Do interventions that include education on dementia progression improve knowledge,mental health and burden of family carers? A systematic review

Abstract

Background and aim

The European Association of Palliative Care recommends that family carers need education on the progression of dementia. This systematic review aimed to explore whether interventions incorporating education regarding the progressive nature of dementia increased carers’ understanding of dementia and improved mental health and burden.

Method

MEDLINE, PsycINFO and CINAHL were searched to April 2018. Randomised controlled trials with samples of family carers of someone with dementia were eligible. Included interventions involved a component aimed to increase the carer’s understanding of the progression of dementia. Outcomes of interest included: knowledge of dementia, depression, burden and pre-death grief.

Results

Searches identified 3221 unique citations of which 11 studies were eligible for review. Interventions ranged from 4 to 16 sessions of which 1 to 3 sessions focused on the progression of dementia. Knowledge: Two studies evaluated carers’ knowledge of dementia. One found no difference between the trial arms immediately after the intervention or three months later. The second found a significant intervention effect at the end of the intervention but not at three-month follow-up. Depression: Seven studies evaluated intervention effects on depression. Meta-analysis of three trials showed significant differences in mean follow-up scores favouring intervention over control. The remaining four studies did not show differences in depression between intervention and control groups. Burden: Nine studies evaluated burden and were examined in two meta-analyses (mean scores at follow-up and mean change scores from baseline to follow-up), neither of which found a benefit for intervention over control. Using the grading of recommendations assessment, development and evaluation system, we judged the quality of evidence to be very low for depression and low for burden, knowledge and pre-death grief, reducing our confidence in any of the effect estimates.

Conclusion

The evidence was not sufficient to support or refute the effectiveness of education on progression of dementia on carers’ knowledge and mental health.

Keywords

burden caregiver intervention carer intervention dementia progression depression education intervention pre-death grief systematic review

Introduction

Globally, an estimated 46.8 million people have dementia and this figure is expected to almost treble by 2050 (Alzheimer’s Disease International, 2015). Deaths due to Alzheimer’s disease and other dementias more than doubled between 2000 and 2015 making it the seventh leading cause of death worldwide (World Health Organization, 2017). Most people, particularly in the early stages of dementia, live at home (Wimo, Jonsson, Bond, Prince, & Winblad, 2013) and are cared for by family members or friends (hereafter referred to as ‘carers’) estimated to provide the equivalent of more than 40 million full-time workers worldwide (Wimo, Gauthier, & Prince, 2018). However, one study found only 43% of 161 family carers of nursing home residents considered dementia a disease you can die from (van der Steen, Onwuteaka-Philipsen, Knol, Ribbe, & Deliens, 2013).

Being a carer is associated with a high level of burden (Abreu, Tolson, Jackson, & Costa, 2018) and dementia carers have a greater risk of having a depressive disorder (Cuijpers, 2005). Grief is also common among dementia carers, with between 47% and 71% of carers experiencing grief before the death of the person with dementia as they experience losses in their roles and the relationship with the person with dementia (Chan, Livingston, Jones, & Sampson, 2013). This ‘pre-death grief’ has been defined as the carers “emotional and physical response to the perceived losses in a valued care recipient.” (p. 2203) (Lindauer & Harvath, 2014).

Whilst the death of a relative can sometimes be a relief for carers of people with dementia, around 20% will go on to experience prolonged grief disorder after death (Chan et al., 2013). Feeling unprepared for end of life is associated with prolonged grief disorder, depression and anxiety (Barry, Kasl, & Prigerson, 2002; Hebert, Dang, & Schulz, 2006). Feeling prepared for end of life is multifaceted but a key element is having an understanding of the prognosis and recognising symptoms of decline and having good communication with healthcare providers to address any concerns and questions (Durepos et al., 2018; Hebert, Prigerson, Schulz, & Arnold, 2006). Schulz, Boerner, Shear, Zhang, and Gitlin (2006) showed that interventions that provide education, skills training and support groups for carers of people with dementia reduced their depression and level of burden. These interventions were also found to have a knock-on effect on the level of grief among carers after the death of their loved ones.

Increasing carers’ understanding of prognosis of dementia could also be beneficial to people dying with dementia, as carers’ with more extensive understanding of the disease progression tend to choose less aggressive care (Mitchell et al., 2009), and it also predicts care recipients’ comfort when dying (van der Steen et al., 2013). The importance of education about the progressive and terminal nature of dementia therefore appears to be an important element in supporting family carers whilst they are caring for someone with dementia. This is also reflected in the European Association of Palliative Care’s white paper on optimal palliative care in dementia where they recommend that education on the progression of dementia should be provided to carers alongside treatment options (van der Steen et al., 2014). However, knowledge of the life-limiting course of dementia is an area that is less well understood among nurses, care staff and family carers of those with advanced dementia (Robinson et al., 2014).

Clinical guidelines and studies highlight the importance of increasing carers’ knowledge and understanding of dementia as a progressive disease you can die from. Improving carers’ knowledge of dementia should increase their preparedness for end of life (Steinhauser et al., 2001) which we anticipate will consequently benefit their mental health. Although there has been evidence that educational interventions that aimed to improve carers’ caregiving skills resulted in a small to moderate effect on depression and burden (Jensen, Agbata, Canavan, & McCarthy, 2015), no systematic review to date has explored the evidence on the effect of education about the progression of dementia on carers’ mental health.

Aim

The aim of this review was to explore whether interventions that incorporated education regarding the progressive nature of dementia could increase carers’ understanding of dementia, and in turn, could reduce their depression, burden and pre-death grief.

Method

Inclusion criteria

Randomised controlled trials (RCT) in any setting.

Participants who were family carers defined as non-paid, non-professional carers who provide care for a relative or friend with dementia of any type and severity.

Studies where the intervention of interest involved an educational intervention that aimed to improve carers’ knowledge and understanding of the progression of dementia. Multi-component interventions involving for instance psychotherapeutic support were also included; however, there must have been a specific educational component that focused on progression of dementia.

Interventions delivered in individual or group format.

Studies that assessed at least one of the outcomes of interest: knowledge of dementia, depression, burden and pre-death grief.

Exclusion criteria

Studies not written in English.

Studies where the control group also received education on the progression of dementia.

Outcomes of interest

The primary outcomes of interest were knowledge of dementia, depression, burden and pre-death grief. These may be captured by validated scales such as the Center for Epidemiologic Studies Depression scale (Radloff, 1977) for depression, the Zarit Burden Interview (Zarit, Reever, & Bach-Peterson, 1980) for carer burden or the Marwit-Meuser Caregiver Grief Inventory (MM-CGI) (Marwit & Meuser, 2002) for pre-death grief. We were interested in impact over time and consider post-intervention follow-up and longer term follow-up, anticipating that most would be within a six-month period from baseline assessment.

Search strategy

We searched MEDLINE, PsycINFO and CINAHL to April 2018. The search strategy included terms associated with dementia, intervention and study design as shown in Table 1.

Table 1.

Search terms by database.

Database	Search terms
Ovid PsycINFO	exp Dementia/ or dement* or alzheimer* or (frontotemporal* or FTD or FTLD) or (lew* adj2 bod) caregiver or carer* or famil* or relatives or kin* or spouse* (1 and 2) education or knowledge or information or teach* or train* or health promotion or booklet* or leaflet* (3 and 4) exp Intervention/ or exp Clinical Trials/ or randomly or (randomised or randomized or RCT or trial) or “double-blind” or “single blind” (5 and 6)
Cinahl EBSCO	(MH “Dementia”) or TX dement* or TX alzheimer* or TX “lew* bod” or TX (FTLD or FTD or frontotemporal) TX caregiver* or TX carer* or TX famil* or TX relatives or TX kin* or TX spouse* (1 and 2) TX education or TX knowledge or TX information or TX teach* or TX train* or TX health promotion or TX leaflet* or TX booklet* (3 and 4) (MH “Randomized Controlled Trials”) or TX randomised or TX randomized or AB randomly or AB “double blind” or AB “single blind” or AB RCT (5 and 6)
Ovid Medline	exp Dementia/ or dement* or alzheimer* or (lewy* adj2 bod) or (chronic adj2 cerebrovascular) or (“organic brain disease” or “organic brain syndrome”) or (cerebr adj2 deteriorat) or (cerebral adj2 insufficient) or (pick adj2 disease) or (creutzfeldt or JCD or CJD) or binswanger* caregiver* or carer* or family* or relatives or kin* or spouse* (1 and 2) education or knowledge or information or teach* or train* or health promotion or booklet* or leaflet* (3 and 4) randomized controlled trial or controlled clinical trial or random$ or groups or RCT or intervention (5 and 6)

Asterisks (*) indicate truncated or wildcard search. FTD: frontotemporal dementia; FTLD: frontotemporal lobar degeneration; adj2: adjacent to; RCT: randomised controlled trial; MH: MeSH heading (medical subject headings); CJD: Creutzfeldt-Jakob disease; JCD: Jakob-Creutzfeldt disease; TX: full text, AB: abstract.

Data collection and analysis

Selection of studies

Two review authors (CYL, KJM) independently screened for inclusion the citations retrieved. Full-text versions of citations that were classified as relevant by either author were retrieved for definitive assessment of eligibility. Any disagreement regarding inclusion was resolved through discussion, or where necessary, with reference to a third author (BC).

Data extraction and management

Key data were extracted by one author (CYL) using a data extraction form and was verified by the second author (KJM). Outcome data as appropriate were entered into Review Manager 5 for meta-analysis.

Assessment of risk of bias of included studies

As only RCTs were included, the Cochrane risk of bias tool (Higgins & Green, 2011) was used to assess the quality of the included studies. Two authors (CYL, BC) independently assessed the risk of bias, any disagreements were resolved through discussion, or where necessary, with reference to another author (KM). Five items were considered as reported in Table 2. Each parameter was classified as either low risk, unclear risk or high risk of bias.

Table 2.

Assessment of bias.

Bias	How bias was assessed.
Selection bias: random sequence generation and allocation concealment	We assessed whether the allocation sequence to trial arms was generated randomly. For allocation concealment, we assessed whether the participants’ allocation to treatment groups could be foreseen before assignment.
Detection bias	We attempted to ascertain whether trial outcome assessors were blinded to participants’ treatment allocation.
Attrition bias	We explored the reasons for dropouts or withdrawal, and whether it was clearly reported.
Sample size	As small sample sizes are likely to lead to an overestimation of the treatment effect, we considered small sample sizes, with less than 50 participants per treatment arm to be high risk (Zhang, Xu, & Ni, 2013).

Statistical analysis

We considered meta-analyses for studies assessing the same outcome of interest and with a similar follow-up period (three- to six-month post-baseline). For studies that had more than one follow-up, we used the first assessment within the three- to six-month follow-up period in the meta-analysis. If the same scale was used across the studies, we reported treatment effect in terms of mean difference (MD). If different scales were used across the studies, the treatment effect was reported as the standardised MD (SMD). As we envisaged heterogeneity between trials, we used random effect models for all analyses. We considered an I² statistic of 50% or greater as an indication of substantial heterogeneity across the studies and explored reasons for this through subgroup analyses to explore whether certain characteristics of trials accounted for heterogeneity. We undertook exploratory analyses to explore whether excluding studies with a high risk of bias or a different intervention format (such as group presentation or online) reduced heterogeneity.

Grading of recommendations assessment, development and evaluation

The grading of recommendations assessment, development and evaluation (GRADE) system (Ryan & Hill, 2016; Schünemann et al., 2017) was used to judge the certainty of evidence behind each outcome. The quality of the evidence was graded as follows:

High: We were very confident that the true effect lies close to that of the effect estimate.

Moderate: We were moderately confident in the effect estimate. The true effect is likely to be close to the effect estimate, but there is a possibility that it is substantially different.

Low: Our confidence in the effect estimate was limited. The true effect may be substantially different from the effect estimate.

Very low: We had very little confidence in the effect estimate. The true effect is likely to be substantially different from the effect estimate.

Depending on the seriousness of the limitation, we would downgrade the evidence by one or more levels. The GRADE judgements were undertaken by one author (CYL) and checked by another (BC). Any disagreements were resolved through discussion, or where necessary, with reference to another author.

We first assumed that the quality of the evidence was high but downgraded if there were serious limitations in:

Risk of bias of contributing studies: This was based on the risk of bias assessment described above. For instance, if most information is from studies at an unclear risk of bias then downgrading by one level may be appropriate as it is likely that there is plausible bias that could seriously alter the results.

Indirectness of evidence: Whether the population, intervention, control or outcomes were not directly relevant to this review. For instance, if the focus of the review is only adults but the studies included involved participants of all ages.

Inconsistency of the results: for example, if the individual studies yielded widely differing estimates of effect. If only one study was identified, this could not be judged; however, downgrading would occur if appropriate for other reasons such as imprecision or risk of bias if sample size was small.

Imprecision of results: if a wide confidence interval was identified which represented uncertainty of the magnitude of the estimated effect, or a limited number of events, then evidence would be downgraded.

The probability of publication bias: Whether there is under or over estimation of impact due to selective publication of the studies. This can be assessed by looking at the pattern of the study results, in particular, if small studies tend to report results in a particular direction compared to larger. The presence of small studies alone is not necessarily an indication of this bias.

Results

Selection of studies

We identified 3221 unique citations of which 3205 were excluded at screening. For a number of abstracts, limited information was provided on the intervention so we further reviewed the method section in the full paper where necessary for a more detailed evaluation. Eighty-nine citations were excluded at this additional ‘screening’ stage. The full texts of the remaining 16 citations were retrieved and reviewed in depth. Of the 16 studies, 11 meet our inclusion criteria (Chien & Lee, 2011; Cristancho-Lacroix et al., 2015; Gavrilova et al., 2009; Guerra, Ferri, Fonseca, Banerjee, & Prince, 2011; Hepburn, Tornatore, Center, & Ostwald, 2001; Kurz, Wagenpfeil, Hallauer, Schneider-Schelte, & Jansen, 2010; Lindstrom Bremer, 2007; Martin-Cook, Remakel-Davis, Svetlik, Hynan, & Weiner, 2003; Onor et al., 2007; Pahlavanzadeh, Heidari, Maghsudi, Ghazavi, & Samandari, 2010; Paun et al., 2015). Four studies were excluded, as they did not state whether the intervention consisted of a specific educational component on the progression of dementia (Beauchamp, Irvine, Seeley, & Johnson, 2005; Ducharme et al., 2011; Gallagher-Thompson, Gray, Dupart, Jimenez, & Thompson, 2008; Gallagher-Thompson et al., 2007), and one other study was excluded because the control condition also consisted of an education component on progression of dementia (Bramble, Moyle, & Shum, 2011). Paun et al. (2015) described their study as a quasi-experimental trial where randomisation was by long-term care facility. Through consulting a statistician, we recategorised the study as a cluster-RCT and included it in our review. Two studies were pilot studies (Cristancho-Lacroix et al., 2015; Paun et al., 2015). Figure 1 shows the review flow diagram.

Figure 1.

Study flow diagram.

Participants and settings

The sample sizes of the included studies ranged from 16 to 292. One study (Martin-Cook et al., 2003) did not report any details regarding participants’ characteristics but only recruited primary carers who were spouses or adult children of a person with dementia who also had behavioural disturbance. In the other studies, participants were family or friends of a person with dementia, with the majority being either a spouse or adult child of the person with dementia. Seventy-one percent of all participants across the 11 studies were female carers. Eight studies reported carers’ education level (Chien & Lee, 2011; Cristancho-Lacroix et al., 2015; Hepburn et al., 2001; Kurz et al., 2010; Lindstrom Bremer, 2007; Onor et al., 2007; Pahlavanzadeh et al., 2010; Paun et al., 2015); most participants completed high school and a minority of them had also completed college or above. Seven studies reported care recipients’ characteristics; most had mild to moderate severity dementia (Chien & Lee, 2011; Cristancho-Lacroix et al., 2015; Gavrilova et al., 2009; Guerra et al., 2011; Kurz et al., 2010; Onor et al., 2007; Pahlavanzadeh et al., 2010).

Four studies took place in the United States (Hepburn et al., 2001; Lindstrom Bremer, 2007; Martin-Cook et al., 2003; Paun et al., 2015), three in Europe (Cristancho-Lacroix et al., 2015; Kurz et al., 2010; Onor et al., 2007), three in a range of low- to middle-income countries (Gavrilova et al., 2009; Guerra et al., 2011; Pahlavanzadeh et al., 2010) and one in Hong Kong (Chien & Lee, 2011). Most trials took place in the community. Four studies took place in the carers’ home (Chien & Lee, 2011; Cristancho-Lacroix et al., 2015; Gavrilova et al., 2009; Guerra et al., 2011), two at long-term care facilities and nursing homes (Lindstrom Bremer, 2007; Paun et al., 2015), one at a medical centre (Hepburn et al., 2001), one at a hospital (Pahlavanzadeh et al., 2010) and one at Alzheimer’s Society centres (Kurz et al., 2010). Despite the location of sessions at a nursing home for the Lindstrom Bremer (2007) study, the person with dementia had to be living at home. The remaining two studies did not specify where the intervention took place (Martin-Cook et al., 2003; Onor et al., 2007).

Characteristics of the intervention

Table 3 shows the key intervention characteristics of the 11 studies. Seven evaluated a group-based intervention (Hepburn et al., 2001; Kurz et al., 2010; Lindstrom Bremer, 2007; Martin-Cook et al., 2003; Onor et al., 2007; Pahlavanzadeh et al., 2010; Paun et al., 2015), two used a home-based intervention where any carers (paid and family) could participate in the education sessions (Gavrilova et al., 2009; Guerra et al., 2011), one was a one-to-one intervention (Chien & Lee, 2011) and one was a computerised intervention where participants could access the intervention online from home (Cristancho-Lacroix et al., 2015). The total duration of the intervention ranged from 4 weeks to 15 months, the number of sessions from 4 to 16 and the length of sessions from 15 to 30 minutes to 120 minutes.

Table 3.

Characteristics of included studies.

Study, country, setting	Sample size (N)	Total duration; number of sessions	Format and who delivered the intervention	Contents of the intervention	Comparator	Outcomes	Follow-up
Chien and Lee (2011), Hong Kong; carers’ home	92	6 Months; 10 × 120 min sessions	One-to-one; case manager	Education on dementia, its prognosis and current treatment and care; social and financial support; psychological support; also targets relationships within the family	Routine care through a dementia research centre	Burden (Family Caregiving Burden Inventory)	Immediately after intervention: 6 months Additional follow-up: 18 and 24 months post-baseline
Cristancho-Lacroix et al. (2015), France, online	49	3 Months; 12 × 15– 30 min sessions	Computerised intervention	Education on the progression of Alzheimer’s Disease and what to expect in the future. Other sessions focused on communication skills, strategies for managing everyday difficulties, social and financial support	Usual care	Knowledge of dementia (self-report on a visual analogue scale), Depression (Beck Depression Inventory), Burden (ZBI)	Immediately after intervention: 3 months Additional follow-up: 6 months post-baseline
Gavrilova et al (2009), Russia; home	60	5 Weeks; 5 × 30 min sessions	Family based; newly qualified doctors who received a 2-day training programme	One assessment session. Two sessions provided education on causes and treatment of dementia and what to expect in the future; two sessions provided training in managing care recipient’s problem behaviours.	Medical care as usual	Burden (ZBI)	Immediately after intervention: No Assessment Additional follow-up: 6 months post-baseline
Guerra et al. (2011), Peru; home	58	5 Weeks; 5 × 30 min sessions	Family-based; junior psychologists and social workers who received a 2-day training programme	Same as above (Gavrilova et al, 2009)	Medical care as usual (received intervention after the follow-up assessment)	Burden (ZBI)	Immediately after intervention: No Assessment Additional follow-up: 6 months post-baseline
Hepburn et al (2001), U.S; Medical Centre	117	7 Weeks; 7 × 120 min sessions	Group based; multidisciplinary team (nurse, educator, family therapist, occupational therapist)	Information on the progressive effect of dementia; information regarding the caregiving role; developing beliefs about caregiving; strategies to manage care recipient’s problem behaviours	Usual care (received intervention after the follow-up assessment)	Depression (CES-D), Burden (ZBI)	Immediately after intervention: No assessment Additional follow-up: 5 months post-baseline
Kurz et al. (2010), Austria, Switzerland, Germany; Alzheimer’s Society centres	292	15 Months; 7 × 90 min sessions + 6 refresher meetings	Group based; psychologists or social workers who received study-specific training	Session 1 focuses on general information about Alzheimer’s disease with a focus on diagnosis and therapy options. Session 4 focuses on late stage of disease including decision making, symptoms and issues of role change. Session 5 focuses on legal and insurance issues including legal representation and advance directives	Usual care	Depression (Montogomery-Asberg Depression Rating Scale)	Immediately after intervention: 15 months
Lindstrom Bremer (2007), U.S; Sessions took place in day care centres and nursing homes but the person with dementia had to be living at home	54	6 Weeks; 6 × 120 min sessions	Group based; gerontologist and nurses (concurrent sessions were run for person with dementia by an occupational or music therapist)	Participants were randomised to 1 of 2 intervention groups or the control group. The two interventions were the day-to-day and decision-making groups. The two interventions had similar content and the same number of sessions and statistical analysis found no differences between them so they were merged for analyses. Both included introduction to dementia and progression of dementia in Session 1 and identifying and tailoring activities to dementia stage in Session 3. The Day-to-Day group focused on teaching carers to tailor everyday task and activities to care recipients’ level and also communication skills. The decision-making group focused on strengthening carers’ decision-making skills	Waitlist	Depression (CES-D), Burden (ZBI)	Immediately after intervention: No assessment Additional follow-up: 6 months post-baseline
Martin-Cook et al (2003), U.S; Not specified	37	4 Weeks; 4 × 120 min sessions	Group based; not specified	Session 1 provided an overview of dementia pathology, pathology, symptoms, course, and treatments. The remaining sessions covered strategies for managing problem behaviours; communication skills; environmental, safety and financial issues, carer feelings and coping strategies	Waitlist	Depression (CES-D)	Immediately after intervention: 6 weeks post-baseline Additional follow-up: 14 weeks post-baseline
Onor et al (2007), Italy; not specified	16	4 Months; 16 × 60 min weekly sessions	Group based, psychiatrist and educator	Information about dementia, the course of illness, disease stages, progression of cognitive and behavioural symptoms	No treatment	Depression (Brief Symptom Inventory), Burden (Caregiver Burden Inventory)	Immediately after intervention: 4 months post-baseline Additional follow-up: 2 months post-baseline (half-way through intervention)
Pahlavanzadeh et al (2010), Iran; Hospital	60	5 Weeks; 5 × 90 min sessions	Group based; supervised by researcher (masters in psychiatry) with experience as a university lecturer and educator of patients and their families	Session 1 covered “Changes in the elderly, the definition, stages and symptoms, risk factors, diagnostic methods, and treatment of dementia” (p104). None of the remaining sessions covered disease progression but focused on behavioural symptoms	Did not attend training programme	Burden (ZBI)	Immediately after intervention: 5 weeks post-baseline Additional follow-up: 9 weeks post-baseline
Paun et al (2015), U.S; Long-term care facilities	93	12 Weeks; 12 sessions (60–90 min)	Group based; psychiatric nurses who received 8-hours preparatory workshop	Three sessions provided information about late stage of Alzheimer’s disease or related dementia, community resources and long-term care facility. In the remaining sessions, communication, conflict resolution and grief management skills were addressed	Two check-in calls	Knowledge of dementia (Knowledge of Alzheimer’s Test), Depression(CES-D), The 50-item Marwit-Meuser Caregiver Grief Inventory –subscale of Personal sacrifice burden was used for carer burden and the total score was used for grief	Immediately after intervention: 3 months post-baseline Additional follow-up: 6 months post-baseline

CES-D: The Centre for Epidemiologic Studies Depression Scale; ZBI: Zarit Burden Interview.

Interventions were delivered by various professionals. Two studies reported the involvement of a multi-disciplinary team (Hepburn et al., 2001; Lindstrom Bremer, 2007). In five studies, the intervention was delivered by personnel who had undergone some training before delivering the intervention (Chien & Lee, 2011; Gavrilova et al., 2009; Guerra et al., 2011; Kurz et al., 2010; Paun et al., 2015). In one study, the intervention was delivered by psychiatrists and educators (Onor et al., 2007) and another by the lead author (masters in psychiatry with experience as a university lecturer and educator of patients and their families) (Pahlavanzadeh et al., 2010). The remaining study did not specify who delivered the intervention (Martin-Cook et al., 2003).

All interventions were multi-component, with education/information on progression of dementia being one to three of the components. Most studies provided teaching sessions only (Chien & Lee, 2011; Gavrilova et al., 2009; Guerra et al., 2011; Hepburn et al., 2001; Kurz et al., 2010; Lindstrom Bremer, 2007; Onor et al., 2007). Of the other studies, two provided both education and written information (Martin-Cook et al., 2003; Paun et al., 2015), another education and a CD (Pahlavanzadeh et al., 2010) and one education as written information on a computer (Cristancho-Lacroix et al., 2015).

The proportion of the interventions that focused on progression of dementia was small; for four it was one session of the entire intervention (Hepburn et al., 2001; Lindstrom Bremer, 2007; Martin-Cook et al., 2003; Pahlavanzadeh et al., 2010). Education/information on progression of dementia accounted for 2 out of 5 sessions in both Gavrilova et al. (2009) and Guerra et al. (2011) studies, and 2 and 3 out of 12 sessions in Cristancho-Lacroix et al. (2015) and Paun et al. (2015) studies, respectively. At the most, the intervention component on the progression of dementia accounted for three of seven of the initial teaching sessions in Kurz et al.’s (2010) study. Based on the details in the papers for Chien and Lee’s (2011) and Onor et al.’s (2007) studies, we were unable to determine the proportion of the intervention that focused on the progression of dementia. Control group participants were either provided with usual care or were waitlist controls.

Shorter term interventions (of less than three-month duration)

Six studies involved an intervention of shorter duration, that is lasting less than three months and had less than 10 sessions (Gavrilova et al., 2009; Guerra et al., 2011; Hepburn et al., 2001; Lindstrom Bremer, 2007; Martin-Cook et al., 2003; Pahlavanzadeh et al., 2010). These interventions aimed to provide in addition to education or information to carers, strategies for managing care recipient’s behavioural and psychological symptoms of dementia (BPSD). One also provided information on environmental, safety and financial issues to carers, as well as psychological support (Martin-Cook et al., 2003).

In both the studies by Hepburn et al. (2001) and Lindstrom Bremer (2007), the intervention aimed to help carers to adapt to the caregiving role. Apart from information on progression of dementia, carers in the intervention group in the Hepburn et al.’s (2001) study were also given information regarding what the caregiving role entails, and strategies in managing care recipients’ BPSD. In the Lindstrom Bremer (2007) study, the intervention focused on teaching carers to tailor everyday activities to care recipient’s level and strengthening carers’ decision-making skills.

Longer term interventions (of more than three-month duration)

Five studies consisted of an intervention of longer duration, lasting for three months or more, and consisted of at least 10 sessions (Chien & Lee, 2011; Cristancho-Lacroix et al., 2015; Kurz et al., 2010; Onor et al., 2007; Paun et al., 2015). Apart from education, most interventions also provided psychological support to carers.

In Chien and Lee’s (2011) and Kurz et al.’s (2010) studies, the intervention was tailored to carers’ needs. Psychological supports were offered to carers in two studies (Onor et al., 2007; Paun et al., 2015); the Onor et al.’s (2007) study focused on addressing carers’ stress and emotion, and the Paun et al.’s (2015) study targeted carers’ grief management skills in order to reduce their chronic grief. In Cristancho-Lacroix et al.’s (2015) study, a computerised intervention was used, where information regarding dementia as well as information regarding how to manage BPSD was provided, it also targeted carers’ communication skills, and social and financial support.

Risk of bias of included studies

Figure 2 presents the risk of bias assessment of the studies. Overall, most studies were classified as unclear risk of bias due to under reporting. Six studies provided details on how they generated random sequence generation. None of the studies reported sufficient details regarding allocation concealment. Blinding of assessors was only evident in four studies. There was no evidence of attrition bias in the studies. All but one study had a small sample size, with less than 50 participants per treatment arm; therefore, they were considered high risk of bias. Nonetheless, six studies conducted a power analysis to determine the sample size required in their studies (Chien & Lee, 2011; Cristancho-Lacroix et al., 2015; Gavrilova et al., 2009; Guerra et al., 2011; Kurz et al., 2010; Pahlavanzadeh et al., 2010). Three studies achieved the required sample size (Chien & Lee, 2011; Kurz et al., 2010; Pahlavanzadeh et al., 2010). Table 4 presents more detailed reasons for each risk of bias judgement for each study.

Figure 2.

Risk of bias summary: review authors’ judgements about each Cochrane risk of bias item for each included study.

Table 4.

Risk of bias judgements of included studies.

Author	Random sequence generation (selection bias)	Allocation concealment (selection bias)	Blinding of outcome assessment (detection bias)	Incomplete outcome data (attrition bias)	Sample size
Chien and Lee (2011)	Low risk: Computer-generated randomisation	Unclear risk: No details	Low risk: Assessment was conducted by a research assistant blinded to randomisation status	Low risk: 3 participants from the intervention group dropout of the study due to mortality of the person with dementia or insufficient time and deterioration of care recipient's condition. One participant from the control group dropout due to death of care recipient.	High risk: Less than 50 per treatment arm
Cristancho-Lacroix et al. (2015)	Low risk: Computer-generated randomisation list using blocks and stratified by sex and relationship	Unclear risk: No details	High risk: Unblinded pilot RCT	Low risk: 8 participants from the intervention group and 7 from the control dropped out of the study, with reasons being hospitalisation, institutionalisation, became illegible through the course of the study and ended participation.	High risk: Less than 50 per treatment arm
Gavrilova et al. (2009)	Low risk: Randomisation carried out in London and stratified using a permuted block method based on carer's burden at baseline	Unclear risk: Central randomisation in London and transmitted back to Russia by email	Low risk: Blinded assessment	Low risk: 5 participants from intervention group and 2 participants lost to follow-up due to death of care recipient	High risk: Less than 50 per treatment arm
Guerra et al. (2011)	Low risk: Randomisation carried out in London and stratified using a permuted block method based on carer’s burden at baseline	Unclear risk: Central randomisation in London and transmitted back to Peru by email	Low risk: Blinded assessment	Low risk: 2 participants from intervention group dropped out due to mortality of the person with dementia	High risk: Less than 50 per treatment arm
Hepburn et al. (2001)	Low risk: Computer-generated randomisation		Unclear risk: No details except to indicate questionnaires were self-completed by participants	Low risk: 23 participants dropout from this study, 12 from the intervention group and 11 from control, with reasons being worsening care recipient’s condition, or other reasons (e.g. transportation difficulties)	High risk: Less than 50 per treatment arm
Kurz et al. (2010)	Unclear risk: Randomisation using a block length of six participants, without any extra information regarding method of randomisation	Unclear risk: Central randomisation, with no further details	Low risk: Blinded assessment	Low risk: No dropouts	Unclear risk: Less than 200 in each treatment arm
Lindstrom Bremer (2007)	Low risk: Block randomisation, generated by computer programme	Unclear risk: No details	Unclear risk: No details	Low risk: 9 drop-outs from intervention groups and 4 from control group	High risk: Less than 50 per treatment arm
Martin-Cook et al. (2003)	Unclear risk: No details	Unclear risk: No details	Unclear risk: No details	Low risk: 1 from intervention group and 2 from control group missing both follow-ups plus another control group participant missing the second follow-up	High risk: Less than 50 per treatment arm
Onor et al. (2007)	Unclear risk: Stated ‘simple randomisation was used’	Unclear risk: No details	Unclear risk: No details	Low risk: No dropouts	High risk: Less than 50 per treatment arm
Pahlavanzadeh et al. (2010)	Unclear risk: No details	Unclear risk: No details	Unclear risk: No details	Low risk: 10 participants, 5 from each group, dropout due to lack of presence in programme, not available for data collection at follow-up	High risk: Less than 50 per treatment arm
Paun et al. (2015)	Unclear risk: No details	Unclear risk: No details	Unclear risk: No details	Low risk: 10 participants dropout from the study, 3 from intervention group and 7 from control group due to lack of interest and care recipient’s death	High risk: Less than 50 per treatment arm

Effects of intervention

Knowledge of dementia

Only two studies reported on the impact of the intervention on carers’ knowledge of dementia (Cristancho-Lacroix et al., 2015; Paun et al., 2015), and therefore, we did not pool their data. In Cristancho-Lacroix et al.’s (2015) study, they measured carers’ knowledge about Alzheimer’s disease using a visual analogue scale where carers evaluated their knowledge of Alzheimer’s disease on a scale from 0 (low) to 100 (high). Paun et al. (2015) measured knowledge using the 22-item Knowledge of Alzheimer’s Test (KAT). For both scales, higher scores indicate better knowledge. Both had follow-up at immediately post-intervention (three months) and three months later.

In Paun et al.’s (2015) study, there was no statistically significant difference between the trial arms in mean change from baseline in carers’ knowledge immediately after intervention (three months) (MD = 0.70 (95% CI −0.12 to 1.52)) or at six-month follow-up (MD = 0.19 (−95% CI −0.72 to 1.10)). Scores reported by Cristancho-Lacroix et al. (2015) for the intervention group were baseline = 45.4 (SD = 23.2); three months = 59.2 (SD = 25.9) and six months = 58.6 (SD = 24.4) compared with the control group baseline = 44.5 (SD = 23.5); three months = 44.4 (SD = 21.6) and six months = 51.7 (SD = 18.8). They reported a statistically significant difference favouring the intervention compared with the control group at the three-month assessment (Cohen’s d = .79, p = .008) but not at six months.

Using the GRADE system, we judged the quality of evidence on improvement in carers’ knowledge of dementia at the end of the three-month intervention to be very low and at six months to be low. Our confidence in the effect estimate was limited. We downgraded the evidence from high by two levels because of study limitations at both time points (low sample size in both and not blinding outcome assessors in the Cristancho-Lacroix et al.’s (2015) study). In addition, we downgraded another level due to inconsistency in results at the end of the three-month intervention.

Depression

Seven studies reported on carers’ depression (Cristancho-Lacroix et al., 2015; Hepburn et al., 2001; Kurz et al., 2010; Lindstrom Bremer, 2007; Martin-Cook et al., 2003; Onor et al., 2007; Paun et al., 2015) using four depression scales (see Table 3). In all scales, a higher score indicated more severe symptoms of depression. The study by Martin-Cook et al. (2003), however, did not anticipate an improvement in depression as their intervention aimed to reduce carers’ resentment and attribution of their relative’s behavioural disturbances without increasing depression.

A meta-analysis was undertaken using data from three studies including two (Hepburn et al., 2001; Lindstrom Bremer, 2007) using the first follow-up score at five- and six-month post-baseline, respectively. The third study (Onor et al., 2007) undertook assessments at two months (during the intervention) and four-month post-baseline. We used the four-month post-baseline data in the meta-analysis as this was the first assessment after the intervention and also fitted within our requirement of between three- and six-month post-baseline. The pooled analysis of the three trials (n = 151) showed a significant statistical effect of education on depression at follow-up (SMD=−0.48 (95% CI −0.82 to −0.14); I² = 0%; p = 0.006; Figure 3).

Figure 3.

Meta-analysis: Depression follow-up scores by study arm.

Four studies were not included in the meta-analysis (Cristancho-Lacroix et al., 2015; Kurz et al., 2010; Martin-Cook et al., 2003; Paun et al., 2015). Two (Martin-Cook et al., 2003; Paun et al., 2015) did not provide scores at follow-up and one did not have follow-up within three to six months (Kurz et al., 2010). Cristancho-Lacroix (2015) reported significant baseline differences between the control and intervention groups on depression so we excluded their follow-up scores from the meta-analysis. Only two studies (Cristancho-Lacroix et al., 2015; Paun et al., 2015) reported mean change scores (from baseline to follow-up) with standard deviations, and therefore, this was insufficient to undertake a meta-analysis. Overall, these four studies did not show statistically significant differences in depression between intervention and control groups either as change scores or mean follow-up scores.

Using the GRADE system, we judged the quality of evidence on reducing depression as low. We downgraded the evidence from high by two levels because of study limitations (lack of detail on allocation concealment, random allocation and blinding of assessors) and imprecision of results (high SDs across studies).

Burden

Nine studies measured carer burden (Chien & Lee, 2011; Cristancho-Lacroix et al., 2015; Gavrilova et al., 2009; Guerra et al., 2011; Hepburn et al., 2001; Lindstrom Bremer, 2007; Onor et al., 2007; Pahlavanzadeh et al., 2010; Paun et al., 2015). Four different scales were used (see Table 3). Higher scores represent higher levels of burden in all four scales.

We undertook two meta-analyses based on whether outcomes were reported as mean change scores or follow-up scores. The first analysis reports on mean scores at follow up and includes five studies (Chien & Lee, 2011; Cristancho-Lacroix et al., 2015; Hepburn et al., 2001; Lindstrom Bremer, 2007; Onor et al., 2007). We used the first follow-up assessments for all studies except Onor et al. (2007) where we used the second assessment which fitted within the three- to six-month timeframe and was after the intervention was completed (excluding the two-month follow-up half way through the intervention). The pooled analysis showed a statistically non-significant effect of education on carers’ burden (SMD=−0.31 (95% CI −0.64 to 0.03); I² = 43%; p = 0.07; n = 292; Figure 4).

Figure 4.

Meta-analysis: Burden follow-up scores by study arm.

Four studies reported on mean changes (Cristancho-Lacroix et al., 2015; Gavrilova et al., 2009; Guerra et al., 2011; Paun et al., 2015) using the first follow-up assessment. Meta-analysis of these studies showed a statistically non-significant effect (SMD= −0.26 (95% CI −0.93 to 0.42); I²=85%; p = 0.46; n = 241; Figure 5). The I² for this meta-analysis suggests substantial heterogeneity across the trials. In a sub-group analyses, we removed the Cristancho-Lacroix (2015) study given it used a computerised rather than face-to-face intervention but this did not substantially reduce heterogeneity.

Figure 5.

Meta-analysis: Burden mean changes from baseline to follow-up by study arm.

Only one study measuring burden was excluded from both analyses due to follow-up assessments being undertaken in less than three-month post-baseline (Pahlavanzadeh et al., 2010). There were also limitations in the paper in that there were inconsistencies in the changes they described and the scores presented. We sought clarification from the author but did not receive a response.

Using the GRADE system, we judged the quality of evidence on reducing burden as very low. We downgraded the evidence from high by three levels because of risk of bias (lack of detail on allocation concealment, random allocation and blinding of assessors), imprecision of results (high SDs across studies) and inconsistency in results (high levels of heterogeneity in meta-analysis).

Pre-death grief

One study (Paun et al., 2015) measured carers’ chronic grief using the MM-CGI, with higher scores indicating higher level of grief. Follow-up was at six months. No statistically significant difference was found between intervention and control groups in carers’ grief immediately after the intervention (three months; MD −1.22 (95% CI −10.17 to 7.73)) or at six-month follow-up (MD −1.53 (95% CI −10.27 to 7.21)).

Using the GRADE system, we judged the quality of evidence on reducing pre-death grief as low. We downgraded the evidence from high by two levels because of study limitations (lack of detail on: allocation concealment, random allocation, blinding of assessors and small sample size) and imprecision of results (high SDs).

Discussion

This systematic review is the first review to our knowledge to examine the effect of educating carers of people with dementia about the progression of dementia. It was undertaken in light of the European Association of Palliative Care’s white paper (van der Steen et al., 2014), recommending education regarding the progressive course of dementia should be provided to families of the person with dementia. We found 11 RCTs that met our inclusion criteria.

We were unable to undertake meta-analyses on the outcomes of knowledge and grief due to fewer than three studies measuring these outcomes. Results from two studies (Cristancho-Lacroix et al., 2015; Paun et al., 2015) provided mixed evidence that education or information on progression of dementia could improve carers’ knowledge of dementia. Based on data from one study (Paun et al., 2015), there was no evidence to support the effectiveness of an educational intervention in reducing carers’ pre-death grief. Our meta-analyses showed no treatment effect on burden but found a significant benefit on depression.

These findings, however, need to be taken in light of the quality of the evidence. Using the GRADE system (Schünemann et al., 2017), we judged the quality of evidence to be very low for depression (we have very little confidence in the effect estimate and the true effect is likely to be substantially different) and low for burden, knowledge and pre-death grief (we have limited confidence in the effect estimate and the true effect may be substantially different). We downgraded outcomes for various reasons including small samples sizes, high SDs in change and lack of reporting of allocation concealment and blinding of assessors. Five studies provided unclear evidence of random sequence generation (Kurz et al., 2010; Martin-Cook et al., 2003; Onor et al., 2007; Pahlavanzadeh et al., 2010; Paun et al., 2015). We considered that it is not possible to conceal group allocation to participants or personnel involved in an education intervention and therefore did not include this in our assessment of risk of bias. It is important, however, to blind the outcome assessment to reduce potential bias, but only four studies reported this (Chien & Lee, 2011; Gavrilova et al., 2009; Guerra et al., 2011; Kurz et al., 2010).

There are other reasons why our conclusions on the evidence derived from these studies are limited. The studies we identified also varied in terms of the outcomes they used, the duration and intensity of the interventions and the delivery method. Most were group sessions with carers who cared for someone at home (Hepburn et al., 2001; Kurz et al., 2010; Lindstrom Bremer, 2007; Martin-Cook et al., 2003; Onor et al., 2007; Pahlavanzadeh et al., 2010), whilst one was for carers of people who had moved into a nursing home (Paun et al., 2015). One study used a computerised intervention which found improvements in self-reported knowledge but not burden (Cristancho-Lacroix et al., 2015). The three studies using individual/family-based face-to-face interventions all reported improvements in burden (Chien & Lee, 2011; Gavrilova et al., 2009; Guerra et al., 2011). These studies were the highest quality studies in the review with blinding of assessors and random sequence generation; however, due to different reporting of results, these studies were not included in the same meta-analysis. Two of these individualised studies and another study (Gavrilova et al., 2009; Guerra et al., 2011; Pahlavanzadeh et al., 2010) were from low-middle income countries and also found significant reductions in burden. Further research is needed to examine whether education for carers is more effective on an individual basis or in particular countries. Possibly in higher income countries, education is offered more routinely, and therefore, the interventions offered were not a substantial addition to routine care. The overall small number of studies and low quality, however, prevented subgroup analyses to examine specific intervention features. The moderate to high heterogeneity in the two meta-analyses on burden also limit our interpretations of these findings.

The proportion of the interventions that focused on education/information on progression of dementia in the included studies was small; one to three sessions of interventions ranging from 4 to 16 sessions in total. This further limits the extent to which we can determine whether it was the education around dementia progression rather than other elements of the interventions which were impacting on carer knowledge and mental health.

Whilst most of the studies reported on burden and depression, only one study reported on pre-death grief and two on dementia knowledge. As pre-death grief is common among dementia carers (Chan et al., 2013), it is important to identify effective interventions that could reduce grief which also help to reduce depression and burden (Lindauer & Harvath, 2014). Only two of our studies measured knowledge of dementia as an outcome (Cristancho-Lacroix et al., 2015; Paun et al., 2015). We expect that improving mental health from an educational intervention is a secondary outcome resulting from an improved understanding of disease progression. However, if we are not measuring improved knowledge, it is difficult to determine whether improvements in mental health are due to improved knowledge or some other aspect of the intervention such as social connection or feeling heard.

Current evidence provides limited guidance on how we can improve dementia knowledge. In the UK, a third of older adults have difficulty interpreting basic health information (Bostock & Steptoe, 2012) and low health literacy is more common amongst those from more deprived backgrounds, ethnic minorities, older people and those with chronic health conditions (Coulter & Jo, 2006). How do we know whether verbal and written information provided to family carers is accurate, understood and positively impacting on psychosocial well-being? The need to improve dementia knowledge amongst staff and the general public has been identified as a barrier to good quality palliative care in dementia (Carter, van der Steen, Galway, & Brazil, 2015), yet there is also some evidence that knowledge of the biomedical aspects of dementia is associated with increased anxiety (Proctor, Martin, & Hewison, 2002). Understanding the clinical course of dementia may be helpful for planning and preparing for the future but some carers may be reluctant to think about the impacts on their loved one of the progression of dementia. Denial and avoidance may be common coping strategies used by family carers, but they have been associated with poorer psychological outcomes (Gilhooly et al., 2016). Perhaps greater attention is needed not only on the type of information delivered but how we provide emotional support to carers trying to process and accept distressing information.

One of the promising findings from this review was the low dropout rate in most studies suggesting that educational interventions were generally viewed as acceptable by carers. Moreover, qualitative findings indicated that participants found education on the progression of the disease, the decision-making guidance and caregiving strategies particularly useful (Lindstrom Bremer, 2007), and these findings can be built on in future research.

Strengths and limitations

This review was conducted as outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins & Green, 2011); therefore, the introduction of bias should be minimised. The search strategy used was extensive, increasing the chances of identifying all relevant studies. We also included grey literature if they were available from the selected databases and the full paper could be obtained. It is possible, however, that some of the studies we excluded consisted of an educational component focusing on progression of dementia, but since this was not clearly specified or even suggested in the abstract or method section of the full paper, they were not included in this review.

Although all but one study had been classified as high risk of bias in terms of sample size, three studies conducted a power analysis to determine the required sample size and achieved this sample size (Chien & Lee, 2011; Kurz et al., 2010; Pahlavanzadeh et al., 2010). However, power analysis was not a Cochrane criterion, and this was not taken into account when assessing the possible biases confounded by sample size.

Conclusion

Based on the results obtained, there was not sufficient evidence available to support or refute the effectiveness of education on progression of dementia on carers’ knowledge and mental health. Therefore, we cannot provide any clinical recommendations based on these results. Nonetheless, the meta-analysis revealed that the treatment effect was in favour of intervention, suggesting a possibility that education on progression of dementia could help to improve mental health. Further robust research is required that use educational interventions regarding the progression of dementia which assess the impact on knowledge of dementia, mental health and burden outcomes. Consideration of the format, duration and stage of dementia are also required.

Footnotes

Acknowledgement

We would like to acknowledge Victoria Vickerstaff, the statistician who provided advice for interpreting one of the papers and the meta-analysis.

Declaration of conflicting of Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval

This paper reports findings from a systematic review and did require ethical approval as the study did not involve human or animal participation.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: KJM was supported by a Senior Fellowship from Alzheimer's Society, UK (grant number 325: AS-SF-16-004). ELS's and BC's posts are supported by Marie Curie core grant MCCC‐FCO‐16‐U.

Kirsten J Moore is a social gerontologist and has been engaged in a broad range of research programs to improve the quality of care for older people and people with dementia. Her research interests focus on the care of people with advanced dementia and identifying better ways to support carers to prepare for end of life and to cope with grief. She worked at the National Ageing Research Institute in Australia from 1999 to 2014 and has since worked at the Marie Curie Palliative Care Research Department at the University College London.

Cheuk Yan Lee was involved in this systematic review as part of completion of her Masters of Mental Health Sciences Research in the Division of Psychiatry at the University College London.

Elizabeth L. Sampson is a Reader in the Marie Curie Palliative Care Research Department, Division of Psychiatry, University College London. Her research focusses on liaison psychiatry for older people (epidemiology and outcomes for older people with dementia and delirium in acute hospitals) and palliative and end of life care in dementia (pain, epidemiological and health services research). Liz works clinically as lead consultant for older people’s liaison psychiatry at North Middlesex University Hospital.

Bridget Candy is a principal research fellow within the UCL Marie Curie Palliative Care Research Department where she leads on systematic reviews in topics relating to the care of people with chronic progressive diseases. Her expertise includes the design and conduct of systematic reviews of complex interventions that can involve either or both trial evidence and qualitative evidence on peoples’ perspectives.

References

Abreu

Tolson

Jackson

G. A.

Costa

(2018). A cross-sectional study of family caregiver burden and psychological distress linked to frailty and functional dependency of a relative with advanced dementia. Dementia (London). Advance online publication. DOI: 10.1177/1471301218773842.

Alzheimer’s Disease International. (2015). World Alzheimer report 2015: The global impact of dementia. An analysis of prevalence, incidence, cost and trends. Retrieved from http://www.alz.co.uk/research/WorldAlzheimerReport2015.pdf

Barry

L. C.

Kasl

S. V.

Prigerson

H. G.

(2002). Psychiatric disorders among bereaved persons: The role of perceived circumstances of death and preparedness for death. The American Journal of Geriatric Psychiatry, 10(4), 447–457.

Beauchamp

Irvine

Seeley

Johnson

(2005). Worksite based internet multimedia program for family caregivers of persons with dementia. The Gerontologist, 45(6), 793–801.

Bostock

Steptoe

(2012). Association between low functional health literacy and mortality in older adults: Longitudinal cohort study. BMJ, 344, e1602. DOI: 10.1136/bmj.e1602.

Bramble

Moyle

Shum

(2011). A quasi-experimental design trial exploring the effect of a partnership intervention on family and staff well-being in long-term dementia care. Aging & Mental Health, 15(8), 995–1007.

Carter

van der Steen

J. T.

Galway

Brazil

(2015). General practitioners' perceptions of the barriers and solutions to good-quality palliative care in dementia. Dementia (London). Advance online publication. DOI: 10.1177/1471301215581227.

Chan

Livingston

Jones

Sampson

E. L.

(2013). Grief reactions in dementia carers: A systematic review. International Journal of Geriatric Psychiatry, 28(1), 1–17. DOI: 10.1002/gps.3795.

Chien

W. T.

Lee

I. Y.

(2011). Randomized controlled trial of a dementia care programme for families of home-resided older people with dementia. Journal of Advanced Nursing, 67(4), 774–787.

10.

Coulter

(2006). Patient-focused interventions: A review of the evidence. Retrieved from http://www.health.org.uk/sites/default/files/PatientFocusedInterventions_ReviewOfTheEvidence.pdf

11.

Cristancho-Lacroix

Wrobel

Cantegreil-Kallen

Dub

Rouquette

Rigaud

A.-S.

(2015). A Web-based psychoeducational program for informal caregivers of patients with Alzheimer’s disease: A pilot randomized controlled trial. Journal of Medical Internet Research, 17(5), e117.

12.

Cuijpers

(2005). Depressive disorders in caregivers of dementia patients: A systematic review. Aging & Mental Health, 9(4), 325–330. DOI: 10.1080/13607860500090078.

13.

Ducharme

F. C.

Levesque

L. L.

Lachance

L. M.

Kergoat

M.-J.

Legault

A. I.

Beaudet

L. M.

Zarit

S. H.

(2011). “Learning to become a family caregiver:” Efficacy of an intervention program for caregivers following diagnosis of dementia in a relative. The Gerontologist, 51(4), 484–494.

14.

Durepos

Sussman

Ploeg

Akhtar-Danesh

Punia

Kaasalainen

(2018). What does death preparedness mean for family caregivers of persons with dementia? American Journal of Hospice and Palliative Medicine. Advance online publication. DOI: 10.1177/1049909118814240.

15.

Gallagher-Thompson

Gray

H. L.

Dupart

Jimenez

Thompson

L. W.

(2008). Effectiveness of cognitive/behavioral small group intervention for reduction of depression and stress in non-Hispanic White and Hispanic/Latino women dementia family caregivers: Outcomes and mediators of change. Journal of Rational-Emotive & Cognitive-Behavior Therapy, 26(4), 286–303.

16.

Gallagher-Thompson

Gray

H. L.

Tang

P. C.

C. Y.

Leung

L. Y.

Wang

P. C.

Thompson

L. W.

(2007). Impact of in-home behavioral management versus telephone support to reduce depressive symptoms and perceived stress in Chinese caregivers: Results of a pilot study. American Journal of Geriatric Psychiatry, 15(5), 425–434.

17.

Gavrilova

S. I.

Ferri

C. P.

Mikhaylova

Sokolova

Banerjee

Prince

(2009). Helping carers to care–the 10/66 Dementia Research Group’s randomized control trial of a caregiver intervention in Russia. International Journal of Geriatric Psychiatry, 24(4), 347–354. DOI: 10.1002/gps.2126.

18.

Gilhooly

K. J.

Gilhooly

M. L.

Sullivan

M. P.

McIntyre

Wilson

Harding

Crutch

(2016). A meta-review of stress, coping and interventions in dementia and dementia caregiving. BMC Geriatrics, 16, 106. DOI: 10.1186/s12877-016-0280-8.

19.

Guerra

Ferri

C. P.

Fonseca

Banerjee

Prince

(2011). Helping carers to care: The 10/66 Dementia Research Group’s randomized control trial of a caregiver intervention in Peru. Revista Brasileira de Psiquiatria, 33(1), 47–54.

20.

Hebert

R. S.

Dang

Schulz

(2006). Preparedness for the death of a loved one and mental health in bereaved caregivers of patients with dementia: Findings from the REACH study. Journal of Palliative Medicine, 9(3), 683–693. DOI: 10.1089/jpm.2006.9.683.

21.

Hebert

R. S.

Prigerson

H. G.

Schulz

Arnold

R. M.

(2006). Preparing caregivers for the death of a loved one: A theoretical framework and suggestions for future research. Journal of Palliative Medicine, 9(5), 1164–1171. DOI: 10.1089/jpm.2006.9.1164.

22.

Hepburn

K. W.

Tornatore

Center

Ostwald

S. W.

(2001). Dementia family caregiver training: Affecting beliefs about caregiving and caregiver outcomes. Journal of the American Geriatrics Society, 49(4), 450–457.

23.

Higgins

Green

(2011). Cochrane handbook for systematic reviews of interventions (Vol.4). Hoboken, NJ: John Wiley & Sons.

24.

Jensen

Agbata

I. N.

Canavan

McCarthy

(2015). Effectiveness of educational interventions for informal caregivers of individuals with dementia residing in the community: Systematic review and meta-analysis of randomised controlled trials. International Journal of Geriatric Psychiatry, 30(2), 130–143. DOI: 10.1002/gps.4208.

25.

Kurz

Wagenpfeil

Hallauer

Schneider-Schelte

Jansen

(2010). Evaluation of a brief educational program for dementia carers: The AENEAS Study. International Journal of Geriatric Psychiatry, 25(8), 861–869. DOI: 10.1002/gps.2428.

26.

Lindauer

Harvath

T. A.

(2014). Pre-death grief in the context of dementia caregiving: A concept analysis. Journal of Advanced Nursing, 70(10), 2196–2207. DOI: 10.1111/jan.12411.

27.

Lindstrom Bremer

K. M.

(2007). Partners in caregiving: Effects of a psychoeducational intervention on Alzheimer’s caregiving daughters (Doctoral thesis). Available from ProQuest Dissertations and Theses database. (UMI No. 3269000)

28.

Martin-Cook

Remakel-Davis

Svetlik

Hynan

L. S.

Weiner

M. F.

(2003). Caregiver attribution and resentment in dementia care. American Journal of Alzheimer’s Disease and Other Dementias, 18(6), 366–374.

29.

Marwit

S. J.

Meuser

T. M.

(2002). Development and initial validation of an inventory to assess grief in caregivers of persons with Alzheimer’s disease. Gerontologist, 42(6), 751–765.

30.

Mitchell

S. L.

Teno

J. M.

Kiely

D. K.

Shaffer

M. L.

Jones

R. N.

Prigerson

H. G.

Hamel

M. B.

(2009). The clinical course of advanced dementia. New England Journal of Medicine, 361(16), 1529–1538. DOI: 10.1056/NEJMoa0902234.

31.

Onor

M. L.

Trevisiol

Negro

Signorini

Saina

Aguglia

(2007). Impact of a multimodal rehabilitative intervention on demented patients and their caregivers. American Journal of Alzheimer’s Disease & Other Dementias, 22(4), 261–272.

32.

Pahlavanzadeh

Heidari

F. G.

Maghsudi

Ghazavi

Samandari

(2010). The effects of family education program on the caregiver burden of families of elderly with dementia disorders. Iranian Journal of Nursing and Midwifery Research, 15(3), 102–108.

33.

Paun

Farran

C. J.

Fogg

Loukissa

Thomas

P. E.

Hoyem

(2015). A chronic grief intervention for dementia family caregivers in long-term care. Western Journal of Nursing Research, 37(1), 6–27. DOI: 10.1177/0193945914521040.

34.

Proctor

Martin

Hewison

(2002). When a little knowledge is a dangerous thing . . . : A study of carers’ knowledge about dementia, preferred coping style and psychological distress. International Journal of Geriatric Psychiatry, 17(12), 1133–1139. DOI: 10.1002/gps.762.

35.

Radloff

L. S.

(1977). The CES-D Scale:A Self-Report Depression Scale for Research in the General Population. Applied Psychological Measurement, 1(3), 385–401. DOI: 10.1177/014662167700100306.

36.

Robinson

Eccleston

Annear

Elliott

K. E.

Andrews

Stirling

McInerney

(2014). Who knows, who cares? Dementia knowledge among nurses, care workers, and family members of people living with dementia. Journal of Palliative Care, 30(3), 158–165.

37.

Ryan

Hill

(2016). How to GRADE the quality of the evidence, Version 3. Retrieved from http://cccrg.cochrane.org/author-resources

38.

Schulz

Boerner

Shear

Zhang

Gitlin

L. N.

(2006). Predictors of complicated grief among dementia caregivers: A prospective study of bereavement. The American Journal of Geriatric Psychiatry, 14(8), 650–658. DOI: 10.1097/01.JGP.0000203178.44894.db.

39.

Schünemann

Oxman

Higgins

Vist

Glasziou

Akl

on behalf of the Cochrane GRADEing Methods Group and the Cochrane Statistical Methods Group. (2017). Completing ‘summary of findings’ tables and grading the confidence in or quality of the evidence. pp. 359–389. In J. Higgins, R. Churchill, J. Chandler, & M. Cumpston (Eds.), Cochrane Handbook for Systematic Reviews of Interventions version 5.2.0. The Cochrane Collaboration.

40.

Steinhauser

K. E.

Christakis

N. A.

Clipp

E. C.

McNeilly

Grambow

Parker

Tulsky

J. A.

(2001). Preparing for the end of life: Preferences of patients, families, physicians, and other care providers. Journal of Pain and Symptom Management, 22(3), 727–737.

41.

van der Steen

J. T.

Onwuteaka-Philipsen

B. D.

Knol

D. L.

Ribbe

M. W.

Deliens

(2013). Caregivers' understanding of dementia predicts patients’ comfort at death: A prospective observational study. BMC Medicine, 11, 105. DOI: 10.1186/1741-7015-11-105.

42.

van der Steen

J. T.

Radbruch

Hertogh

C. M.

de Boer

M. E.

Hughes

J. C.

Larkin

Volicer

(2014). White paper defining optimal palliative care in older people with dementia: A Delphi study and recommendations from the European Association for Palliative Care. Palliative Medicine, 28(3), 197–209. DOI: 10.1177/0269216313493685.

43.

Wimo

Gauthier

Prince

(2018). Global estimates of informal care. Retrieved from https://www.alz.co.uk/adi/pdf/global-estimates-of-informal-care.pdf

44.

Wimo

Jonsson

Bond

Prince

Winblad

(2013). The worldwide economic impact of dementia 2010. Alzheimer’s & Dementia, 9(1), 1–11.e13. DOI: 10.1016/j.jalz.2012.11.006.

45.

World Health Organization. (2017). The top 10 causes of death. Fact sheet N°310. Retrieved from http://www.who.int/mediacentre/factsheets/fs310/en/

46.

Zarit

S. H.

Reever

K. E.

Bach-Peterson

(1980). Relatives of the impaired elderly: Correlates of feelings of burden. Gerontologist, 20(6), 649–655. DOI: 10.1093/geront/20.6.649.

47.

Zhang

(2013). Small studies may overestimate the effect sizes in critical care meta-analyses: A meta-epidemiological study. Critical Care (London, England), 17(1), R2. DOI: 10.1186/cc11919.