Evidence profile: caregiver support

Transcription

1 Integrated care for older people (ICOPE) Guidelines on community-level interventions to manage declines in intrinsic capacity Evidence profile: caregiver support Scoping question: Does respite care or psychosocial support produce any benefit or harm for family caregivers of care-dependent older people? The full ICOPE guidelines and complete set of evidence profiles are available at who.int/ageing/publications/guidelines-icope Painting: Wet in Wet by Gusta van der Meer. At 75 years of age, Gusta has an artistic style that is fresh, distinctive and vibrant. A long-time lover of art, she finds that dementia is no barrier to her artistic expression. Appreciated not just for her art but also for the support and encouragement she gives to other artists with dementia, Gusta participates in a weekly art class. Copyright by Gusta van der Meer. All rights reserved

2 Evidence profile: caregiver support Contents Background... 1 Part 1: Evidence review... 2 Scoping question in PICO format (population, intervention, comparison, outcome)... 2 Search strategy... 3 List of systematic reviews identified by the search process... 3 PICO table... 4 Narrative description of the studies that went into analysis... 5 GRADE table 1: Respite care compared with usual care for family caregivers of care-dependent older people... 8 GRADE table 2: Respite care compared with usual care for informal caregivers of care-dependent older people... 9 GRADE table 3: Respite care compared with no intervention for carers of older people GRADE table 4: Psychosocial interventions compared with no intervention for caregivers of older people GRADE table 5: Psychotherapy compared with no intervention for caregivers of older people GRADE table 6: Support interventions compared with no intervention for carers of older people GRADE table 7: Training of care recipient compared with no intervention GRADE table 8: Multicomponent interventions compared with no intervention for carers of older people Additional evidence not mentioned in GRADE tables Part 2: From evidence to recommendations Summary of evidence Evidence-to-recommendation table Guideline development group recommendation and remarks References Annex 1: Search terms Annex 2: PRISMA flow diagram for systematic reviews of reviews World Health Organization 2017 Some rights reserved. This work is available under the Creative Commons Attribution- NonCommercial-ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO;

3 1 Evidence profile: caregiver support Background Worldwide, 349 million people are estimated to be care-dependent, of whom 18 million (5%) are children aged under the age of 15 years, and 101 million (29%) are older people 60 years of age and over. Care-dependence is defined as the need for frequent human help or care beyond that habitually required by a healthy adult. In older people, coexisting chronic diseases (multimorbidity) is frequently associated with the need for health and social care (1). In most countries, care for older people is provided by informal caregivers (including spouses, adult offspring and other relatives or friends), and the majority of primary caregivers are women (2). Evidence shows that caregivers of people with severe declines in capacity are at high risk of experiencing psychological distress and depression (3). In many low- and middle-income countries, formal systems of long-term care are poorly developed; the negative effects of caregiving therefore have a strong impact on the physical, emotional and economic status of family caregivers (2). In the past two decades, psychosocial interventions to support informal caregivers have been extensively studied. Psychosocial support includes different types of service provision, such as psychoeducational, counselling, skill-building and information or emotional support, which may be provided through agency-based settings or in the carer s home (4). These interventions focus on improving the carer s ability to manage everyday caregiving tasks (4). Furthermore, recent research has explored technologybased interventions, including the use of telephone and computer services, to provide adequate support and education to caregivers, with accessibility being a key advantage (5). Another popular intervention is respite care, defined as the provision of a temporary break in caregiving activities for the informal carer aimed at reducing distress and improving the carer s well-being (6). Respite can be delivered in different ways, including in-home services, adult day care or in institutions, such as care homes or hospices (7). Evidence from earlier reviews suggest that most of caregiver interventions were from high-income countries and were largely dominated by a range of programmes and services developed to assist caregivers of people with dementia. Therefore, the effectiveness of these interventions in different social, cultural and geographical contexts is unclear (2). Of all the chronic diseases, dementia is a particularly important contributor to caregiver strain. However, the extent to which these interventions can be administered to informal caregivers of care-dependent older people is unknown.

4 2 Evidence profile: caregiver support Part 1: Evidence review Scoping question in PICO format (population, intervention, comparison, outcome) Population Family caregivers (both male and female) of care-dependent older people of 60 years of age and over Interventions Respite care Psychosocial support Technology-based interventions Comparisons Usual or standard care Waiting list control Active control intervention Outcomes Critical: Caregiver burden, caregiver depression, care recipients symptoms Important: Well-being, ability/knowledge, quality of life, anger, anxiety

5 3 Evidence profile: caregiver support Search strategy The search for studies using combined intervention terms was conducted on October 2015 in Ovid MEDLINE (see Annex 1). The Cochrane Database of Systematic Reviews and Cochrane Central Register of Controlled Clinical Trials databases were searched, using combinations of the following terms: ("caregivers"[mesh Terms] OR "caregivers"[all Fields] OR "caregiver"[all Fields]) AND support[all Fields]) OR respite[all Fields] OR (psychosocial[all Fields] AND interventions[all Fields])) AND interventions[all Fields] AND ("frail elderly"[mesh Terms] OR ("frail"[all Fields] AND "elderly"[all Fields]) OR "frail elderly"[all Fields] OR ("frail"[all Fields] AND "older"[all Fields] AND "adults"[all Fields]) OR "frail older adults"[all Fields]) (caregiver support OR respite care OR psychosocial intervention) AND frail*. List of systematic reviews identified by search Included in GRADE 1 tables (8 10) Mason A, Weatherly H, Spilsbury K, Golder S, Arksey H, Adamson J et al. The effectiveness and cost-effectiveness of respite for caregivers of frail older people. J Am Geriatr Soc. 2007;55(2): doi: /j x. [Review was updated by WHO in 2015]. Shaw C, McNamara R, Abrams K, Cannings-John R, Hood K, Longo M et al. Systematic review of respite care in the frail elderly. Health Technol Assess. 2009;13(20):1 224, iii. doi: /hta [Review was updated by WHO in 2015]. Sorensen S, Pinquart M, Duberstein P. How effective are 1 GRADE: Grading of Recommendations Assessment, Development and Evaluation. More information: interventions with caregivers? An updated meta-analysis. Gerontologist. 2002;42(3): [Review was updated by WHO in 2015]. Pinquart M, Sorensen S. Helping caregivers of persons with dementia: which interventions work and how large are their effects? Int Psychogeriatr. 2006;18(4): doi: /S [Review updated by WHO in 2015]. Excluded from GRADE tables and footnotes (12, 13) Lopez-Hartmann M, Wens J, Verhoeven V, Remmen R. The effect of caregiver support interventions for informal caregivers of community-dwelling frail elderly: a systematic review. Int J Integr Care. 2012;12:e133. (Reason: eligible studies included were cited in the included reviews) Cassie KM, Sanders S. Familial caregivers of older adults. J Gerontol Soc Work. 2008;50(Suppl 1): doi: / (Reason: eligible studies included were cited in the included reviews)

6 4 Evidence profile: caregiver support PICO table Intervention/ comparison Outcomes Systematic reviews used for GRADE Explanation 1 Respite care vs usual care or waiting list control Reduction in caregiver burden Caregiver depression Mason A, Weatherly H, Spilsbury K, Golder S, Arksey H, Adamson J et al. The effectiveness and costeffectiveness of respite for caregivers of frail older people. J Am Geriatr Soc. 2007;55(2): (8) Systematic review relevant to the area GRADE table 1 2 Respite care vs usual care or waiting list control Reduction in caregiver burden Caregiver depression Caregiver anxiety Caregiver anger Shaw C, McNamara R, Abrams K, Cannings-John R, Hood K, Longo M et al. Systematic review of respite care in the frail elderly. Health Technol Assess. 2009;13(20): (9) Systematic review relevant to the area GRADE table 2 3 Respite care vs no intervention Psychosocial interventions vs no intervention Reduction in caregiver burden Caregiver depression Subjective well-being Ability/knowledge Care recipients symptoms Sorensen S, Pinquart M, Duberstein P. How effective are interventions with caregivers? An updated metaanalysis. Gerontologist. 2002;42(3): (10) GRADE tables 3 to 7 Systematic review relevant to the area

7 5 Evidence profile: caregiver support Narrative description of the studies that went into analysis Respite care and psychosocial interventions GRADE table 1 Mason et al. conducted a systematic review and meta-analysis to examine the effectiveness and cost-effectiveness of respite care for caregivers of frail older people (8). An extensive literature search was conducted and relevant studies were identified and assessed for methodological quality by two of the authors. A total of 22 studies were included: 10 randomized controlled (RCTs), seven quasi-experimental studies and five uncontrolled studies. Noticeably, even though the search was not restricted to a particular disease, most of the studies included older people with cognitive impairment exclusively (n = 13). Although physical impairment was also described as a common condition, it was inconsistently reported. Moreover, different types of respite care were covered across the different including adult day care, multidimensional packages, respite packages, in-home respite, host-family respite, institutional respite and video respite. Caregiver burden and caregiver depression were the two main outcomes measured. Pooled estimates obtained from four RCTs and four quasi-experimental studies (N = 989) assessing respite package (n = 1), in-home respite (n = 2) and adult day care (n = 5) found no statistically significant effect of respite on caregiver burden (SMD: 0.15, 95% CI: to 0.05). On the other hand, pooled estimates from one RCT and two quasi-experimental studies (N = 295) examining caregiver depression showed a statistically significant positive effect (SMD: 0.32, 95% CI: to -0.02).This overall beneficial effect was attributed mainly to results in one trial on day care, however, which undermines the reliability of the pooled results. The authors concluded that there is evidence suggesting that respite for caregivers of frail older people may have a small positive effect reducing caregiver depression and burden (although the latter effect was not significant when including RCTs in the analysis). They found no reliable evidence that respite care may delay institutionalization or may be more cost-effective than usual care. However, it should be noted that several limitations have been reported by the authors regarding the methodological quality of the studies and the variability of relative effects based on structural differences in the interventions provided. GRADE table 2 Shaw et al. conducted a systematic review and meta-analysis to assess the effectiveness of respite care on the well-being of informal carers of frail and disabled older people living in the community (9). An extensive search was conducted, including qualitative studies, and methodological quality was assessed by two authors independently. From a total of 104 quantitative studies selected, 16 were included in the meta-analysis (9 RCTs and quasi-experimental studies and seven longitudinal before-and-after studies). All the studies were conducted in high-income countries, the majority assessing day care and mixed respite care interventions, while some assessed in-home care and institutional care. Overall, care recipients included frail older people or older adults with dementia or experiencing mixed problems. Two RCTs and a quasi-experimental study assessing day care respite showed (continued next page)

8 6 Evidence profile: caregiver support no significant effects of respite care on caregiver burden after a six-month follow-up period (SMD: 0.11, 95% CI: to 0.17). However, even though studies were rated as moderate to high quality, sampling characteristics of one of the were not generalizable to the carer population and another study presented limitations as the uptake of respite was low in the sample. In addition, caregiver depression was assessed as a primary outcome. Due to significant heterogeneity in the studies, random effect model results were used and these are presented in the GRADE tables. Authors found no significant results in favour of day care respite to address carer depression either at short-term (six months) or long-term follow-up (12 months) (SMD: -0.23, 95% CI: to 0.03; and SMD: -0.08, 95% CI: to 0.24, respectively). Although the overall quality of studies in this group was moderate, authors remarked that the low level of respite provision in two of the studies might explain the intervention s low impact. Further analysis revealed no significant benefits of respite in terms of caregiver anxiety (SMD: 0.27, 95% CI: to 0.82) and singlegroup studies indicated that carer s quality of life was worse after respite use. However, pooled estimates derived from one moderate-quality RCT and one quasi-experimental study providing day care showed that anger was significantly lower after three months post-intervention in the intervention group (SMD: -0.38, 95% CI: to -0.17). GRADE tables 3 to 7 Sorensen et al. conducted a meta-analytic study to investigate the effectiveness of interventions for informal carers of older adults (10). The authors included a total of 78 studies in order to explore six different interventions (psychoeducation, supportive interventions, psychotherapy, respite care, training of care recipient and multicomponent interventions) and six outcome variables (carer s burden, depression, subjective wellbeing, ability/knowledge and care recipient s symptoms). The number of sessions of the interventions ranged from 1 to 180 (median = 8 sessions) with follow-up assessments conducted in only 22% of the cases after an average of seven months (SD = 5.1 months). The number of carers receiving interventions ranged from 4 to 2268 (mean = 24) with a mean age of 62.3 years for carers and a mean age of 77.3 years for care recipients. Nearly 60% of the studies explored group treatments, 22% individual interventions, 18% combined (group and individual) and in 1% this was not reported. Attrition rates varied from over 35% for respite care to 11.7% for psychotherapy. Moreover, almost 70% of the caregivers were female, near 77% of the carers lived with the care receiver and 50% were spouses. Noticeably, more than 60% of the studies included by the authors focused exclusively on caregivers for people with dementia. Also, most of the other heterogeneous samples included people with dementia along with people with other physical or mental disabilities/disorders. All of the studies were conducted in high-income countries. When analysing the effect of the different interventions including only RCTs, the authors found that psychotherapy was significantly beneficial for all the outcomes measured (carer s burden SMD: -0.22, 95% CI: to -0.03, n = 8; depression SMD: -0.27, 95% CI: to -0.09, n = 9; well-being SMD: 0.52, 95% CI: 0.08 to 0.96, n = 2; ability/knowledge SMD: 0.38, 95% CI: 0.14 to 0.61, n = 4). Psychoeducation for carers showed significant effects for improving caregiver (continued next page)

9 7 Evidence profile: caregiver support burden (n = 21), depression (n = 15) and ability/knowledge (n = 19) (SMD: -0.12, 95% CI: to -0.01; SMD: -0.23, 95% CI: to -0.08; and SMD: 0.37, 95% CI: 0.24 to 0.51, respectively). Supportive interventions (including professionally and peer-led support and discussion groups) revealed positive effects for caregiver burden (SMD: -0.35, 95% CI: to -0.10, n = 4) and ability/knowledge (SMD: 0.54, 95% CI: 0.30 to 0.78, n = 5). Regarding these results, the authors suggested that the small number of RCTs examining respite care may have reduced the chance of finding significant effects for this intervention. Moreover, significant positive effects on care recipient s symptoms were observed after training and supportive interventions. Furthermore, in this study, a regression model was used to compare all the studies in which care recipients had dementia exclusively versus those studies with no dementia patients or in which they only constituted a proportion of the total sample. After conducting the analysis, authors found that when all care-receivers in the sample had dementia, interventions were less effective at improving caregiver s burden, depression, well-being and ability/knowledge. Moreover, authors identified several limitations in relation to missing data for relevant outcomes in several studies (e.g. drop-out rates), and in relation to the variability of intervention effects observed according to the different outcome measurements (particularly for depression and burden). The authors concluded that psychotherapeutic interventions and psychoeducation for frail older people s caregivers showed the most consistent effects on the outcomes measures. Pinquart et al. replicated the systematic review but focused exclusively on the effectiveness of interventions for dementia caregivers (11). In addition to the interventions analysed in their previous article (10), the authors also included measures such as: (i) counselling/case management, (ii) psychoeducation stratified into active participation or information provision only and (iii) structured and unstructured multicomponent approaches. Furthermore, authors added institutionalization as a main outcome measure. A detailed analysis of the evidence reported in this study is presented in the WHO mhgap guidelines on interventions for carers of people with dementia (14). Overall, when evidence from both studies is compared (10, 11), similar findings can be observed showing that psychoeducation is significantly beneficial for all the outcomes across both studies and, similarly, psychotherapy (e.g. cognitive behavioural therapy [CBT]) had significant effects in terms of reducing caregiver burden and depression in all caregivers. However, while respite care showed significant benefits for improving burden, depression and subjective well-being (SWB) for dementia caregivers (SMD: -0.26, 95% CI: to -0.12; SMD: -0.12, 95% CI: to 0; SMD: 0.27, 95% CI: 0.03 to 0.51, respectively) (11), no positive effects remained when only RCTs were analysed for older people s carers (10).

10 8 Evidence profile: caregiver support GRADE table 1: Respite care compared with usual care for family caregivers of care-dependent older people Author: WHO systematic review team Date: 20 October 2015 Question: What is the effectiveness of respite care compared with usual care for family caregivers of care-dependent older people? Setting: Community Bibliography (8): Mason A, Weatherly H, Spilsbury K, Golder S, Arksey H, Adamson J et al. The effectiveness and cost-effectiveness of respite for caregivers of frail older people. J Am Geriatr Soc. 2007;55(2): [Review updated by WHO in 2015] Quality assessment Number of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Respite care Usual care Absolute (95% CI) Quality Importance Caregiver burden (follow-up 2 12 months; assessed with Caregiver Burden Inventory, Caregiver Strain Index, Zarit Burden Interview; lower score = better performance) 8 randomized very a not b not none see the footnote c see the SMD 0.15 footnote c lower (0.36 lower to 0.05 Caregiver depression (follow-up 12 months; assessed with Center for Epidemiological Studies Depression Scale [CES-D] and Hamilton Depression Scale; lower score = better performance) 3 randomized d not b not none see the footnote c see the footnote c SMD 0.32 lower (0.62 lower to 0.02 lower) MD: mean difference; RR: relative risk. a. Risk of bias: Downgraded twice as four included that had a quasi-experimental design and the method of allocation concealment was inadequate in six. b. Risk of bias: Downgraded once as were mainly from high-income countries, and data therefore may not be generalizable to resource-poor settings. c. Number of participants recruited for each group was not reported. d. Risk of bias: Downgraded once as two included applied quasi-experimental design.

11 9 Evidence profile: caregiver support GRADE table 2: Respite care compared with usual care for informal caregivers of care-dependent older people Author: WHO systematic review team Date: 20 October 2015 Question: What is the effectiveness of respite care compared with usual care for informal caregivers of care-dependent older people? Setting: Community Bibliography (9): Shaw C, McNamara R, Abrams K, Cannings-John R, Hood K, Longo M et al. Systematic review of respite care in the frail elderly. Health Technol Assess. 2009;13(20): [Review updated by WHO in 2015]. Quality assessment Number of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Respite care Usual care Absolute (95% CI) Quality Importance Caregiving burden (follow-up 3 12 months; assessed with CBI, PDS, ZBI; lower score = better performance) 3 randomized a b c not none SMD 0.11 lower (0.38 lower to 0.17 Caregiver depression (follow-up 2 weeks to 6 months; assessed with BSI, CES-D, GHQ; lower score = better performance) 5 randomized d e c not none see the footnote f see the footnote f SMD 0.23 lower (0.49 lower to 0.03 (continued next page)

12 10 Evidence profile: caregiver support Caregiver depression (follow-up 12 months; assessed with HADS, CES-D, PDS; lower score = better performance) 3 randomized g h c not none SMD 0.08 lower (0.41 lower to 0.24 Caregiver anxiety (follow-up 12 months; assessed with HADS, PDS; lower score = better performance) 2 randomized a i c not none SMD 0.27 higher (0.28 lower to 0.82 IMPORTANT Caregiver anger (follow-up 3 months; assessed with BSI, CES-D; lower score = better performance) 2 randomized a not c not none SMD 0.38 lower (0.6 lower to 0.17 lower) LOW IMPORTANT BSI: Brief Symptoms Inventory; CBI: Caregiver Burden Inventory; CES-D: Center for Epidemiological Studies Depression Scale; GHQ: General Health Questionnaire; HADS: Hospital Anxiety and Depression Scale; PDS: Psychological Distress Scale; SMD: standardized mean difference; ZBI: Zarit Burden Interview. a. Risk of bias: Downgraded once as one of the included trial applied a quasi-experimental design. b. Inconsistency: Downgraded once as moderate heterogeneity was observed for pooled estimate (Chi 2 = 5.02 [df = 2], P = 0.081, I 2 = 60.2%). c. Indirectness: Downgraded once as all included were conducted in high-income countries, therefore evidence may not be generalizable to resource-poor settings. d. Risk of bias: Downgraded once as two included applied a quasi-experimental design. e. Inconsistency: Downgraded once as moderate heterogeneity was observed for pooled estimate (Chi [df = 4], P = 0.039, I 2 = 60.4%). f. Number of participants recruited for each group was not reported separately (total sample size was more than 400). g. Risk of bias: Downgraded once as one of the included applied quasi-experimental design. h. Inconsistency: Downgraded once as moderate heterogeneity was observed (Chi 2 = 6.18 [df = 2], P = 0.045, I 2 = 67.7%). i. Inconsistency: Downgraded once as moderate heterogeneity was observed (Chi 2 = 3.64 [df = 1], P = 0.057, I 2 = 72.5%).

13 11 Evidence profile: caregiver support GRADE table 3: Respite care compared with no intervention for carers of older people Author: WHO systematic review team Date: 20 October 2015 Question: What is the effectiveness of respite care compared with no intervention for carers of older people? Setting: Community Bibliography (10): Sorensen S, Pinquart M, Duberstein P. How effective are interventions with caregivers? An updated meta-analysis. Gerontologist. 2002;42(3): [Review updated by WHO in 2015] Quality assessment Number of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Respite care No intervention Absolute (95% CI) Quality Importance Caregiver burden (follow-up 3 months; assessed with Zarit Burden Interview; lower score = better performance) 1 randomized a not b very c none 16 Footnote e MD 0.34 higher (0.36 lower to 1.04 Caregiver depression (follow-up 6 months; assessed with CES-D, BDI, BSI; lower score = better performance) 2 randomized a not b very c none 38 Footnote e SMD 0.29 lower (0.74 lower to 0.16 Subjective well-being (follow-up 3 6 months; assessed with ABS, LSS, PNES; lower score = better performance) 4 randomized a not b d none 156 Footnote e SMD 0.06 higher (0.16 lower to 0.28 IMPORTANT (continued next page)

14 12 Evidence profile: caregiver support Ability/knowledge (follow-up 3 6 months; assessed with questionnaires on knowledge, coping abilities or caregiving-related self-efficacy; higher score = better performance) 1 randomized a not b very c none 16 Footnote e MD 1.36 lower (2.14 lower to 0.58 lower) IMPORTANT Recipients symptoms (follow-up 3 months; assessed with BPC, IADLs; lower score = better performance) 2 randomized a not b d none 112 Footnote e SMD 0.12 higher (0.14 lower to 0.39 ABS: Affect Balance Scale; BDI: Beck Depression Inventory; BPC: Behavioural Problem Checklist; BSI: Depression Subscale of the Brief Symptom Inventory; CES-D: Center for Epidemiological Studies Depression Scale; IADLs: instrumental activities of daily living; LSS: Life Satisfaction Scales; MD: mean difference; PGCM: Philadelphia Geriatric Center Morale Scale; PANAS: Positive and Negative Affect Scale; SMD: standardized mean difference; ZBI: Zarit Burden Interview. a Risk of bias: Downgraded once as high average dropout rate was reported in most (> 35%). b Indirectness: Downgraded once as recruited unselected target population; caregivers for older people with and without dementia were included, therefore it is difficult to distinguish the effect. Included were from high-income countries, therefore data may not be generalizable to resource-poor settings. c Imprecision: Downgraded twice as sample size was very small (less than 100 participants). d Imprecision: Downgraded once as sample size was small. e Number of participants in control group was not reported.

15 13 Evidence profile: caregiver support GRADE table 4: Psychosocial interventions compared with no intervention for caregivers of older people Author: WHO systematic review team Date: 20 October 2015 Question: What is the effectiveness of psychoeducation compared with no intervention for caregivers of older people? Settings: Community and long-term care Bibliography (10): Sorensen S, Pinquart M, Duberstein P. How effective are interventions with caregivers? An updated meta-analysis. Gerontologist. 2002;42(3): [Review updated by WHO in 2015] Quality assessment No. of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Psychoeducation No intervention Absolute (95% CI) Quality Importance Caregiver burden (follow-up 3 12 months; assessed with ZBI, MBBS; lower score = better performance) 21 randomized not not a not none 582 Footnote c SMD 0.12 lower (0.24 lower to 0.01 lower) Caregiver depression (follow-up 6 12 months; assessed with BDI, CES-D, BSI; lower score = better performance) 15 randomized not not a not none 370 Footnote c SMD 0.23 lower (0.38 lower to 0.08 lower) (continued next page)

16 14 Evidence profile: caregiver support Subjective well-being (follow-up 6 months; assessed with ABS, LSS, PANAS; lower score = better performance) 3 randomized not not a b none 66 Footnote c SMD 0.25 lower (0.63 lower to 0.14 LOW IMPORTANT Caregiver ability/knowledge (follow-up 6 12 months; assessed with questionnaires on knowledge, coping abilities or caregiving-related self-efficacy; higher score = better performance) 19 randomized not not a not none 523 Footnote c SMD 0.37 higher (0.24 higher to 0.51 IMPORTANT Care-recipients symptoms (follow-up 3 12 months; assessed with BPC, IADL, ADL; lower score = better performance) 15 randomized not not a not none 508 Footnote c SMD 0.09 lower (0.22 lower to 0.03 ABS: Affect Balance Scale; ADLs: activities of daily living; BDI: Beck Depression Inventory; BPC: Behavioural Problem Checklist; BSI: Depression Subscale of the Brief Symptom Inventory; CES-D: Center for Epidemiological Studies Depression Scale; IADLs: instrumental activities of daily living; LSS: Life Satisfaction Scales; MBBS: Montgomery and Borgatta s Burden Scale; MD: mean difference; PGCM: Philadelphia Geriatric Center Morale Scale; PANAS: Positive and Negative Affect Scale; ZBI: Zarit Burden Interview. a Indirectness: Downgraded once as recruited unselected target population; caregivers for older people with and without dementia were included, therefore it is difficult to distinguish the effect. b Imprecision: Downgraded once as sample size was very small (smaller than 100 participants). c Number of participants in control group was not reported.

17 15 Evidence profile: caregiver support GRADE table 5: Psychotherapy compared with no intervention for caregivers of older people Author: WHO systematic review team Date: 20 October 2015 Question: What is the effectiveness of psychotherapy compared with no intervention for caregivers of older people? Settings: Community and long-term care Bibliography (10): Sorensen S, Pinquart M, Duberstein P. How effective are interventions with caregivers? An updated meta-analysis. Gerontologist. 2002;42(3): [Review updated by WHO in 2015] Quality assessment Number of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Psychotherapy No intervention Absolute (95% CI) Quality Importance Caregiver burden (follow-up 12 months; assessed with ZBI, MBBS; lower score = better performance) 8 randomized not not a not none 240 Footnote c SMD 0.22 lower (0.41 lower to 0.03 lower) Caregiver depression (follow-up 2 12 months; assessed with BDI, CES-D, BSI; lower score = better performance) 9 randomized not not a not none 271 Footnote c SMD 0.27 lower (0.45 lower to 0.09 lower) Subjective well-being (SWB) (follow-up 2 3 months; assessed with ABS, LSS; lower score = better performance) 2 randomized not not a very b none 58 Footnote c SMD 0.52 higher (0.08 higher to 0.96 IMPORTANT Ability/knowledge (follow-up 6 12 months; assessed with questionnaires on knowledge, coping abilities or caregiving-related self-efficacy; higher score = better performance) 4 randomized not not a not none 158 Footnote c SMD 0.38 higher (0.14 higher to 0.61 IMPORTANT (continued next page)

18 16 Evidence profile: caregiver support Quality assessment Number of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Psychotherapy No intervention Absolute (95% CI) Quality Importance Recipients symptoms (follow-up 2 12 months; assessed with BPC, IADLs, ADLs; lower score = better performance) 7 randomized not not a not none 228 Footnote c SMD 0.19 lower (0.38 lower to 0 lower) ABS: Affect Balance Scale; ADLs: activities of daily living; BDI: Beck Depression Inventory; BPC: Behavioural Problem Checklist; BSI: Depression Subscale of the Brief Symptom Inventory; CES-D: Center for Epidemiological Studies Depression Scale; IADLs: instrumental activities of daily living; LSS: Life Satisfaction Scales; MBBS: Montgomery and Borgatta s Burden Scale; PGCM: Philadelphia Geriatric Center Morale Scale; PNES: Positive and Negative Affect Scale; SMD: standardized mean difference; ZBI: Zarit Burden Interview. a Indirectness: Downgraded once as recruited unselected target population; caregivers for older people with and without dementia were included, therefore it is difficult to distinguish the effect. Included were from high-income countries, therefore data may not be generalizable to resource-poor settings. b Imprecision: Downgraded twice as sample size was very small (less than 100 participants). c Number of participants recruited in the control group was not reported.

19 17 Evidence profile: caregiver support GRADE table 6: Support interventions compared with no intervention for carers of older adults Author: WHO systematic review team Date: 20 October 2015 Question: What is the effectiveness of support interventions compared with no intervention for carers of older adults? Settings: Community and long-term care Bibliography (10): Sorensen S, Pinquart M, Duberstein P. How effective are interventions with caregivers? An updated meta-analysis. Gerontologist. 2002;42(3): [Review updated by WHO in 2015] Quality assessment Number of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Support interventions No intervention Absolute (95% CI) Quality Importance Caregiver burden (follow-up 12 months; assessed with ZBI, MBBS; lower score = better performance) 4 randomized not not a not none 121 see the footnote d SMD 0.35 lower (0.6 lower to 0.1 lower) Caregiver depression (follow-up 4 12 months; assessed with BDI, CES-D, BSI; lower score = better performance) 5 randomized not not a b none 127 see the footnote d SMD 0.09 lower (0.33 lower to 0.16 LOW Subjective well-being (follow-up 3 months; assessed with ABS, LSS; lower score = better performance) 2 randomized not not a very c none 58 see the footnote d SMD 0.17 higher (0.18 lower to 0.52 IMPORTANT Ability/knowledge (follow-up 6 months; assessed with questionnaires on knowledge, coping abilities or caregiving-related self-efficacy; higher score = better performance) 5 randomized not not a not none 131 see the footnote d SMD 0.54 higher (0.3 higher to 0.78 IMPORTANT (continued next page)

20 18 Evidence profile: caregiver support Recipients symptoms (follow-up 12 months; assessed with BPC, ADLs; lower score = better performance) 2 randomized not not a b none 63 see the footnote d SMD 0.17 lower (0.52 lower to 0.18 LOW ABS: Affect Balance Scale; ADLs: activities of daily living; BDI: Beck Depression Inventory; BPC: Behavioural Problem Checklist; BSI: Depression Subscale of the Brief Symptom Inventory; CES-D: Center for Epidemiological Studies Depression Scale; IADLs: instrumental activities of daily living; LSS: Life Satisfaction Scales; MBBS: Montgomery and Borgatta s Burden Scale; PGCM: Philadelphia Geriatric Center Morale Scale; PANAS: Positive and Negative Affect Scale; SMD: standardized mean difference; ZBI: Zarit Burden Interview. a Indirectness: Downgraded once as recruited unselected target population; caregivers for older people with and without dementia were included, therefore it is difficult to distinguish the effect. Included were from high-income countries, therefore data may not be generalizable to resource-poor settings. b Imprecision: Downgraded once as sample size was small (less than 200). c Imprecision: Downgraded twice as sample size was very small. d Number of participants recruited in the control group was not reported.

21 19 Evidence profile: caregiver support GRADE table 7: Training of care recipient compared with no intervention Author: WHO systematic review team Date: 15 October 2015 Question: What is the effectiveness of training interventions for care recipients compared with no intervention or usual care? Settings: Community and long-term care Bibliography (10): Sorensen S, Pinquart M, Duberstein P. How effective are interventions with caregivers? An updated meta-analysis. Gerontologist. 2002;42(3): [Review updated by WHO in 2015] Quality assessment Number of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Training interventions No intervention Absolute (95% CI) Quality Importance Caregiver burden (follow-up 12 months; assessed with ZBI, MBBS; lower score = better performance) 5 randomized a not b c none 79 see the footnote e SMD 0.13 lower (0.46 lower to 0.2 VERY LOW Caregiver depression (follow up 4 12 months; assessed with BDI, CES-D, BSI; lower score = better performance) 4 randomized a not b very d none 56 see the footnote e SMD 0.27 lower (0.67 lower to 0.13 VERY LOW Subjective well-being (follow-up 3 months; assessed with ABS, LSS; lower score = better performance) 1 randomized a not b very d none 21 see the footnote e MD 0.74 higher (0.06 higher to 1.42 VERY LOW IMPORTANT (continued next page)

22 20 Evidence profile: caregiver support Ability/knowledge (follow-up 6 months; assessed with questionnaires on knowledge, coping abilities or caregiving-related self-efficacy; higher score = better performance) 2 randomized a 7 b very d none 44 see the footnote e SMD 0.16 lower (0.59 lower to 0.27 VERY LOW IMPORTANT Recipients symptoms (follow-up 12 months; assessed with BPC, IADLs, ADLs; lower score = better performance) 5 randomized a not b c none 79 see the footnote e SMD 0.51 lower (0.84 lower to 0.18 lower) VERY LOW ABS: Affect Balance Scale; ADLs: activities of daily living; BDI: Beck Depression Inventory; BPC: Behavioural Problem Checklist; BSI: Depression Subscale of the Brief Symptom Inventory; CES-D: Center for Epidemiological Studies Depression Scale; IADLs: instrumental activities of daily living; LSS: Life Satisfaction Scales; MBBS: Montgomery and Borgatta s Burden Scale; MD: mean difference; RR: relative risk; SMD: standardized mean difference; ZBI: Zarit Burden Interview. a Risk of bias: Downgraded once as high drop-out rate was reported in the included (over 30%). b Indirectness: Downgraded once as included were from high-income countries, therefore, data may not be generalizable to resourcepoor settings. Trials recruited unselected target population: both caregivers for older people with and without dementia were included, therefore, it is difficult to distinguish the effect. c Imprecision: Downgraded once as sample size was small (less than 200). d Imprecision: Downgraded twice once as sample size was small (less than 100). e Number of participants recruited in the control group was not reported.

23 21 Evidence profile: caregiver support GRADE table 8: Multicomponent interventions compared with no intervention for carers of older people Author: WHO systematic review team Date: 15 October 2015 Question: What is the effectiveness of multicomponent interventions compared with no intervention for carers of older people? Settings: Community and long-term care Bibliography (10): Sorensen S, Pinquart M, Duberstein P. How effective are interventions with caregivers? An updated meta-analysis. Gerontologist. 2002;42(3): [Review updated by WHO in 2015]. Quality assessment Number of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Multicomponent interventions No intervention Absolute (95% CI) Quality Importance Caregiver burden (follow-up 6 12 months; assessed with ZBI, MBBS; lower score = better performance) 4 randomized not a b not none 366 see the footnote e SMD 0.65 lower (0.84 lower to 0.46 lower) LOW Caregiver depression (follow up 3 12 months; assessed with BDI, CES-D, BSI; lower score = better performance) 4 randomized not not b not none 190 see the footnote e SMD 0.02 lower (0.22 lower to 0.18 Subjective well-being (follow-up 3 months; assessed with ABS, LSS; lower score = better performance) 2 randomized not not b very c none 42 see the footnote e SMD 0.78 higher (0.27 higher to 1.29 IMPORTANT Ability/knowledge (follow-up 6 months; assessed with questionnaires on knowledge, coping abilities or caregiving-related self-efficacy; higher score = better performance) 3 randomized not not b very c none 50 see the footnote e SMD 0.86 higher (0.42 higher to 1.31 IMPORTANT (continued next page)

24 22 Evidence profile: caregiver support Quality assessment Number of patients Effect Number of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Multicomponent interventions No intervention Absolute (95% CI) Quality Importance Recipients symptoms 2 randomized not not b d none 81 Not reported SMD 0.04 higher (0.53 lower to 0.61 LOW ABS: Affect Balance Scale; BDI: Beck Depression Inventory; BSI: Depression Subscale of the Brief Symptom Inventory; CES-D: Center for Epidemiological Studies Depression Scale; LSS: Life Satisfaction Scales; MBBS: Montgomery and Borgatta s Burden Scale; SMD: standardized mean difference; ZBI: Zarit Burden Interview. a Inconsistency: Downgraded once as Chi 2 test indicates heterogeneity (P = 0.05), but I 2 value was not reported. b Indirectness: Downgraded once as recruited unselected target population; caregivers for older people with and without dementia were included, therefore it is difficult to distinguish the effect. Included were from high-income countries, therefore data may not be generalizable to resource-poor settings. c Imprecision: Downgraded twice as sample size was very small (less than 100). d Imprecision: Downgraded once as sample size was small (less than 200).

25 23 Evidence profile: caregiver support Additional evidence not mentioned in GRADE tables Psychosocial interventions Yin et al. conducted a systematic review and meta-analysis to examine the effectiveness of group and individual interventions on reducing the burden experienced by carers of frail older people (15). A total of 26 studies were included (17 RCTs and nine quasi-experimental studies) with sample sizes ranging from 20 to 486 (total N = 1970). Caregiver mean age was 60.1 years, nearly 80% of participating caregivers were female and more than 80% reported living with the care recipient. On the other hand, the mean age for care recipients was 78.7 years and 54% of the total sample had a diagnosis of dementia. Eighteen studies explored group interventions of which more than 90% combined an educational component with a support programme. Duration of the interventions varied from two weeks to one year and 33% of the studies included up to six month of follow-up. Moreover, the eight studies examining individual interventions (N = 472) had a duration ranging from one single session to eight months follow-up. All studies included caregiver burden as a primary outcome measure and the Zarit Burden Interview (ZBI) was mostly employed as a measure scale. The authors calculated that the weighted mean effect size was 0.41 (95% CI: 0.32 to 0.51), showing a significant beneficial treatment effect for group interventions. Moreover, they found that the effect size remained significant when only RCTs were included (mean: 0.26, 95% CI: 0.15 to 0.37, n = 11). However, the homogeneity test was significant (Q17 = 56.37, P < ), which could indicate variations in the different study effect sizes attributable to study characteristics. Furthermore, similar results were observed for individual interventions where the weighted effect size was 0.48 (95% CI: 0.30 to 0.67), indicating a positive treatment effect. Nevertheless, the authors reported that most of the studies were flawed: six showed high risk of bias due to uncertainties around randomization and seven studies failed to report features significant to study outcomes which may jeopardize the generalizability of the results. Information and communication technology support There is very limited evidence regarding the effectiveness of technology-based interventions for caregivers who care for people exclusively with chronic diseases other than dementia. Klemm et al. conducted a randomized longitudinal study to evaluate whether psychosocial outcomes in family caregivers of people with chronic disease who participate in online support groups (OSGs) differ from those who are non-active participants in OSGs (16). A total of 86 caregivers were randomized into two types of OSGs (professionally facilitated/ psychoeducational or moderated/peer-directed OSGs). Nonactive participants were those who completed informed consent and questionnaires, but did not actively participate in the online discussions. The OSG programmes had a duration of 12 weeks, and outcomes were measured at baseline and at two later time points with outcomes including depression measured by the Center for Epidemiology Studies Depression Scale (CES-D), burden measured by the Caregiver Strain Index (CSI) and quality of life measured by the Caregiver Quality-of-Life Index (CQoL-I). The authors found that both professionally facilitated/psychoeducational and moderated/peer-directed OSGs reduced depressive symptoms and improved quality of life for caregivers who actively participated (P = CES-D; (continued next page)

26 24 Evidence profile: caregiver support P = CSI; CQoL-I). Moreover, they found no significant differences in the effectiveness of the two types of OSGs. However, generalizability of the results may be limited considering the sample characteristics (most caregivers were middle-aged white women who read and spoke English) and considering the fact that the intervention is limited to those who have access to a computer and the Internet and who have some degree of computer literacy. Moreover, risk of bias is unclear due to high attrition rates, mainly in the non-active participant group. Economic incentives were awarded to those who completed the study. The authors suggest that future caregiver research should target men and minorities. Evidence from technology-based interventions for dementia caregivers Mahoney et al. conducted a small RCT (n = 100) to examine the effectiveness of a computer-mediated automated interactive voice response (IVR) intervention for assisting caregivers of people with disruptive behaviours associated with Alzheimer s disease (17). Caregivers were randomized to a 12-month intervention group or a control group. IVR provided stress monitoring and counselling information, personal voic linkage to Alzheimer s disease experts, a telephone support group and a distraction call for care recipients. Outcomes including depression, anxiety, mastery and bothersome nature of caregiving were measured at baseline, 6, 12 and 18 months. Authors found a significant effect for caregivers with lower mastery at baseline on all the other outcomes bother (P = 0.04), anxiety (P = 0.01) and depression (P = 0.007) meaning this intervention could be particularly beneficial for caregivers with these characteristics. Moreover, another study based on the REACH programme (Resources for Enhancing Alzheimer s Caregiver Health) explored the efficacy of technology-based interventions and family therapy in reducing depression among dementia caregivers (18). A total of 225 (114 Cuban and 111 American) caregivers were randomized to receive a structural family therapy (SFT), a combined intervention (SFT plus computer telephone integrated system [CTIS]) or minimal support (control group). The CTIS system covered several aspects including: regular or conference phone calls with family or friends living far apart, monthly web-based telephone conferences with other caregivers, access to voic box, therapist appointment reminders and access to the Alzheimer s Association Resource Guide. Randomization was stratified according to the ethnicity of the caregiver and the REACH core battery of assessments (i.e. the CES-D, the Revised Memory and Behaviour Problems Checklist [RMBPC] to measure burden, activities of daily living [ADLs] scale, Mini- Mental State Examination [MMSE] and a Likert scale to measure global satisfaction) was administered at 6, 12 and 18 months. All measures were adequately translated to Spanish for Spanish-speaking participants. Authors found that at the six-month follow-up depressive symptoms in caregivers were significantly reduced among those who had received the combined intervention.

27 25 Evidence profile: caregiver support Part 2: From evidence to recommendations Summary of evidence Intervention Effect size Caregiver burden Caregiver depression Caregiver depression Caregiver anxiety Caregiver anger Respite care GRADE table 1 Mason et al. (8) SMD 0.15 lower (0.36 lower to 0.05 SMD 0.32 lower (0.62 lower to 0.02 lower) No data No data No data Respite care GRADE table 2 Shaw et al. (9) SMD 0.11 lower (0.38 lower to 0.17 SMD 0.23 lower (0.49 lower to 0.03 SMD 0.08 lower (0.41 lower to 0.24 SMD 0.27 higher (0.28 lower to 0.82 SMD 0.38 lower (0.6 lower to 0.17 lower) Respite care GRADE table 3 Sorenson et al. (10) SMD 0.34 higher (0.36 lower to 1.04 SMD 0.29 lower (0.74 lower to 0.16 SMD 0.06 higher (0.16 lower to 0.28 SMD 1.36 lower (2.14 lower to 0.58 lower) SMD 0.12 higher (0.14 lower to 0.39 Psychoeducation GRADE table 4 Sorenson et al. (10) SMD 0.12 lower (0.24 lower to 0.01 lower) SMD 0.23 lower (0.38 lower to 0.08 lower) SMD 0.25 lower (0.63 lower to 0.14 LOW SMD 0.37 higher (0.24 higher to 0.51 SMD 0.09 lower (0.22 lower to 0.03 (continued next page)

28 26 Evidence profile: caregiver support Intervention Caregiver burden Caregiver depression Caregiver subjective well-being Caregiver ability/ knowledge Care recipients symptoms Psychotherapy GRADE table 5 Sorenson et al. (10) SMD 0.22 lower (0.41 lower to 0.03 lower) SMD 0.27 lower (0.45 lower to 0.09 lower) SMD 0.52 higher (0.08 higher to 0.96 SMD 0.38 higher (0.14 higher to 0.61 SMD 0.19 lower (0.38 lower to 0 lower) Support GRADE table 6 Sorenson et al. (10) SMD 0.35 lower (0.6 lower to 0.1 lower) SMD 0.09 lower (0.33 lower to 0.16 LOW SMD 0.17 higher (0.18 lower to 0.52 SMD 0.54 higher (0.3 higher to 0.78 SMD 0.17 lower (0.52 lower to 0.18 LOW Training care recipients GRADE table 7 Sorenson et al. (10) SMD 0.13 lower (0.46 lower to 0.2 SMD 0.27 lower (0.67 lower to 0.13 SMD 0.74 higher (0.06 higher to 1.42 SMD 0.16 lower (0.59 lower to 0.27 SMD 0.51 lower (0.84 lower to 0.18 lower) Multicomponent GRADE table 8 Sorenson et al. (10) SMD 0.65 lower (0.84 lower to 0.46 lower) LOW SMD 0.02 lower (0.22 lower to 0.18 SMD 0.78 higher (0.27 higher to 1.29 SMD 0.86 higher (0.42 higher to 1.31 SMD 0.04 higher (0.53 lower to 0.61 LOW

29 27 Evidence profile: caregiver support Evidence-to-recommendation table Problem Is the problem a priority? Yes No Uncertain Benefits and harms Do the desirable effects outweigh the undesirable effects? Yes No Uncertain Explanation In most countries, care of older people is provided by informal caregivers (including spouses, adult offspring, other relatives and/or friends) and the majority of primary caregivers are women. Evidence shows that caregivers of people with severe declines in capacity are at high risk of experiencing psychological distress and depression. In many low- and middle-income counties, formal systems of long-term care are poorly developed; therefore, providing support to caregivers of care-dependent older people is important. Explanation There is sufficient very low- to moderate-quality evidence to suggest that psychosocial interventions produce small to moderate effects on critical outcomes, including: caregiver burden, depression, subjective well-being, ability or knowledge, and care-recipients symptoms. Noticeably, psychoeducational interventions and psychotherapy (mainly based on cognitive behavioural therapy [CBT] interventions) showed significant beneficial effects on all of the outcomes. Conversely, other interventions such as support and training of the care recipient were beneficial only for specific domains, and may be more effective as a parts of a multicomponent intervention. There is very low-quality evidence on the effectiveness of respite care interventions for caregivers who care for frail older people. Three showed that a respite care intervention brought significant benefit in terms of reducing caregiver depression (SMD: -0.32, 95% CI: to -0.02). However, no benefit was observed for other critical outcomes. A meta-analysis examining the effectiveness of respite care for carers of older adults showed positive results in terms of improving caregiver s burden, depression and subjective well-being. However, the results were not replicated when the analysis was limited to RCTs only. A significant favourable effect of respite care on reducing caregiver (continued next page)

30 28 Evidence profile: caregiver support (continued from previous page) Do the desirable effects outweigh the undesirable effects? Yes No Uncertain Values and preferences/acceptability Is there important uncertainty or variability about how much people value the options? Major variability Minor variability Is the option acceptable to key stakeholders? Major variability Minor variability Uncertain Uncertain depression was described in the report of another analysis. However, several methodological limitations were identified in the included. Information and communication technology support interventions, such as professionally facilitated/psychoeducational or moderated/peer-directed online support groups (OSGs) reduced depressive symptoms and improved the quality of life for carers of people with chronic diseases who actively participated in the OSGs. Moreover, computer-mediated automated interactive voice response (IVR), as well as a combined structural family therapy and computer telephone integrated system (CTIS) intervention, showed some beneficial effects. However, these findings are based on single RCTs conducted in high-income countries and aimed at caregivers of people with dementia. No harm was identified. Explanation Caregiving tasks for frail older people may cause impacts on the caregiver s health, representing an important source of burden. Evidence shows that there are available and effective interventions for improving the health of such caregivers, including psychotherapy and multicomponent interventions based upon psychoeducational interventions, support and training. Further research is needed to continue exploring and improving the quality of interventions. Most of the available studies analysed did not focus specifically on interventions for caregivers of frail older people; there was frequently a preponderance of caregivers of people with dementia in the samples. No subgroup analysis was conducted in most of the studies. Moreover, the lack of studies comparing interventions limits our understanding as to whether some interventions are more effective than others. When assessing the effectiveness of respite interventions, respite care was not found to significantly reduce the caregiver s burden, depression or anxiety. Further analysis including only RCTs showed similar results, as well as finding no significant effects of respite care on caregivers subjective well-being, ability and knowledge, or on the reduction of the care (continued next page)

31 29 Evidence profile: caregiver support (continued from previous page) Is the option acceptable to key stakeholders? Major variability Minor variability Feasibility/resource use Uncertain How large are the resource requirements? Major Minor Uncertain Is the option feasible to implement? Yes No Uncertain Equity Would the option improve equity in health? Yes No Uncertain recipients symptoms. However, positive results for reducing caregivers burden, depression and improving their subjective well-being were observed when the analysis was not limited to RCTs. Significant beneficial effects of respite care on reducing caregiver depression were described in another meta-analysis. However, the reliability of these results remains unclear as only one RCT was included in the analysis. Further research is required to explore the effectiveness of respite care interventions for caregivers of frail older people exclusively, as well as an appropriate subgroup analysis within the caregiver population and for each type of respite care service provided (i.e. day care, residential home, etc.). Explanation Most of the interventions evaluated in the included in the analysis were resource intensive. Psychoeducational interventions are a feasible approach for low-resource settings. These interventions can be conducted by non-specialist health-care providers with minimal training. Psychological interventions (e.g. CBT) are generally administered by trained specialists and require supervision. Moreover, the training time required and the total duration of the interventions might represent an additional burden for the caregivers. However, even though such approaches would be unlikely to be feasible in most low-resource settings, they could be adapted for use in these settings, e.g. brief interventions administered by trained and supported non-specialists. Considering that these interventions are frequently resource intensive, implementation of such approaches would be more difficult to achieve and would vary according to the resources available at each different location. Explanation The guideline development group strongly believed that this recommendation would increase equity in health.

32 30 Evidence profile: caregiver support Guideline development group recommendation and remarks Recommendation Psychological intervention, training and support should be offered to family members and other informal caregivers of care-dependent older people, particularly but not exclusively when the need for care is complex and extensive and/or there is significant caregiver strain. Quality of the evidence: Moderate Strength of the recommendation: Strong (No recommendation: For respite care, no recommendation could be made due to the uncertainty of evidence. The quality of the evidence was very low. Ten members of the guideline development group voted for no recommendation and four members voted in favour of a conditional recommendation on respite care.) Remarks Psychotherapeutic interventions (including counselling and/or cognitive behavioural therapy) should be offered when indicated, assuming that specialists accredited in the approaches or trained and supervised non-specialists are available to deliver the interventions. The evidence for information and communication technology (ICT) interventions (mainly) from the field of dementia suggests ICT technology as a potentially useful means, with increasing coverage due to the low cost. The vast majority of the primary studies available were carried out in high-income countries. Moreover, few available focused specifically on caregivers of older adults with chronic or disabling conditions other than dementia. Therefore, more research concentrating on the effectiveness of interventions for this particular group is required, as well as subgroup analyses within the caregiver population (children and other relatives, friends, neighbours, and so on). Further studies are also needed to explore the different types, content, duration and format of delivery of each intervention. Psychoeducational interventions delivered by trained non-specialists could represent a feasible and effective alternative, which could be beneficial to all health systems worldwide, given the increasing health demands in the context of ageing populations.

33 31 Evidence profile: caregiver support References 1. Fortin M, Soubhi H, Hudon C, Bayliss EA, van den Akker M. Multimorbidity's many challenges. BMJ : British Medical Journal. 2007;334(7602): International. WHOAsD. Supporting informal caregivers of people living with dementia Douglas SL, Daly BJ, Kelley CG, O Toole E, Montenegro H. IMpact of a disease management program upon caregivers of chronically critically ill patients*. Chest. 2005;128(6): Lopez-Hartmann M, Wens J, Verhoeven V, Remmen R. The effect of caregiver support interventions for informal caregivers of communitydwelling frail elderly: a systematic review. International Journal of Integrated Care. 2012;12:e Cassie KM, Sanders S. Familial caregivers of older adults. Journal of gerontological social work. 2008;50 Suppl 1: Hanson EJ, Tetley J, Clarke A. Respite care for frail older people and their family carers: concept analysis and user focus group findings of a pan-european nursing research project. Journal of advanced nursing. 1999;30(6): Shaw C, McNamara R, Abrams K, Cannings-John R, Hood K, Longo M, et al. Systematic review of respite care in the frail elderly. Health Technol Assess. 2009;13(20):1-224, iii. 8. Mason A, Weatherly H, Spilsbury K, Golder S, Arksey H, Adamson J et al. The effectiveness and cost-effectiveness of respite for caregivers of frail older people. J Am Geriatr Soc. 2007;55(2): doi: /j x. 9. Shaw C, McNamara R, Abrams K, Cannings-John R, Hood K, Longo M et al. Systematic review of respite care in the frail elderly. Health Technol Assess. 2009;13(20):1 224, iii. doi: /hta Sorensen S, Pinquart M, Duberstein P. How effective are interventions with caregivers? An updated meta-analysis. Gerontologist. 2002;42(3): Pinquart M, Sorensen S. Helping caregivers of persons with dementia: which interventions work and how large are their effects? Int Psychogeriatr. 2006;18(4): doi: /s Lopez-Hartmann M, Wens J, Verhoeven V, Remmen R. The effect of caregiver support interventions for informal caregivers of communitydwelling frail elderly: a systematic review. Int J Integr Care. 2012;12:e Cassie KM, Sanders S. Familial caregivers of older adults. J Gerontol Soc Work. 2008;50(Suppl 1): doi: / Interventions for carers of people with dementia. Geneva: World Health Organization; Yin T, Zhou Q, Bashford C. Burden on family members: caring for frail elderly: a meta-analysis of interventions. Nurs Res. 2002;51(3): Klemm PR, Hayes ER, Diefenbeck CA, Milcarek B. Online support for employed informal caregivers: psychosocial outcomes. Comput Inform Nurs. 2014;32(1): doi: /cin Mahoney DF, Tarlow BJ, Jones RN. Effects of an automated telephone support system on caregiver burden and anxiety: findings from the REACH for TLC Intervention Study. Gerontologist. 2003;43(4): , Eisdorfer C, Czaja SJ, Loewenstein DA, Rubert MP, Argüelles S, Mitrani VB, et al. The effect of a family therapy and technology-based intervention on caregiver depression. The Gerontologist. 2003;43(4):

34 32 Evidence profile: caregiver support Annex 1: Search terms MEDLINE database up to October respite$.af. 2. (community care$ or community nurs$).mp. 3. exp Community Health Services/or community health service$.mp. 4. (community health nurs$ or community mental health).mp. 5. Exp Community Psychiatry/or community psychiat$.mp. 6. (community healthcare$ or community health care$).mp. 7. (home care$ or home health care$).mp. 8. home nurs$.mp. 9. health service$ for the ag$.mp. 10. informal care$.mp. 11. (day centre$ or day center$).mp. 12. (day care or daycare$).mp. 13. night care$.mp. 14. (night sitt$ or night service$).mp. 15. domiciliary.mp. 16. (short break$ or break$ in car$).mp. 17. psychosocial.mp 18. psychotherapy.mp 19. support.mp 20. training.mp 21. old age assistance.mp. 22. temporary care$.mp. 23. exp Nursing Homes/or nursing home$.mp. 24. exp Residential Facilities/or residential facilit$.mp. 25. home$ for the ag$.mp. 26. (residential home$ or residential care$).mp. 27. (cancer care$ or oncologic$ care$).mp. 28. palliative$.mp. 29. exp Terminal Care/or terminal care$.mp. 30. terminal ill$.mp. 31. end of life.mp. 32. dying$.mp. 33. hospice$.mp. 34. exp Aged/or aged.mp. 35. exp Aging/or aging.mp. 36. ageing.mp. 37. old$.mp. 38. elder$.mp. 39. frail$.mp. 40. senior$.mp. 41. veteran$.mp. 42. (geriatric$ or gerontolog$).mp. 43. psychogeriatric$.mp. 44. (caregiver$ or care giver$).mp. 45. carer$.mp.

35 Included Eligible Identified Screened 33 Evidence profile: caregiver support Annex 2: PRISMA 2 flow diagram for systematic reviews of reviews Records identified through database searching (n = 214) Additional records identified through other sources (n = 7) Records after duplicates removed (n = 191) Records screened (n = 191) Records excluded (n = 182) Full-text articles assessed for eligibility (n = 9) Full-text articles excluded (n = 2) Not relevant: n = 2 Studies included in narrative synthesis (n = 3) Studies included in GRADE Table (n = 3) v 2 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). For more information: