Effectiveness of interventions for informal caregivers of people with end-stage chronic illness: a systematic review

Abstract

Background

People living with advanced, non-malignant chronic conditions often have extensive and complex care needs. Informal or family caregivers often provide the care and support needed by those with advanced chronic conditions at home. These informal caregivers experience many challenges associated with their caring role, which can impact their own wellbeing. Whilst there is growing evidence around the impact on carers, guidance on support for informal caregivers of patients with advanced, non-malignant, chronic conditions is lacking, with little evidence available on effective psychosocial carer interventions. This systematic review explored existing interventions for caregivers of those with advanced, non-malignant, chronic illness, in order to assess the effectiveness of these interventions in improving psychosocial outcomes.

Methods

Electronic databases, Medline, CINAHL, EMBASE, and PsycINFO, were searched up to the end of March 2023. Studies meeting the inclusion criteria, focusing on interventions to improve psychosocial outcomes, such as depression, anxiety, quality of life, and caregiver burden, in this cohort of caregivers were included. Data were extracted regarding study setting, design, methods, intervention components, and outcomes. Risk of bias and quality assessment were conducted.

Results

A total of 5281 articles were screened, ultimately identifying 12 studies for inclusion, reported in 13 publications. A narrative synthesis revealed mixed results. Psychosocial interventions resulted in more significant improvements in psychosocial outcomes than psychoeducational or support interventions, with interventions for carer-patient dyads also reflecting more positive outcomes for caregivers. Evidence-based interventions, guided by an appropriate theoretical model, were reportedly more effective in improving caregiver outcomes. Differences in outcomes were related to intervention development, design, delivery, and outcome assessment.

Conclusions

This review, to our knowledge, is the first to explore the effectiveness of interventions in improving psychosocial outcomes for caregivers of those with advanced, non-malignant, chronic conditions. The review highlights the need for more robust, sufficiently powered, high-quality trials of evidence-based interventions for caregivers of people with advanced chronic illness. Optimal intervention duration and frequency of sessions are unclear and need further exploration.

Introduction

The number of older adults with advanced, non-malignant conditions in need of end-of-life care exceeds those with malignant conditions [1]. This is expected to grow due to an aging population and the rise in the number of people with chronic illness [2]. As such, it is anticipated that there will be increased demand from people living with chronic conditions who are frail and require complex care [3]. Current UK figures estimate around 1 in 8 people living here are carers (6.5 ~ 6.8 million) [4, 5]. Informal caregivers provide the patient with care and support in all aspects of advanced illness. This may include provision of emotional support, support with daily activities, providing physical care, completing household tasks, cooking, collecting prescriptions, monitoring medications and medication adherence, monitoring symptoms, possibly even actively participating in care planning and decision-making, and often acting as an advocate for the patient [6, 7].

As such, informal carers in the UK serve not only to provide care, but also to reduce the burden on the health service; however, these caregivers are often unsupported placing them at risk of burnout [8]. Caregiver need remains largely unaddressed with research to date in this area generally guided by work in cancer care [1, 9]. Caregiving in advanced chronic disease is demanding in nature [3, 10], impacting the caregiver’s physical, psychological, emotional, and social wellbeing [11, 12]. Advanced ‘chronic and uncertain conditions’ present a huge challenge for informal carers [13] p. 2). Caregivers may have acted in their capacity as caregiver for a long period of time [14] with patient needs changing over time, and advancing illness bringing different or increasing symptoms [13]. Caregiver burden in chronic illness will generally increase as the condition progresses [12]. Informal caregivers cannot manage this burden alone; they need support, guidance, knowledge, and skills to manage the complex care needs of advanced chronic illness [14, 15].

Caregiving can take its toll on the caregiver, affecting their physical, psychological, and emotional wellbeing [10,11,12]. Providing supportive interventions to address caregiver needs and challenges in advanced, non-malignant, chronic illness is essential, as left unaddressed, these challenges can affect condition management for patients, serving to increase the complexity of chronic condition management [10]. Non-malignant refers to a condition which is not cancerous, such as neurological conditions, coronary heart disease, or kidney disease [16]. It is argued that intervention among this cohort of caregivers is essential to guard against caregiver burden and burnout, and is becoming increasingly more urgent in light of an aging population [17]. There is a growing body of literature highlighting the experiences, burden, and negative impact of caring on caregivers for those with advanced chronic illness, (e.g. [10, 17, 18]). However, supportive interventions for these caregivers remain underdeveloped [10, 12]. The literature highlights the need for improved provision focused on ‘developing tools to help caregivers cope and manage their own needs’ [17]p. 9).

Research to date has highlighted the needs of caregivers in advanced illness, with clear demand for support responsive to their emotional and psychosocial needs, information on condition management and practical aspects, as well as advice on self-care [18, 19]. There is a call from caregivers that this support should be more proactive in nature, responding at an earlier stage rather than when the situation becomes unmanageable [19, 20]. There is agreement within the literature that interventions should ensure ‘prioritisation of psychological impact from caring’ [4] p. 356), particularly considering the increased incidence of depression, anxiety, stress, and burden among caregivers for those with advanced chronic illness [1, 4, 19]. The need to explore effective interventions for this group of caregivers is increasingly acute due to the growth of chronic illness and our aging population, both of which will be indicative of future demand for care [21]. Given the lack of existing collated information on effective intervention for carers, the aim of this systematic review was to explore the effectiveness of interventions in improving psychosocial outcomes for caregivers of those with advanced, non-malignant, chronic conditions.

Materials and methods

Registration

This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [22]. The review was registered, and accepted, in the international Prospective Register of Systematic Reviews (PROSPERO) [CRD42021279151].

Search strategy

Four electronic databases, Medline, CINAHL, EMBASE, and PsycINFO, were searched up to 31st March 2023, with no date restrictions applied. Appropriate key words and medical subject heading terms (MeSH) for studies, that were relevant to psychosocial outcomes in caregivers for patients with end-stage/advanced chronic illness, were developed and verified by the research team (see Supplementary Materials Table 1).

Eligibility

The review utilised the Population, Intervention, Comparison, Outcomes and Study design (PICOS) search tool for inclusion and exclusion criteria.

Inclusion criteria:

P — Adult, informal caregivers for individuals with end-stage/palliative/advanced chronic conditions (≥ 50% of sample to fit this patient cohort in mixed samples).

I — Targeting caregiver psychosocial outcomes to include quality of life (QoL), depression, anxiety, caregiver burden; or secondary outcomes of interest including distress, stress, and self-efficacy.

C — Comparator group not necessary. Baseline and follow-up data (pre-post intervention at a minimum) to evidence any changes affected by intervention in caregivers.

O — Quantitative (can be as part of mixed methods design); QoL, depression or/and anxiety, caregiver burden; or secondary outcomes of interest as above. Reported as standardised mean difference (SMD) from baseline.

S — Randomised controlled trials (RCTs), or intervention/non-traditional/quasi-experimental trials, e.g. pre and post intervention testing, non-randomised with control.

Exclusion criteria are as follows: dissertations; study protocols; case studies; studies with incomplete data; cross-sectional studies were excluded. Studies were also excluded if they focused on the following: patient rather than caregiver outcomes; caregiver involvement in patient management only; caregivers for children; malignant conditions; or qualitative outcomes.

Study selection and data extraction

All records identified through database searches were imported into Covidence, online systematic review management software [23]. Duplicates (n = 1134) were removed. All remaining 5281 titles and abstracts were screened by the lead author (KMG) and independently by two researchers (GL and AC). Full-text (n = 67) review was completed by the lead author (KMG) with two independent researchers (GL and AC) sharing the full-text screening to ensure agreement. Any disagreements were to be resolved by a fourth researcher (CC); however, no disagreements arose.

Data were extracted, independently, by three reviewers (KMG, GL, AC). For the included studies, key data were extracted, including the following: study details/characteristics (first author, country and year of publication), study design, population, intervention description/components, relevant outcome measures, and intervention effects.

Risk of bias and quality assessment

Risk of bias was assessed using the Cochrane Risk of Bias 2.0 (RoB 2.0) [24]. RoB 2.0 allows for the assessment of risk of bias across six domains, with resulting low, unclear, or high risk of bias. The quality of RCTs in this review was assessed using this tool. For the remaining trials, the Joanna Briggs Institute reviewer manual was used to assess the risk of bias and quality across nine domains [25].

Data synthesis

Due to the wide variation in clinical population, outcome measures, and intervention design, a narrative synthesis of the data was conducted, in line with the Centre for Reviews and Dissemination guidance [26]. Study characteristics, intervention characteristics, quality, and findings are reported.

Results

Selection of studies

In total, 6401 records were identified via database searches. With duplicates removed, 5281 were screened, with 5214 excluded after title and abstract screening. Fourteen records were identified from other sources including hand searching of study reference lists (backward citations), citing literature (forward citations), and trial databases. After a full-text review of the remaining articles, 12 studies (in 13 publications) were included (see Fig. 1: Flowchart). An inter-rater reliability analysis was conducted using Fleiss Kappa, an adaptation of Cohen’s Kappa utilised with 3 or more raters [27]. The analysis reflected moderate agreement for title and abstract screening (κ = 0.502) and substantial agreement for full-text screening (κ = 0.801) [28].

Fig. 1

Flowchart of screening and selection process

Risk of bias and quality assessment

Results from the RoB 2.0 assessment are reported in Figs. 2 and 3. From these figures, it can be seen that of the 8 RCTs included in this review, 3 studies had an overall low risk of bias [29,30,31], whilst the remaining 5 had an overall unclear risk of bias, noted as ‘some concerns’ [32,33,34,35]. Figure 3 shows in which domains the concerns arose.

Fig. 2

Cochrane RoB 2.0: summary of bias for each study (RCT)

Fig. 3

Cochrane RoB 2.0: summary of bias across all included RCTs

The remaining 5 studies were assessed using the Joanna Briggs Institute reviewer manual [25]. In 2 of the quasi-experimental studies, all or nearly all of the appraised domains are endorsed positively [36, 37]. The remaining 3 studies have positive endorsements of many of the applicable domains [38,39,40]. The results of this assessment are presented in Table 1.

Table 1 Risk of bias assessment for experimental studies

Study characteristics

A total of 12 studies (13 publications) were included. Seven studies took place in the USA [29, 30, 34,35,36, 38, 40], 1 in Sweden (in 2 publications: [31, 41], 1 in Singapore [32], 2 in Hong Kong [33, 39], and 1 in Israel [37]. The majority of the included studies were RCTs [29,30,31,32,33,34,35, 41].

The overall sample comprised 1353 caregivers (Table 2). Study sample sizes ranged from 10 to 365 caregivers. The mean caregiver age ranged from 44 to 69.5 years. In all studies, a much higher proportion of caregivers were female (56–97%). Included studies reported on interventions for carers of those with a range of end-stage or advanced, non-malignant chronic conditions, including the following: heart failure [30, 31, 35, 38, 40, 41], renal failure [32, 33, 37], critical chronic illness [29, 34], and mixed chronic illness [36, 39].

Table 2 Study characteristics, population, study design, intervention delivery, structure, type, and follow-up

Cs caregivers, Pts patients, IG intervention group, CG control group, RCT randomised controlled trial.

Intervention characteristics

Interventions were delivered in the home [30, 35,36,37,38,39,40], hospital [29, 32, 34], or condition-specific clinic [31, 33, 41]. Interventions were described as delivered by the following: nurses, interventionists, or teams trained in the intervention approach [29,30,31, 36, 38, 41]; or professionals and volunteers with experience in the intervention type/method [32, 37, 39]. The remaining interventions reported no specific additional training in the intervention approach/methods but did report the involvement of chronic condition specialists [33,34,35, 40]. Intervention duration ranged from 4 days to 48 weeks, comprising 2–36 sessions. Where session duration was reported, sessions lasted from 30 min to 2 h (see Table 2).

Control group

Three studies did not include a control/comparison group [38,39,40]. From the remaining studies, 7 reported the control group (CG) received ‘usual care’ [30,31,32,33,34, 37, 41]. Two studies reported the CG received usual care with additional informational materials [29, 35]. One study stated CG received minimal, non-specific support [36].

Psychosocial outcomes for caregivers

All interventions measured changes in psychosocial outcomes (see Supplementary Materials: Table 2) in caregivers of those with advanced/end-stage chronic illness, including caregiver stress/burden/strain (n = 9: [30, 32,33,34,35,36, 38, 39, 41]); depression (n = 9: [29,30,31, 33, 34, 36,37,38, 40]); anxiety (n = 7: [29, 30, 32, 33, 37, 38, 40]); and quality of life (n = 7: [30,31,32, 34, 38, 40]). Two studies explored caregiver psychological wellbeing using a 3-item and single-item measure, respectively [36, 39], whilst another captured PTSD symptomology [29] and a further study measured caregiver distress [37].

Overall changes

Significant improvements in at least one psychosocial outcome were reported in 9 of the studies [29, 31,32,33, 35,36,37,38,39]. These changes are reported below and described further in Table 3.

Table 3 Intervention description, outcomes, and results

Caregiver burden

Of the 9 studies reporting on caregiver stress/burden/strain, 5 evidenced significant reductions in burden/strain [32, 33, 36, 38, 39], with 3 of these studies [32, 36, 38] evidencing significant reductions in caregiver burden within the intervention group (IG) across time. One study [39] had no CG but highlighted significant reduction in caregiver strain across time. The remaining study [33] evidenced significant differences between IG and CG, with significantly lower levels of caregiver burden reported in the IG at follow-up.

Depression

Of the 9 studies reporting on depression, only one reported a significant reduction in depression within the IG across time, and also a significant difference in depression between IG and CG at follow-up, favouring the IG, i.e. the IG had significantly lower levels of depression when compared to the CG [37].

Anxiety

Of the 7 studies reporting anxiety, two reported significant reductions in anxiety within the IG across time [32, 37]. Two studies reported significant differences in anxiety between IG and CG, with significantly lower levels of anxiety evidenced in the IG post intervention [33, 37].

Quality of life (QoL)

Two of the 7 studies reporting on QoL reported significant changes among caregivers [31, 35]. One study reported significant improvement in IG QoL across time [35], with both studies reporting a significant improvement in QoL in the IG compared with the CG at follow-up [31, 35].

Other psychosocial outcomes

One study [37] reported significant improvements in self-efficacy in the IG across time, but also significant differences in self-efficacy between the IG and CG, with higher levels of self-efficacy reported in the IG. Results were mirrored in relation to distress, with levels of IG distress significantly lower across time and levels of distress significantly lower among IG compared to CG at follow-up [37]. One study reported changes in PTSD, with significant differences found between IG and CG, with the intervention seemingly having a detrimental effect on PTSD, with IG levels significantly higher than those reported by CG post intervention [29].

Intervention type

Details on intervention development, content, and components are reported in Table 3. Seven studies were described as psychosocial interventions [30, 32, 33, 36,37,38,39]. Four of these studies resulted in significant reduction in caregiver burden among IG participants [32, 33, 36, 38]. Law [39] reported significant reduction in caregiver strain among caregivers in their study across time. One study evidenced significant reduction in IG depression [37], with two psychosocial studies reporting significant reductions in IG anxiety [33, 37]. One psychosocial intervention reported significant improvement in self-efficacy, and significant reduction in distress, among IG caregivers [37].

Three interventions were psychoeducational in nature [31, 35, 41]. Significant improvements in IG QoL were noted in two psychoeducational interventions [31, 35].

The remaining three interventions were described as providing information, support, or structured provision for caregivers [29, 34, 40]. No significant changes were noted in psychosocial outcomes among participants in these studies.

Intervention development

Four studies report development of the intervention from an existing evidence base, relying on findings around caregiver need, caregiver intervention, and effective approaches [33,34,35, 39]. Of these studies, two reported significant reduction in caregiver burden/strain [33, 39], with another reporting a significant improvement in QoL among the IG [35]. Three studies cite clear theoretical frameworks guiding intervention development [31, 32, 41]. Again, two studies report a significant reduction in caregiver burden [32, 39], with another reporting significant improvement in IG QoL [31].

Two further studies were highlighted as theory-based interventions, but these studies based their intervention on an existing oncology model and intervention adapted for an advanced chronic illness population and caregivers [30, 38]. One study reported a significant reduction in caregiver burden [38].

Two studies described intervention development guided by both theory and evidence [36, 37]. One study reported significant reductions in caregiver burden [36]. The second study highlighted significant reductions in depression, anxiety, and distress among the IG, with significant improvement noted in self-efficacy among IG also [37].

One study reports a theory and evidence-based intervention for caregivers, adapted from a dementia care intervention [40]. The remaining study provides little detail on intervention development [29]. No significant positive intervention effects were reported for psychosocial outcomes in either of these studies.

Intervention delivery

Of the 7 studies involving patient and carer dyads [31, 33, 36,37,38, 41], five reported significant changes in psychosocial outcomes. Significant reductions in caregiver burden were noted in three of these studies [33, 36, 38]. Significant reductions in depression, anxiety, and distress, and significant improvements in self-efficacy, were recorded in the IG in one dyad study [37], whilst significant improvements in QoL were recorded in another [31]. Law [39], although not explicitly described as a dyad study, involves both patients and caregivers, with significant reductions in caregiver strain reported among caregivers.

The remaining studies [29, 30, 32, 34, 35] were aimed at caregivers only (Table 2). One study evidenced significant improvement in caregiver burden [32]. Another evidenced significant reduction in caregiver anxiety in the IG [32], with a further study reporting higher levels of QoL among the IG [35].

Outcomes associated with intervention delivery among the studies are mixed, with no clear conclusions able to be drawn in relation to intervention delivery methods, setting, duration, or frequency.

Discussion

Summary of findings

This review, to our knowledge, is the first to explore the effectiveness of interventions in improving psychosocial outcomes for caregivers of those with advanced, non-malignant, chronic conditions. The studies included in this review were rigorously assessed for risk of bias and quality. From the 8 RCTs in this review, 3 had an overall low risk of bias [29,30,31], whilst the remaining 5 had an overall unclear risk of bias, noted as ‘some concerns’ [32,33,34,35, 41]. The quality of all studies was assessed ahead of inclusion in the study. Given the quality of the included studies, we hope the findings will provide useful insight and guidance for future research in this area.

The aim of this review was to identify what interventions exist for caregivers of those living with advanced, non-malignant, chronic illness and to explore the efficacy of these. The review search strategy identified numerous records for screening and review, evidencing that a lot of research had been conducted to highlight the impact of caring on this cohort of carers via cross-sectional studies, and also to evidence their experiences via detailed qualitative work; however, this information has not been as readily translated into interventions for carers of those with advanced chronic illness [10, 17].

Twelve studies (n = 13 publications) fit the review inclusion criteria, despite the incidence of advanced chronic illness within our aging population. This perhaps supports the calls within the literature for improved provision [18] and the need for tailored, culturally appropriate, psychosocial interventions for this population [10, 17].

The findings highlight some trends, seemingly associated with improved psychosocial outcomes among caregivers for those with advanced chronic illness. Intervention delivery yielded mixed results; however, interventions reported in the review tended to most commonly be delivered face-to-face, at home. Given the involvement of dyads in these interventions (n = 6), delivery in the home may be reflective of preference of patients at end-of-life to be cared for, and to die, at home [42]. Although significant changes were reported for dyad interventions and interventions for carers only, dyad interventions were more readily associated with positive changes in outcomes for caregivers [31, 33, 36,37,38]. Perhaps unsurprisingly, psychosocial interventions were more likely to see positive changes in psychosocial outcomes for caregivers, with 6 psychosocial interventions evidencing improvements in caregiver outcomes including caregiver burden/strain, depression, anxiety, self-efficacy, and distress [32, 33, 36,37,38,39]. Psychoeducational interventions did not significantly improve psychosocial outcomes, except for QoL [31, 35].

Excluding those adapted from other interventions, some studies pointed to intervention development informed by evidence [33,34,35, 39], theory [31, 32, 41], or both [36, 37]. Although findings in relation to intervention development were mixed, the most effective intervention, in terms of the number of outcomes improved, was combined evidence based and theory driven [37].

Interventions adapted from other conditions, i.e. oncology and dementia [30, 38, 40] did not seem to translate the desired outcomes to the advanced chronic illness caregiver populations targeted, with only a significant within-group reduction in caregiver burden evidenced in one study [38]. It should be noted that the ENABLE CHF-PC intervention [30, 38] did undergo consultation to adapt to a new caregiver population, but translation was informed by the literature, expert consultation, and clinician input, with caregiver input only at the testing phase to assess acceptability and satisfaction [43]. Perhaps co-production of the adapted intervention, involving the target population of caregivers in the development and adaptation of the intervention, may have a greater impact [44].

The findings appear to support the case for interventions for this cohort of caregivers to be as follows: evidence-based, psychosocial, developed within an appropriate psychological framework, delivered at home, and involving the patient-carer dyad. However, it may be important to note that no explicit mention of caregiver involvement in the earliest stages of intervention development is outlined in the included studies. Given the growing body of literature in relation to co-production of caregiver interventions, this may be an important consideration which would improve the impact or effectiveness of interventions for caregiving populations [44, 45]. Interventions are likely to be most effective when targeted at the recognised needs of the caregiver population [42].

Another finding from the review recognises female caregivers outnumber males in all included studies, regardless of whether the patient being cared for is male or female. It may be important to consider gender-specific aspects in intervention development, e.g. around caregiver burden and coping styles [46, 47].

Limitations

There are some limitations to the current literature. All studies rely on the use of self-report measures to capture caregiver data at different points in time; however, measures used in the studies differed, with a lack of consistency in the measures used, making comparison across outcomes more difficult. Future studies should carefully consider the measures used to capture caregiver outcomes to ensure reliability and validity of findings. Some studies struggled with small sample sizes, which may have implications for statistical power, and although a potential limitation, it is not unexpected in studies focused on those with advanced illness and their caregivers [48]. This is often compounded by caregivers’ lack of recognition of their caring role, as some do not identify as a carer, or do not consider their own needs and wellbeing [3].

Optimal intervention duration and frequency of sessions need further exploration, with results from this review unable to shed any significant light on these aspects. It may be important to also consider that the evidence on the effectiveness of interventions for caregivers in cancer has seen substantial growth within the literature over the last two decades [42]. It is hoped this pattern will be mirrored for caregiver interventions in chronic illness, with research in this area seemingly still in its infancy.

Conclusion

The results of this review suggest that interventions for caregivers of those with advanced, non-malignant, chronic conditions can positively affect psychosocial outcomes among this population. However, the effects of these interventions are mixed, with some studies having greater impacts than others on caregiver burden, depression, anxiety, and quality of life. The interventions in this review vary in relation to design, delivery, duration, content, structure, and outcomes. It is clear information on interventions for caregivers for patients with advanced chronic illness is scant. Longitudinal studies, for example longer-term RCTs and observational studies, on intervention effectiveness over time, are needed to add to our understanding of efficacy and sustained impact. More sufficiently powered, robust, high-quality trials assessing the efficacy of interventions developed for use among this population are needed. Given the proposed growth in chronic illness, consideration should be given to increasing the reach and scalability of effective interventions for this cohort of caregivers, with online delivery offering an option for this. Given the isolation that can be experienced by caregivers, further research should explore the effectiveness of group interventions, as well as those targeting caregivers or dyads. With this in mind, this review suggests interventions developed for use and testing among caregivers of those with advanced, non-malignant, chronic illness should be as follows: evidence-based, developed within an appropriate theoretical framework, target both caregiver and patient dyad, delivered at home as these appear to hold more promise.