative descriptive study. BMC Palliat Care 16 7:15-684X-7-15. Holm M, Henriksson A, Carlander I, Wengstrom Y, Ohlen J (2014) Preparing for family caregiving in specialized palliative home care: an ongoing process. Palliat Support Care 9:1–9. doi:10.1017/ S1478951514000558 Williams AL, McCorkle R (2011) Cancer family caregivers during the palliative, hospice, and bereavement phases: a review of the descriptive psychosocial literature. Palliat Support Care 9(3):315– 325. doi:10.1017/S1478951511000265 Davis EL, Deane FP, Lyons GC (2014) Acceptance and valued living as critical appraisal and coping strengths for caregivers dealing with terminal illness and bereavement. Palliat Support Care 30: 1–10. doi:10.1017/S1478951514000431 Burton AM, Sautter JM, Tulsky JA, Lindquist JH, Hays JC, Olsen MK, et al. (2012) Burden and well-being among a diverse sample of cancer, congestive heart failure, and chronic obstructive pulmonary disease caregivers. J Pain Symptom Manag 44(3):410–420. doi:10.1016/j.jpainsymman.2011.09.018 Thomas K, Hudson P, Trauer T, Remedios C, Clarke D (2014) Risk factors for developing prolonged grief during bereavement in family carers of cancer patients in palliative care: a longitudinal study. J Pain Symptom Manag 47(3):531–541. doi:10.1016/j.jpainsymman. 2013.05.022 Nanni MG, Biancosino B, Grassi L (2014) Pre-loss symptoms related to risk of complicated grief in caregivers of terminally ill cancer patients. J Affect Disord 60:87–91. doi:10.1016/j.jad.2013. 12.023 Hudson PL, Thomas K, Trauer T, Remedios C, Clarke D (2011) Psychological and social profile of family caregivers on commencement of palliative care. J Pain Symptom Manag 41(3):522–534. doi:10.1016/j.jpainsymman.2010.05.006 Tomarken A, Holland J, Schachter S, Vanderwerker L, Zuckerman E, Nelson C, et al. (2008) Factors of complicated grief pre-death in caregivers of cancer patients. Psychooncology 17(2):105–111 Parker Oliver D, Albright DL, Washington K, Wittenberg-Lyles E, Gage A, Mooney M, et al. (2013) Hospice caregiver depression: the evidence surrounding the greatest pain of all. J Soc Work End Life Palliat Care 9(4):256–271. doi:10.1080/15524256.2013.846891 Grunfeld E, Coyle D, Whelan T, Clinch J, Reyno L, Earle CC, et al. (2004) Family caregiver burden: results of a longitudinal study of breast cancer patients and their principal caregivers. CMAJ 8 170(12):1795–1801 Morishita M, Kamibeppu K (2014) Quality of life and satisfaction with care among family caregivers of patients with recurrent or metastasized digestive cancer requiring palliative care. Support Care Cancer 22(10):2687–2696. doi:10.1007/s00520-014-2259-3 Kenny P, King MT, Hall J (2014) The physical functioning and mental health of informal carers: evidence of care-giving impacts from an Australian population-based cohort. Health Soc Care Community 22(6):646–659. doi:10.1111/hsc.12136 Adelman RD, Tmanova LL, Delgado D, Dion S, Lachs MS (2014) Caregiver burden: a clinical review. JAMA 311(10):1052–1060. doi:10.1001/jama.2014.304 Bachner YG, O’Rourke N, Davidov E, Carmel S (2009) Mortality communication as a predictor of psychological distress among family caregivers of home hospice and hospital inpatients with terminal cancer. Aging Ment Health 13(1):54–63. doi:10.1080/ 13607860802154473 Hauksdottir A, Steineck G, Furst CJ, Valdimarsdottir U (2010) Long-term harm of low preparedness for a wife’s death from
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
cancer—a population-based study of widowers 4-5 years after the loss. Am J Epidemiol 15 172(4):389–396. doi:10.1093/aje/kwq147 Romito F, Goldzweig G, Cormio C, Hagedoorn M, Andersen BL (2013) Informal caregiving for cancer patients. Cancer 1 119(11): 2160–2169. doi:10.1002/cncr.28057 Hudson P (2013) Improving support for family carers: key implications for research, policy and practice. Palliat Med 27(7):581– 582. doi:10.1177/0269216313488855 Danish Medicines Agency; (2013). Antallet af ansøgninger om individuelle tilskud i 2012 [Applications for individual drug reimbursement in 2012]. Available at: http://laegemiddelstyrelsen.dk/ da/udgivelser/2013/antallet-af-ansoegninger-om-individuelletilskud-i-2012. Accessed 10 January 2016 Danish Health and Medicines Authority (2015). Plejeorlov til pasning af nærtståendei hjemmet [Compassionate leave to care for a close relative at home]. Available at: https://www.sundhed. dk/borger/sygdomme-a-aa/sociale-ydelser/sociale-ydelser/ydelser/ plejeorlov-til-pasning-af-naertstaaende-i-hjemmet/. Accessed 08 April 2015 Statistics Denmark. Data for research (2014). Available at: http:// www.dst.dk/en/TilSalg/Forskningsservice.aspx. Accessed 05 May 2015 Vandenbroucke JP, von Elm E, Altman DG, Gotzsche PC, Mulrow CD, Pocock SJ, et al. (2014) Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. Int J Surg 12(12):1500–1524. doi:10.1016/j.ijsu.2014.07. 014 Beck A, Steer R, Brown G (1996) Manual: Beck Depression Inventory, 2nd ed. The Psychological Corp., 2., edition edn. Hartcourt &Brace, San Antonio, Tx Guldin MB, Vedsted P, Jensen AB, Olesen F, Zachariae R (2013) Bereavement care in general practice: a cluster-randomized clinical trial. Fam Pract 30(2):134–141. doi:10.1093/fampra/cms053 Warmenhoven F, van Rijswijk E, Engels Y, Kan C, Prins J, van Weel C, et al. (2012) The Beck Depression Inventory (BDI-II) and a single screening question as screening tools for depressive disorder in Dutch advanced cancer patients. Support Care Cancer 20(2): 319–324. doi:10.1007/s00520-010-1082-8 Christensen KS, Bech P, Fink P (2010) Measuring mental health by questionnaires in primary care-unidimensionality, responsiveness and compliance. European Psychiatric Review 3(1):8–12 Keller SD, Ware JE Jr, Bentler PM, Aaronson NK, Alonso J, Apolone G, et al. (1998) Use of structural equation modeling to test the construct validity of the SF-36 Health Survey in ten countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 51(11):1179–1188 Kurtz ME, Kurtz JC, Given CW, Given BA (2004) Depression and physical health among family caregivers of geriatric patients with cancer—a longitudinal view. Med Sci Monit 10(8):CR447–CR456 Brogaard T, Neergaard MA, Guldin MB, Sokolowski I, Vedsted P (2013) Translation, adaptation and data quality of a Danish version of the Burden Scale for Family Caregivers. Scand J Caring Sci 27(4):1018–1026. doi:10.1111/j.1471-6712.2012.01092.x Grau H, Graessel E, Berth H (2015) The subjective burden of informal caregivers of persons with dementia: extended validation of the German language version of the Burden Scale for Family Caregivers (BSFC). Aging Ment Health 19(2):159–168. doi:10. 1080/13607863.2014.920296 Asgeirsdottir HG, Valdimarsdottir U, Furst CJ, Steineck G, Hauksdottir A (2013) Low preparedness before the loss of a wife to cancer and the widower’s chronic pain 4-5 years later-a population-based study. Psychooncology 22(12):2763–2770. doi: 10.1002/pon.3345 Bachner YG, Gesis-Za DE, Carmel S (2008) Caregivers’ communication with patients about illness and death: initial validation of a scale. Omega (Westport) 57(4):381–397
Support Care Cancer 33.
Statistics Denmark (2015). Quality declarations Available at: http:// www.dst.dk/en/Statistik/dokumentation/declarations. Accessed 05 May 2015 34. Gjerstorff ML (2011) The Danish Cancer Registry. Scand J Public Health 39(7 Suppl):42–45. doi:10.1177/1403494810393562 35. Lynge E, Sandegaard JL, Rebolj M (2011) The Danish National Patient Register. Scand J Public Health 39(7 Suppl):30–33. doi:10. 1177/1403494811401482 36. Sundararajan V, Quan H, Halfon P, Fushimi K, Luthi JC, Burnand B, et al. (2007) Cross-national comparative performance of three versions of the ICD-10 Charlson index. Med Care 45(12):1210– 1215 37. Thygesen SK, Christiansen CF, Christensen S, Lash TL, Sorensen HT (2011) The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the populationbased Danish National Registry of Patients. BMC Med Res Methodol 28 11:83. doi:10.1186/1471-2288-11-83 38. Ware JEJ, Gandek B, Kosinski M, Aaronson NK, Apolone G, Brazier J, et al. (1998) The equivalence of SF-36 summary health scores estimated using standard and country-specific algorithms in 10 countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 51(11):1167–1170 39. Li Q, Loke AY (2013) A spectrum of hidden morbidities among spousal caregivers for patients with cancer, and differences between the genders: a review of the literature. Eur J Oncol Nurs 17(5):578– 587. doi:10.1016/j.ejon.2013.01.007
40.
41.
42.
43. 44. 45.
46.
Ellervik C, Kvetny J, Christensen KS, Vestergaard M, Bech P (2014) Prevalence of depression, quality of life and antidepressant treatment in the Danish General Suburban Population Study. Nord J Psychiatry 29. doi:10.3109/08039488.2013.877074 Gandek B, Ware JE Jr (1998) Methods for validating and norming translations of health status questionnaires: the IQOLA Project approach. International Quality of Life Assessment. J Clin Epidemiol 51(11):953–959 Given B, Wyatt G, Given C, Sherwood P, Gift A, DeVoss D, et al. (2004) Burden and depression among caregivers of patients with cancer at the end of life. Oncol Nurs Forum, Nov 16 31(6):1105– 1117 Joanna Briggs Institute (2012) Caregiver burden of terminally-ill adults in the home setting. Nurs Health Sci 14(4):435–437 Frank L (2000) Epidemiology. When an entire country is a cohort. Science 31 287(5462):2398–2399 Guldin MB, Vedsted P, Zachariae R, Olesen F, Jensen AB (2012) Complicated grief and need for professional support in family caregivers of cancer patients in palliative care: a longitudinal cohort study. Support Care Cancer 20(8):1679–1685. doi:10.1007/ s00520-011-1260-3 Statistics Denmark (2015). Dødsfald og middellevetid [Deaths and average lifespan]. Available at: http://www.dst.dk/da/ Statistik/emner/doedsfald-og-middellevetid/doedsfald.aspx. Accessed 05 May 2015.
ORIGINAL ARTICLE
Psychological distress, health, and socio-economic factors in caregivers of terminally ill patients: a nationwide population-based cohort study Mette Kjaergaard Nielsen 1 & Mette Asbjoern Neergaard 2 & Anders Bonde Jensen 3 & Flemming Bro 1 & Mai-Britt Guldin 1,2
Received: 10 September 2015 / Accepted: 9 February 2016 # Springer-Verlag Berlin Heidelberg 2016
Abstract Purpose At some point in life, most people become caregivers to a terminally ill relative. Previous studies have shown that many caregivers experience psychological distress and declining physical health, but these studies have predominantly been conducted in specialized palliative care settings. Therefore, caregiver studies with a population-based approach are needed. We aimed to describe socio-economic characteristics, situational factors, pre-loss grief symptoms, depressive symptoms, caregiver burden, and health status in a general population of caregivers to terminally ill patients. Method We conducted a nationwide population-based cohort study. Caregivers were systematically recruited through patients registered with drug reimbursement for terminal illness in 2012. Data on socio-economic characteristics was mainly obtained from Danish registries, whereas data on situational factors, distress, and health was measured in questionnaires. Results Of patients to responding caregivers (n = 3635), 89 % suffered from cancer, predominantly lung cancer (23 %). Of responding caregivers, 62 % were partners and 29 % were adult children. In total, one third of caregivers reported severe outcome, 15 % reported severe pre-loss grief symptoms, 16.1 % had moderate to severe depressive symptoms, and
* Mette Kjaergaard Nielsen [email protected]
1
Research Unit for General Practice, Aarhus University, Bartholins Allé 2, 8000 Aarhus C, Denmark
2
The Palliative Team, Department of Oncology, Aarhus University Hospital, Noerrebrogade 44, 8000 Aarhus C, Denmark
3
Department of Oncology, Aarhus University Hospital, Noerrebrogade 44, 8000 Aarhus C, Denmark
12 % experienced high caregiver burden. Partners had the highest levels of pre-loss grief and depressive symptoms, while adult children reported the highest levels of caregiver burden. Conclusions From this cohort, which was estimated to be representative of caregivers to terminally ill relatives in the general population, we found high levels of pre-loss grief, depressive symptoms, and/or caregiver burden in one third of all caregivers. These findings call for increased focus on caregivers’ need of support. Keywords Informal caregivers . Terminal illness . Pre-loss grief . Depression . Caregiverburden . Socio-economicfactors
Background Severe illness in a close relative is a big challenge to face. Cancer, COPD, heart failure, and other severe diseases cause a large number of deaths in modern society and inevitably place relatives in the role as caregivers. The need for informal caregiving is expected to increase due to an aging population in Western countries; fortunately, most caregivers wish to take part in giving care to their close relatives [1, 2]. However, the need for caregiving may increase as the disease progresses [3, 4], and caregivers may be at risk of developing illness themselves [4] if the demands for care exceed their resources [5]. Informal caregivers are at risk of experiencing severe preloss grief symptoms [6, 7]. This implies that the grief symptoms in caregivers can be very severe early in the terminal illness trajectory and affect their daily life functioning. Furthermore, severe pre-loss grief symptoms have been found to strongly predict complicated grief (i.e., severe grief impairing caregiver functioning for more than 6 months after the loss) [6, 7]. The few existing studies on pre-loss grief [6–9]
Support Care Cancer
are mainly based on data from specialized palliative care settings, and severe pre-loss grief was found in about 15 % of caregivers to these patients [8]. Caregiving may cause other adverse outcomes. Former studies have shown that depressive symptoms are present in up to half of caregivers to dying patients [10, 11], that the physical functioning of caregivers is impaired during illness [12, 13], and that caregiver burden is reported in up to 32 % of caregivers [14]. These adverse outcomes in caregivers seem to be associated with situational factors during the patient’s illness trajectory; these include low level of communication between the patient and the caregiver about illness and death [15], perceived low preparedness in the caregiver for the impending death [16], and a high number of caregiving hours [14]. In a recent review, females and young caregivers were found to be at particular risk of developing psychological distress, while inconclusive results were reported regarding the caregiver’s relations with the patient [3]. Large-scale population-based studies of informal caregivers to adult patients with terminal illness are needed [17, 18]. Specifically, we need more knowledge on adverse outcomes and socio-economic characteristics of caregivers, including caregivers to terminally ill patients who do not receive specialized palliative care. Such knowledge may provide a solid foundation for more targeted and effective support to caregivers [18]. Therefore, in a large-scale nationwide caregiver cohort, we aimed to: 1) describe socio-economic characteristics of caregivers and situational factors including number of caregiving hours and level of preparedness, and 2) investigate levels of pre-loss grief symptoms, depressive symptoms, caregiver burden, and health status in caregivers Furthermore, we wanted to investigate differences between caregivers in the abovementioned factors on the basis of the caregiver’s relation to the patient.
Methods Setting The Danish health-care system is tax-funded, and welfare services provided at hospitals and in primary care are free of charge. However, medication expenses are only partly covered, but citizens assessed by a physician as incurably ill are entitled to full drug reimbursement due to terminal illness after formal registration at the Danish Medicines Agency [19]. Furthermore, a relative of the terminally ill patient may take
compassionate leave and has a statutory right to receive a special allowance during such leave [20]. For the present study, we systematically approached caregivers of all terminally ill patients in both a general and a specialized palliative care setting through letters to patients receiving drug reimbursement for terminal illness in Denmark in 2012. Data from caregiver questionnaires and Danish national registers were linked at an individual level using the Danish personal identification number (CPR number). All data were anonymized and accessed at Statistics Denmark through an authorized connection [21] and reported according to the STROBE Statement [22]. Sampling procedure On a weekly basis throughout 2012, the first author received a list from the Danish Health and Medicines Authority of all patients newly registered for drug reimbursement. The patients received a letter within 2 days and were asked to forward an enclosed questionnaire to their closest relative. The letter contained a participant information leaflet and a questionnaire with written consent form. Variables and data sources Age and gender were calculated from CPR numbers of patients and caregivers, and patients were followed-up for survival until 1 June 2014. Questionnaire-based scales and variables A pre-loss version of the Prolonged Grief-13 scale (PG-13) measured grief symptoms presented as a sum score and categorical criteria scores (yes, no) based on the criteria for the PG-13 scale [8]. The 21-item Beck’s Depression Inventory-II (BDI-II) measured depressive symptoms [23]. An additional category (does not apply) was added to item 21 on sexuality as the question was considered offensive in a previous caregiver study [24]. Sum scores and categorizations were handled in accordance with the manual [23, 25]. The 36-item Short Form Health Survey (SF-36) measured health status on a physical component score (PCS) and a mental component score (MCS); these ranged from 0 (worst) to 100 (best) [26–28]. The Burden Scale for Family Caregivers (BSFC) was reported as a sum score and a categorical score (severe, moderate, mild, no symptoms) according to the manual [29, 30]. Ad hoc items addressed the caregiver’s relation to the patient (partner, child, other), the amount of hours spent on both practical and total care (0–2, 3–8, 9–16, 17–24 h/day), and whether the caregiver had compassionate leave (yes, no, not on labor market). Caregiver’s preparedness for death was measured on a single-item question and reported as a dichotomized value (high, low) [31]. For the five-item Couples’
Support Care Cancer
Communication about Illness and Death scale (CCID) [15, 32], a sum score and a dichotomized value (high, low) was reported. Register-based data The socio-economic variables retrieved from Statistics Denmark [33] for both patients and relatives were cohabitation status (married/cohabiting, living alone), educational level (≤10 years, >10 and ≤15 years, >15 years of education), ethnicity (immigrant/descendant, non-immigrant/descendant), and urbanicity (50,000 inhabitants in the community) per 1 January 2012. For caregivers, data were also retrieved on children living at home (yes, no), residency (owned, rented), and gross income in 2011 (50,000 euros/year). Diagnoses for cancer patients were obtained from the ICD-10 codes recorded in the Danish Cancer Registry (DCR) [34]; these were categorized into colorectal, lung, breast, prostate, hematological, and other cancers. Patients without a cancer diagnosis were categorized with a Bnon-cancer^ diagnosis. Data on diagnoses from hospital admissions for both patients and caregivers were retrieved from the Danish National Patient Registry (NPR) [35]; these were used to identify the most frequent noncancer diagnosis among patients and the number of chronic diseases on the Charlson Comorbidity Index (CCI) [36, 37] for caregivers. Statistical analyses Variables were reported using a mean value with 95 % confidence intervals (CI) or a median with interquartile range or proportion, whichever appropriate. Patients of responding and non-responding relatives were compared by t tests or chi square tests to identify any differences in the socioeconomic characteristics between the two patient groups. In accordance with previous studies, continuous summary score variables for BSFC, SF-36, BDI-II, and pre-loss PG-13 were presented as a mean with 95 % CI [8, 23, 30, 38]. Only respondents with complete responses on the CCID, SF-36, BDIII, and pre-loss PG-13 or less than two missing values on the BSFC [29] were included in the analysis. Data were analyzed using the statistical software Stata 13.0 (StataCorp LP, TX).
Results Study population We obtained 11,628 CPR numbers of patients registered with drug reimbursement in 2012, and 9512 eligible patients were contacted (Fig. 1). A total of 3635 caregivers responded positively (38 %), and 3488 caregivers provided their CPR
number for inclusion in the register-based analysis (Fig. 1). The patients of responding caregivers were significantly more likely to be male, be young, have a partner, have a higher education, be non-immigrants, have a cancer diagnosis, and have a longer survival time after receiving drug reimbursement than patients of non-responding caregivers (Table 1). The median time to death of patients of non-participating relatives was 64 days after drug reimbursement and 93 days for patients with a participating caregiver (Table 1). Socio-economic and situational factors The majority of caregivers were partners (62 %), and 29 % were adult children. The vast majority (89 %) were caregivers to cancer patients, predominantly lung cancer (23 %) (Table 2). Most caregivers (66 %) spent more than 3 h per day providing care for the patient, and most caregivers reported high preparedness for the patient’s impending death, although 17 % of all caregivers stated that they were not prepared (Table 3). Pre-loss grief, depressive symptoms, health, and caregiver burden The criteria for severe pre-loss grief symptoms were fulfilled for 15 % of caregivers, and moderate to severe depressive symptoms was found in 16 %. The SF-36 mean PCS sum score was 50.2 (95 % CI 50.4–51.1), and the mean MCS sum score was 41.0 (95 % CI 40.6–41.4) (Table 3). Moderate to very severe caregiver burden was experienced by 12 % of caregivers. In total, 30 % of the caregivers reported symptoms of pre-loss grief, depression, or caregiver burden, and 13 % had more than one of the mentioned conditions (data not shown). Severe pre-loss grief symptoms were reported by 17 % of partners and 11 % of adult children (Table 3). In total, 17 % of partners and 15 % of adult children had moderate to severe depressive symptoms, 11 % of partners reported moderate to very severe caregiver burden, and 14 % of adult children experienced caregiver burden.
Discussion In this nationwide large-scale caregiver cohort, two third of the caregivers was partners of the patient, and 17 % reported low preparedness for the impending death. More than two third of the caregivers spent more than 3 h per day providing care for the patient. The criteria for severe pre-loss grief symptoms were met for 15 % of participants, moderate to severe depressive symptoms were found in 16 %, and high caregiver burden was experienced by 12 %. Adult children reported a higher level of caregiver burden than partners. However,
Support Care Cancer Fig. 1 Flow chart of study population of patients (pt.) granted drug reimbursement and their caregivers
Population source: Patients granted drug reimbursement in 2012 11,628 Pt. dead prior to contact: 1,348 (11.6%) Pt. protected from contact*: 757 (6.5%) Pt. 15 years of education Missing data Gross incomeb 50,000 euros/year Missing data Residency Owned Rented Missing Ethnicity Not immigrant/descendant Immigrant/descendant Missing data Urbanicity 50,000 inhabitants Missing data Number of chronic diseasesc 0 1–2 >2 Missing data
All caregivers
Partners n = 2254 (62.0 %)
Children n = 1037 (28.5 %)
Other relation n = 269 (7.4 %)
61.2 (60.8–61.6)
66.5 (66.1–66.9)
49.9 (49.2–50.5)
58.8 (57.2–60.3)
659 (18.1) 855 (23.5) 1067 (29.4) 907 (25.0) 147 (4.0)
152 (6.8) 377 (16.7) 848 (37.6) 848 (37.6) 29 (1.3)
450 (43.4) 400 (38.6) 134 (12.9) 9 (0.9) 44 (4.2)
57 (21.2) 77 (28.6) 83 (30.9) 47 (17.5) 5 (1.9)
2420 (66.6) 1068 (29.4) 147 (4.0)
1419 (63.0) 806 (35.8) 29 (1.3)
787 (75.9) 206 (19.9) 44 (4.2)
214 (79.5) 50 (18.6) 5 (1.9)
3026 (83.3) 451 (12.4) 158 (4.4)
2128 (94.4) 92 (4.1) 34 (1.5)
731 (70.5) 258 (24.9) 48 (4.6)
163 (60.6) 99 (36.8) 7 (2.6)
717 (19.7) 2760 (75.9) 158 (4.4)
198 (8.8) 2022 (89.7) 34 (1.5)
452 (43.6) 537 (51.8) 48 (4.6)
67 (24.9) 195 (72.5) 7 (2.6)
980 (27.0) 1651 (45.4) 802 (22.1) 202 (5.6)
741 (32.9) 1028 (45.6) 416 (18.5) 69 (3.0)
152 (14.7) 511 (49.3) 320 (30.8) 54 (5.2)
85 (31.6) 111 (41.3) 63 (23.4) 10 (3.7)
921 (25.3) 922 (25.4) 880 (24.2) 760 (20.9) 152 (4.2)
786 (34.9) 610 (27.1) 460 (20.4) 366 (16.2) 32 (1.4)
86 (8.3) 216 (20.8) 347 (33.5) 343 (33.1) 45 (4.3)
48 (17.8) 92 (34.2) 73 (27.2) 50 (18.6) 6 (2.2)
2512 (69.1) 942 (25.9) 181 (5.0)
1644 (72.9) 565 (25.1) 45 (2.0)
690 (66.5) 291 (28.1) 56 (5.4)
175 (65.1) 83 (30.9) 11 (4.1)
3389 (93.2) 88 (2.4) 158 (4.4)
2161 (95.9) 59 (2.6) 34 (1.5)
966 (93.2) 23 (2.2) 48 (4.6)
256 (95.2) 6 (2.2) 7 (2.6)
1391 (38.3) 1041 (28.6) 1038 (28.6) 165 (4.5)
924 (41.0) 662 (29.4) 629 (27.9) 39 (1.7)
369 (35.6) 303 (29.2) 317 (30.6) 48 (4.6)
96 (35.7) 76 (28.3) 88 (32.7) 9 (3.3)
2448 (67.4) 488 (13.4) 91 (2.5) 608 (16.7)
1483 (65.8) 372 (16.5) 72 (3.2) 327 (14.5)
761 (73.4) 85 (8.2) 14 (1.4) 177 (17.1)
200 (74.4) 30 (11.2) 5 (1.9) 34 (12.6)
All variables as of 1 January 2012, unless otherwise stated a
Information on relation to the patient was missing for 75 caregivers (2.1 %); these account for missing values in the relation variable
b
For the year 2011
c
Charlson Comorbidity Index (CCI) based on hospital registered diagnoses during 2001–2011. Therefore, missing values include persons unregistered at the hospital
Support Care Cancer Table 3
Self-reported caregiver data on situational factors, distress and health Relation to patient (n = 3560)a All caregivers (n = 3635)
Partners 2254 (62.0 %)
Adult children 1037(28.5 %)
Other relation 269 (7.4 %)
Situational factors Employment Working
1252 (34.5)
462 (20.5)
661 (63.7)
129 (48.0)
Compassionate/other leave
557 (15.3)
302 (13.4)
226 (21.8)
29 (10.8)
Not working (retired/unemployed)
1687 (46.4)
1442 (64.0)
139 (13.4)
104 (38.7)
Missing data
139 (3.8)
48 (2.1)
11 (1.1)
7 (2.6)
Caregiving timeb 0–2 h/day
1121 (30.8)
406 (18.0)
551 (53.1)
145 (53.9)
3–8 h/day
890 (24.5)
521 (23.1)
294 (28.4)
65 (24.2)
9–16 h/day
535 (14.7)
426 (18.9)
81 (7.8)
18 (6.7)
17–24 h/day
964 (26.5)
816 (36.2)
93 (9.0)
34 (12.6)
Missing data
125 (3.4)
85 (3.8)
18 (1.7)
7 (2.6) 195 (72.5)
Practical care timec 0–2 h/day
2167 (59.6)
1120 (49.7)
820 (79.1)
3–8 h/day
762 (21.0)
594 (26.4)
124 (12.0)
31 (11.5)
9–16 h/day
203 (5.6)
156 (6.9)
36 (3.5)
8 (3.0)
17–24 h/day
292 (8.0)
248 (11.0)
27 (2.6)
10 (3.7)
Missing data
211 (5.8)
136 (6.0)
30 (2.9)
25 (9.3)
Preparedness Low level
598 (16.5)
402 (17.8)
150 (14.5)
34 (12.6)
High level
2851 (78.4)
1702 (75.5)
871 (84.0)
227 (84.4)
Not relevant
66 (1.8)
55 (2.5)
5 (0.5)
5 (1.9)
Missing data
120 (3.3)
95 (4.2)
11 (1.1)
3 (1.1)
Communication Low level
295 (8.1)
216 (9.6)
58 (5.6)
14 (5.2)
High level
3108 (85.5)
1860 (82.5)
951 (91.7)
243 (90.3)
Missing data
232 (6.4)
178 (7.9)
28 (2.7)
12 (4.5)
29.1 (28.8–29.4)
30.7 (30.3–31.1)
26.2 (25.6–26.8)
26.6 (25.3–27.8)
No
2941 (80.9)
1765 (78.3)
901 (86.9)
228 (84.8)
Yes
544 (15.0)
386 (17.1)
117 (11.3)
28 (10.4)
150 (4.1)
103 (4.6)
19 (1.8)
13 (4.8)
12.2 (11.9–12.5)
12.9 (12.5–13.3)
11.3 (10.8–11.9)
9.7 (8.7–10.7) 182 (67.7)
Pre-loss grief symptomsd Sum score mean(95 % CI) Categorized:
Missing data Depressive symptomse Sum score mean (95 % CI) Categorized: No
2141 (58.9)
1269 (56.3)
659 (63.6)
Mild
692 (19.1)
452 (20.0)
188 (18.1)
42 (15.6)
Moderate
405 (11.1)
261 (11.6)
102 (9.8)
29 (10.8)
Severe
181 (5.0)
115 (5.1)
54 (5.2)
7 (2.6)
Missing data
216 (5.9)
157 (7.0)
34 (3.3)
9 (3.3)
25.6 (25.2–26.0)
25.5 (25.0–26.1)
26.6 (25.7–27.4)
22.5 (20.9–24.1)
Caregiver burden Sum score mean (95 % CI) Categorized None or mild
3050 (83.9)
1901 (84.3)
867 (83.6)
227 (84.4)
Moderate
377 (10.4)
212 (9.4)
134 (12.9)
24 (8.9)
Severe
40 (1.1)
26 (1.2)
13 (1.3)
1 (0.4)
Missing data
168 (4.6)
115 (5.1)
23 (2.2)
17 (6.3)
Support Care Cancer Table 3 (continued) Relation to patient (n = 3560)a All caregivers (n = 3635)
Partners 2254 (62.0 %)
Adult children 1037(28.5 %)
Other relation 269 (7.4 %)
Physical Component Score (PCS), mean (95% CI)
50.2 (50.4–51.1)
49.8 (49.3–50.2)
53.0 (52.5–53.6)
50.5 (49.2–51.7)
Mental Component Score (MCS), mean (95% CI)
41.0 (40.6–41.4)
40.4 (39.6–40.8)
41.5 (40.6–42.3)
44.8 (43.3–46.4)
Missing data
168 (4.6)
126 (5.6)
19 (1.8)
12 (4.5)
Healthf
a
Information on relation to the patient was not obtained from the questionnaire for 75 caregivers (2.1 %)
b
The question was: BHow many hours did you spend per day (24 h) on providing care for your relative?^
c
The question was: BHow many hours did you spend per day (24 h) on practical care for your relative, e.g. bath, food, toilet visits, medication?^
d
Pre-loss version of the Prolonged Grief-13 scale (PG-13)
e
Beck’s Depression Inventory-II (BDI-II)
f
The 36-item Short Form Health Survey-36 (SF-36): Bcomponent scores^ are summary scores for physical and mental health, respectively
measured with the same scale [29, 30]. Caregiver burden was higher among adult children compared to partners. This association has previously been sparsely reported, and our finding was consistent with a prior study [42]. The large subgroup of adult children in the present study (n = 1037) were predominantly female (76 %), many had children living at home (44 %), most attended a full-time job (64 %), and 22 % had compassionate leave. Almost half of the adult children spent at least 3 h per day providing care for their ill parent although only one fifth had compassionate leave. Taking care of an ill parent while working full-time and attending one’s own family may be a complex and stressful situation. More adult children than partners reported low preparedness and a low level of communication about illness and death, which may also contribute to the increase in caregiver distress. Caregiver burden has previously been associated with younger age, being a wife (compared to being a daughter) [43], female gender, living with the patient, a high number of caregiving hours, and inability to continue regular employment [14]. The present study differs partly from these findings as adult children of terminally ill patients were found to have the highest level of caregiver burden. On the other hand, adult children were younger than partners, and it seems likely that the level of caregiver burden would be strongly affected by having young children and full-time work. Our findings add to the understanding of the caregivers’ situation and indicate that lack of time and concurrent commitments may play a central role for caregiver burden. The main advantage of this study was the systematic population-based sampling procedure and nationwide recruitment. This provided a high generalizability and a large number of participating caregivers of terminally ill patients, which gave high precision in the estimates and allowed us to perform subgroup analyses, which are scarce in the existing literature [3]. Further strengths of this study were the completeness of
the questionnaire data and the precise classification in the used registries [44]. Thus, the risk of information bias is estimated to be very low. The response rate was 38 %, which is comparable to other studies of vulnerable populations [6]. Higher participation rates have been reported in studies recruiting caregivers directly through health professionals, but these studies had considerably fewer participants and were not population-based [29, 45]. The sampling procedure allowed recruitment of a large sample of patients without the influence from health professionals; this approach reduced the risk of selection bias and permitted non-response analysis. Regarding the generalizability of the findings, the largest group of non-respondents (n = 4468) stated no reason for nonparticipation. The reported main reasons for non-participation were that the patient was imminently dying (n = 284), lack of caregiver energy (n = 122), and irrelevance as no symptoms of terminal illness were present (n = 139) (Fig. 1). The first two reported reasons may cause underestimation of distress to the results reported, while the last may cause overestimation. This information point to diverging reasons for non-participation, and no systematic selection of participating caregivers seems to have occurred in our caregiver cohort. However, patients of responding caregivers had a higher educational level and were younger than patients with no responding caregiver, which might influence the generalizability of our results and cause overestimation of psychological distress in the caregiver due to younger patient age. The median survival time from formal registration of drug reimbursement to death was 93 days for patients with responding caregivers compared to 47 days for patients with non-responding caregivers. This indicates that only a minority of the patients of responding caregivers are in the late terminal phase, which is the period when most caregivers tend to report depressive symptoms [11]. This would potentially lead caregivers to report lower levels of
Support Care Cancer
psychological problems compared to caregivers to patients in the late terminal phase. The nationwide cohort of approximately 11,000 terminally ill patients with drug reimbursement status during 2012 may comprise a selected group of terminally ill patients as nearly 50,000 patients die from illness-related death each year in Denmark [46]. Due to multiple causes of death, these patients may comprise a heterogeneous group. In comparison, patients registered with drug reimbursement may be more likely to have a low socioeconomic position due to their need for financial remuneration. Higher levels of psychological distress in caregivers may be seen in those with low socio-economic position, which might have caused higher levels of distress in our study. However, formal registration with drug reimbursement due to terminal illness may also provide an opportunity for communication with health professionals regarding the patient’s illness stage, which may reduce caregiver distress. In conclusion, the caregivers in this study comprise a systematically sampled large-scale population-based cohort, which appears to be representative of caregivers of patients with life-threatening illness in the terminal phase.
Implications for clinical practice and future research Providing care to a severely ill loved one is an important human activity; it helps the patient, benefits the caregiver, and relieves the health-care system. Increased focus is required to ensure that the demands do not exhaust the resources of the caregivers, reduce the risk of adverse outcomes, and improve the terminal phase for both patient and caregivers. A strategy for systematic assessment of caregiver resources and needs during caregiving in both primary care and hospital settings is warranted along with guidelines for interdisciplinary support. For caregivers on the labor market, additional practical arrangements such as compassionate leave and/or flexible working hours should be an option to ensure sufficient time and energy for supportive caregiving and ultimately a decent end-oflife period for the patient. Assessment of symptoms of grief and depression in caregivers before the patient’s death may prove useful to reduce distress during caregiving and prevent adverse outcome during bereavement. Ongoing contact between family and health professionals is essential for accurate assessment of caregivers’ distress and need for targeted support or treatment for caregivers during the patient’s illness trajectory. Interdisciplinary collaboration between health professionals such as primary care physicians, community nurses, hospital staff, and specialized palliative care teams may ensure smooth transitions between
different health-care units and continuity in the care for both patients and caregivers. Adverse outcomes may be prevented if health professionals have the opportunity to prepare caregivers of terminally ill patients for the caregiving role through information on the practical aspects of end-of-life care and possibility to get assistance from their general practitioner, community nurses, hospital staff, or the family network. Future research may be directed at in-depth analysis of specific risk factors (e.g., low preparedness) of adverse outcomes in caregivers and also during the patient’s illness trajectory.
Conclusion This nationwide large-scale cohort comprised 3635 caregivers who were recruited through terminally ill patients, regardless of needs for specialized palliative care. The analyses revealed substantial levels of psychological distress in this group. In total, 15 % of all caregivers reported severe pre-loss grief symptoms, 16 % suffered from moderate to severe depressive symptoms, and 12 % experienced caregiver burden. Caregiver burden seemed more prevalent in adult children, while partners reported significantly higher levels of pre-loss grief and depressive symptoms. These findings implicate a need for directing attention towards the individual caregivers and their specific needs for support, e.g., in terms of preparation for the patient’s end-oflife period, practical arrangements regarding home care, and regular contact with health-care professionals. Caregivers are placed in a vulnerable position and face increased risk of distress; there seems to be a need to place this at the center of attention in future end-of-life care initiatives. Acknowledgments We wish to extend our profound gratitude towards patients and caregivers who participated in this study. Thanks to the staff at the Research Unit for General Practice, in particular IT specialist Hanne Beyer for setting up the database, professor Peter Vedsted and IT specialist Kaare Rud Flarup for providing us with access to the registry data from the Cancer in Primary and Secondary Sector (CAPS) database, statistician Anders Helles Carlsen for statistical assistance, and language editor Lone Niedziella for proof reading the manuscript. This work was supported by the Danish Cancer Society, the Danish foundation TrygFonden, and the Danish Health Foundation (Helsefonden). Compliance with ethical standards According to the Committee on Health Research Ethics of the Central Denmark Region, the Danish Act on Research Ethics Review of Health Research Projects did not apply to this study as questionnaire surveys generally do not require ethical clearance. The study was approved by the Danish Data Protection Agency (File no. 2013–41-2603). Disclosures None
Support Care Cancer
References 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Weibull A, Olesen F, Neergaard MA (2008) Caregivers’ active role in palliative home care - to encourage or to dissuade? A qualitative descriptive study. BMC Palliat Care 16 7:15-684X-7-15. Holm M, Henriksson A, Carlander I, Wengstrom Y, Ohlen J (2014) Preparing for family caregiving in specialized palliative home care: an ongoing process. Palliat Support Care 9:1–9. doi:10.1017/ S1478951514000558 Williams AL, McCorkle R (2011) Cancer family caregivers during the palliative, hospice, and bereavement phases: a review of the descriptive psychosocial literature. Palliat Support Care 9(3):315– 325. doi:10.1017/S1478951511000265 Davis EL, Deane FP, Lyons GC (2014) Acceptance and valued living as critical appraisal and coping strengths for caregivers dealing with terminal illness and bereavement. Palliat Support Care 30: 1–10. doi:10.1017/S1478951514000431 Burton AM, Sautter JM, Tulsky JA, Lindquist JH, Hays JC, Olsen MK, et al. (2012) Burden and well-being among a diverse sample of cancer, congestive heart failure, and chronic obstructive pulmonary disease caregivers. J Pain Symptom Manag 44(3):410–420. doi:10.1016/j.jpainsymman.2011.09.018 Thomas K, Hudson P, Trauer T, Remedios C, Clarke D (2014) Risk factors for developing prolonged grief during bereavement in family carers of cancer patients in palliative care: a longitudinal study. J Pain Symptom Manag 47(3):531–541. doi:10.1016/j.jpainsymman. 2013.05.022 Nanni MG, Biancosino B, Grassi L (2014) Pre-loss symptoms related to risk of complicated grief in caregivers of terminally ill cancer patients. J Affect Disord 60:87–91. doi:10.1016/j.jad.2013. 12.023 Hudson PL, Thomas K, Trauer T, Remedios C, Clarke D (2011) Psychological and social profile of family caregivers on commencement of palliative care. J Pain Symptom Manag 41(3):522–534. doi:10.1016/j.jpainsymman.2010.05.006 Tomarken A, Holland J, Schachter S, Vanderwerker L, Zuckerman E, Nelson C, et al. (2008) Factors of complicated grief pre-death in caregivers of cancer patients. Psychooncology 17(2):105–111 Parker Oliver D, Albright DL, Washington K, Wittenberg-Lyles E, Gage A, Mooney M, et al. (2013) Hospice caregiver depression: the evidence surrounding the greatest pain of all. J Soc Work End Life Palliat Care 9(4):256–271. doi:10.1080/15524256.2013.846891 Grunfeld E, Coyle D, Whelan T, Clinch J, Reyno L, Earle CC, et al. (2004) Family caregiver burden: results of a longitudinal study of breast cancer patients and their principal caregivers. CMAJ 8 170(12):1795–1801 Morishita M, Kamibeppu K (2014) Quality of life and satisfaction with care among family caregivers of patients with recurrent or metastasized digestive cancer requiring palliative care. Support Care Cancer 22(10):2687–2696. doi:10.1007/s00520-014-2259-3 Kenny P, King MT, Hall J (2014) The physical functioning and mental health of informal carers: evidence of care-giving impacts from an Australian population-based cohort. Health Soc Care Community 22(6):646–659. doi:10.1111/hsc.12136 Adelman RD, Tmanova LL, Delgado D, Dion S, Lachs MS (2014) Caregiver burden: a clinical review. JAMA 311(10):1052–1060. doi:10.1001/jama.2014.304 Bachner YG, O’Rourke N, Davidov E, Carmel S (2009) Mortality communication as a predictor of psychological distress among family caregivers of home hospice and hospital inpatients with terminal cancer. Aging Ment Health 13(1):54–63. doi:10.1080/ 13607860802154473 Hauksdottir A, Steineck G, Furst CJ, Valdimarsdottir U (2010) Long-term harm of low preparedness for a wife’s death from
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
cancer—a population-based study of widowers 4-5 years after the loss. Am J Epidemiol 15 172(4):389–396. doi:10.1093/aje/kwq147 Romito F, Goldzweig G, Cormio C, Hagedoorn M, Andersen BL (2013) Informal caregiving for cancer patients. Cancer 1 119(11): 2160–2169. doi:10.1002/cncr.28057 Hudson P (2013) Improving support for family carers: key implications for research, policy and practice. Palliat Med 27(7):581– 582. doi:10.1177/0269216313488855 Danish Medicines Agency; (2013). Antallet af ansøgninger om individuelle tilskud i 2012 [Applications for individual drug reimbursement in 2012]. Available at: http://laegemiddelstyrelsen.dk/ da/udgivelser/2013/antallet-af-ansoegninger-om-individuelletilskud-i-2012. Accessed 10 January 2016 Danish Health and Medicines Authority (2015). Plejeorlov til pasning af nærtståendei hjemmet [Compassionate leave to care for a close relative at home]. Available at: https://www.sundhed. dk/borger/sygdomme-a-aa/sociale-ydelser/sociale-ydelser/ydelser/ plejeorlov-til-pasning-af-naertstaaende-i-hjemmet/. Accessed 08 April 2015 Statistics Denmark. Data for research (2014). Available at: http:// www.dst.dk/en/TilSalg/Forskningsservice.aspx. Accessed 05 May 2015 Vandenbroucke JP, von Elm E, Altman DG, Gotzsche PC, Mulrow CD, Pocock SJ, et al. (2014) Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. Int J Surg 12(12):1500–1524. doi:10.1016/j.ijsu.2014.07. 014 Beck A, Steer R, Brown G (1996) Manual: Beck Depression Inventory, 2nd ed. The Psychological Corp., 2., edition edn. Hartcourt &Brace, San Antonio, Tx Guldin MB, Vedsted P, Jensen AB, Olesen F, Zachariae R (2013) Bereavement care in general practice: a cluster-randomized clinical trial. Fam Pract 30(2):134–141. doi:10.1093/fampra/cms053 Warmenhoven F, van Rijswijk E, Engels Y, Kan C, Prins J, van Weel C, et al. (2012) The Beck Depression Inventory (BDI-II) and a single screening question as screening tools for depressive disorder in Dutch advanced cancer patients. Support Care Cancer 20(2): 319–324. doi:10.1007/s00520-010-1082-8 Christensen KS, Bech P, Fink P (2010) Measuring mental health by questionnaires in primary care-unidimensionality, responsiveness and compliance. European Psychiatric Review 3(1):8–12 Keller SD, Ware JE Jr, Bentler PM, Aaronson NK, Alonso J, Apolone G, et al. (1998) Use of structural equation modeling to test the construct validity of the SF-36 Health Survey in ten countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 51(11):1179–1188 Kurtz ME, Kurtz JC, Given CW, Given BA (2004) Depression and physical health among family caregivers of geriatric patients with cancer—a longitudinal view. Med Sci Monit 10(8):CR447–CR456 Brogaard T, Neergaard MA, Guldin MB, Sokolowski I, Vedsted P (2013) Translation, adaptation and data quality of a Danish version of the Burden Scale for Family Caregivers. Scand J Caring Sci 27(4):1018–1026. doi:10.1111/j.1471-6712.2012.01092.x Grau H, Graessel E, Berth H (2015) The subjective burden of informal caregivers of persons with dementia: extended validation of the German language version of the Burden Scale for Family Caregivers (BSFC). Aging Ment Health 19(2):159–168. doi:10. 1080/13607863.2014.920296 Asgeirsdottir HG, Valdimarsdottir U, Furst CJ, Steineck G, Hauksdottir A (2013) Low preparedness before the loss of a wife to cancer and the widower’s chronic pain 4-5 years later-a population-based study. Psychooncology 22(12):2763–2770. doi: 10.1002/pon.3345 Bachner YG, Gesis-Za DE, Carmel S (2008) Caregivers’ communication with patients about illness and death: initial validation of a scale. Omega (Westport) 57(4):381–397
Support Care Cancer 33.
Statistics Denmark (2015). Quality declarations Available at: http:// www.dst.dk/en/Statistik/dokumentation/declarations. Accessed 05 May 2015 34. Gjerstorff ML (2011) The Danish Cancer Registry. Scand J Public Health 39(7 Suppl):42–45. doi:10.1177/1403494810393562 35. Lynge E, Sandegaard JL, Rebolj M (2011) The Danish National Patient Register. Scand J Public Health 39(7 Suppl):30–33. doi:10. 1177/1403494811401482 36. Sundararajan V, Quan H, Halfon P, Fushimi K, Luthi JC, Burnand B, et al. (2007) Cross-national comparative performance of three versions of the ICD-10 Charlson index. Med Care 45(12):1210– 1215 37. Thygesen SK, Christiansen CF, Christensen S, Lash TL, Sorensen HT (2011) The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the populationbased Danish National Registry of Patients. BMC Med Res Methodol 28 11:83. doi:10.1186/1471-2288-11-83 38. Ware JEJ, Gandek B, Kosinski M, Aaronson NK, Apolone G, Brazier J, et al. (1998) The equivalence of SF-36 summary health scores estimated using standard and country-specific algorithms in 10 countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 51(11):1167–1170 39. Li Q, Loke AY (2013) A spectrum of hidden morbidities among spousal caregivers for patients with cancer, and differences between the genders: a review of the literature. Eur J Oncol Nurs 17(5):578– 587. doi:10.1016/j.ejon.2013.01.007
40.
41.
42.
43. 44. 45.
46.
Ellervik C, Kvetny J, Christensen KS, Vestergaard M, Bech P (2014) Prevalence of depression, quality of life and antidepressant treatment in the Danish General Suburban Population Study. Nord J Psychiatry 29. doi:10.3109/08039488.2013.877074 Gandek B, Ware JE Jr (1998) Methods for validating and norming translations of health status questionnaires: the IQOLA Project approach. International Quality of Life Assessment. J Clin Epidemiol 51(11):953–959 Given B, Wyatt G, Given C, Sherwood P, Gift A, DeVoss D, et al. (2004) Burden and depression among caregivers of patients with cancer at the end of life. Oncol Nurs Forum, Nov 16 31(6):1105– 1117 Joanna Briggs Institute (2012) Caregiver burden of terminally-ill adults in the home setting. Nurs Health Sci 14(4):435–437 Frank L (2000) Epidemiology. When an entire country is a cohort. Science 31 287(5462):2398–2399 Guldin MB, Vedsted P, Zachariae R, Olesen F, Jensen AB (2012) Complicated grief and need for professional support in family caregivers of cancer patients in palliative care: a longitudinal cohort study. Support Care Cancer 20(8):1679–1685. doi:10.1007/ s00520-011-1260-3 Statistics Denmark (2015). Dødsfald og middellevetid [Deaths and average lifespan]. Available at: http://www.dst.dk/da/ Statistik/emner/doedsfald-og-middellevetid/doedsfald.aspx. Accessed 05 May 2015.
