The impact of end-stage kidney disease (ESKD) on close persons: a literature review

Introduction

Kidney disease is defined as end-stage when a patient's glomerular filtration rate has fallen to <15 ml/min/ 1.73 m² [1]. Mortality associated with end-stage kidney disease (ESKD) is high [2]. The incidence of treated ESKD is rising in the western world, with a corresponding increase in the incidence of diabetes and cardiovascular disease, especially in ethnic minority groups. Survival on dialysis has been shown to be poorer in the older age group, especially in patients with increased comorbidity and in those whose functional status at the start of dialysis is poor [3]. Whilst renal transplant rates vary between different countries [4], the liberalization in the acceptance of older people into renal replacement therapy (RRT) programmes, together with changes in population demographics and the fact that kidney transplantation is less suitable for this group of older patients, means that dialysis may be the only treatment option available for an increasing number of patients aged 65 years and over [5]. Recent health policy changes in the OECD countries [6,7] acknowledge that end of life care may be more appropriate for some of these people, and maximum conservative management programmes (where residual renal function is supported, haemoglobin levels maintained and symptoms relieved) have been introduced into many renal units, particularly in the United Kingdom [6].

The onset of ESKD and subsequent recommendation of dialysis as a treatment option involves a change in lifestyle for both patients and close persons [8]. Even before end-stage disease is reached, as renal function deteriorates, patients frequently require additional support, and it is often family members who provide this [9]. In the UK it is estimated that 9 out of 10 carers of patients with either physical or neurological disabilities will be close relatives. In particular when home haemodialysis is undertaken, family members have been involved in supporting patients [10,11]. Studies have shown that good family support is associated with successful adaptation to dialysis and compliance with dietary restrictions [12,13]. Conversely, one of the main factors associated with patients discontinuing dialysis is patients’ perception that they have become a ‘burden’ to close family members [14]. There is therefore a need for health professionals to be aware of the important contribution that close persons make to the care of renal patients, to communicate effectively with them and to provide bereavement support for this group when appropriate [15].

The literature on close persons of patients with renal disease has identified two main areas of impact. Firstly, both haemodialysis and peritoneal dialysis may have a disruptive influence on family members’ social lives [16] and the structure of the week may be geared towards dialysis sessions. Secondly, some patients become frail and lose functional independence, leaving family members to provide greater physical support. Family members may have health and social care needs of their own that need to be addressed [16,17]. Qyinan [18] reported that close persons commonly felt overwhelmed and stressed, although this review was limited to an evidence base of four articles and considered home dialysis only. Campbell [19] used findings from the general carer literature to illustrate demands of ageing partners with ESKD. In other chronic illnesses such as stroke [20] or in palliative care for cancer and mental health [21], interventions aimed at providing family members with training to support patients with their rehabilitation, or to address unmet needs as a result of the patient's illness, have been developed and evaluated. Results have been mixed. In the case of stroke, carers in the intervention group experienced less depression and anxiety and better quality of life [20], whilst in the palliative care study, no statistically significant differences were found between the intervention and control group on carers’ psychological outcomes [21]. However, before such interventions can be developed in ESKD, it is important to understand better the emotional and physical needs of close persons.

This review aims to identify all studies involving close persons caring for ESKD patients, to describe the main findings and critique the methodology. Specific attention has been paid to (a) studies exploring the impact of ESKD on close persons, in particular for those close persons where the patients are either withdrawing from dialysis or being provided with end of life care and (b) studies looking at the provision of health care for close persons.

Methodology

Search strategies

A literature search for relevant articles was conducted in five databases: Medline (1950–2006), Embase (1991–2006), CINAHL (1982–2006), PsycINFO (1970–2006) and AMED (1985–2006), employing the following key words: carers, caregivers, end-stage kidney disease, end-stage renal disease, haemodialysis, peritoneal dialysis and renal replacement therapy. These keywords were used both in word search options and exploded as thesaurus terms to obtain the maximum number of articles. The abstracts for each article were read to check for inclusion into the main review, using the following criteria:

1. Published in peer-reviewed journals.

2. Research studies with an introduction, a methodology and results section and a conclusion.

3. Involve close persons, defined as either a family member or the person identified by the patient as an informal carer. By an informal carer, we mean a person who provides the majority of a patient's physical and emotional care needs and who is neither a volunteer nor in the employment of statutory services.

4. Use a sample of close persons caring for adult ESKD patients (over 18 years).

5. Non-English language articles were considered if the English translation of the abstract met the above criteria.

Using these criteria, 334 articles were identified from the five databases (139 in Medline, 121 in Embase, 80 in CINAHL, 8 in AMED and 34 in PsycINFO). J.L. went through the abstracts of each of the 382 articles, of which 37 initially met the inclusion criteria. One was later excluded on closer inspection, as it was specifically a validation study of a fatigue severity scale. Of the remaining 36 studies, 16 exclusively looked at family members, 12 specifically at the patient-family dyad and 3 at the family-health professional dyad. Whilst the latter two types of studies did not concentrate solely on family concerns, we decided to include them in the analysis, because these findings further contribute to the limited number of studies in this field. Thirty-six studies were included in the review.

Both J.L. and G.S. first went through the remaining 36 studies independently and extracted the following information for each study: the number of carers in the study sample, the RRT population they were caring for, authors’ definition of a carer, demographic details of the sample, patients’ dialysis history, caring history, study design, outcome measures used and main findings. J.L. and G.S. then met together to discuss these findings and obtain an initial consensus before meeting with A.B. and L.J. to obtain a final consensus.

Results

Initial exploration of the 36 reviewed studies

We undertook an exploration of the aims of these studies, their study design, the sample of participants and the outcome measures highlighted in the quantitative studies.

Main aims of the reviewed studies

The three main themes explored were (a) the impact of caring for a patient with ESKD on dialysis on close persons, in particular quality of life, psychological morbidity, close person responsibilities—which authors often referred to as ‘burden’ or ‘carer burden’—and their life situation; (b) the coping strategies employed by these close persons and (c) factors that influence psychological morbidity. No studies of health provision for close persons of patients with ESKD were identified.

Four studies looking at end of life issues explored the following themes for close persons: (a) their perceptions of patients’ terminal symptoms; (b) their reasons for why patients decided to stop dialysis; (c) the long-term impact of patient death following dialysis cessation and (d) their perceptions of advance directives. End of life care may be provided by the renal multi-disciplinary team alone, or it may involve referral for specialist palliative care advice. Such advice is likely to include symptom control, attention to spiritual and psychological issues for patients and, where possible, involvement of their families in decision-making.

Whilst most studies were interested in the direct impact on close persons only, one triangulated close person data with patient data.

Study designs

The majority of studies reviewed used a cross-sectional design; there was only one longitudinal study. Whilst most were quantitative (24/36), there were some qualitative studies (11/36) and one used a mixed methods approach.

Sample

The total sample was predominantly female, with mean ages ranging from 41 to 68 years (analysis possible in only 18 studies). Sample sizes also tended to be small, with a median sample of 55 participants for the quantitative and 15 participants for the qualitative studies.

Most of the sample was recruited in studies where associated patients were undergoing haemodialysis or peritoneal dialysis. Three studies also involved kidney transplant patients. Only five studies looked at close persons dealing with end of life issues or with patients withdrawing from dialysis. Many studies did not focus on close persons in their potential role as ‘informal carers’. Thirteen studies actively sought close persons who also considered themselves to be informal carers, of which eight provided full definitions of what they meant by this term.

All studies included spouses as part of their sample, of which eight specifically concentrated on this group alone. Adult children were included in seven of these studies and parents in five.

Outcome measures (quantitative studies)

A wide variety of outcome measures were used to rate health-related quality of life, anxiety, depression, coping strategies and patient disease severity. Some studies used standardized outcome measures; others used simple self-rated tools.

The most commonly used standardized measures were the Zarit Burden Interview [22] to evaluate the sense of carer responsibility (3/8), the Beck Depression Scale [23] to evaluate depression (2/5), SF-36 [24] to evaluate health-related quality of life (3/8), Jalowiec Coping Scale [25] to evaluate close persons’ use of coping strategies (3/3) and End-Stage Renal Disease Severity Index [26] to evaluate patients’ disease severity (2/4).

Country of origin

A breakdown of the country of origin for each study showed that over half originated from either the USA (11/36) or Canada (7/36), with five originating from Australia and only seven from the European Union, of which only one was conducted in the UK. The remaining six studies came from the following countries: Japan (2/36), Brazil (2/36), China (1/36) and Turkey (1/36).

Main findings

The 36 studies have shown mixed results. They have mainly explored the following areas associated with caring for an ESKD patient: the impact on close persons and their social life and the factors affecting close persons’ psychological health. There have been very few studies looking at palliative care issues and these have primarily concentrated on dealing with end of life issues rather than the provision of supportive care in the pre-terminal phase.

Family life (Tables 1 and 2)

Table 1.

Impact on family life (quantitative and mixed methods studies)

Authors	No. of carers	Population involved in the study	Carer definition	Demographic details (1) mean age (years) (2) female (%) (3) spouses (%) (4) in employment (%) (5) mean patient dialysis duration (years) (6) mean length of caring (years)	Country of origin	Study design	Outcomes (carer) (a) ‘burden' (b) disruptions to daily life (c) family adjustment (d) quality of life (e) psychological (f) patient-related (g) miscellaneous	Main findings
Devins et al. [42]	19	Dialysis and transplant  spouses	Not defined	(1) 41 (2) 52 (3) 100 (4) 68 (5), (6) not reported	Canada	Cross-sectional	(b) IIRS (c) FES (a), (d), (e), (f), (g) N/A	Spouses of transplant recipients more assertive,   self-sufficient and able to make own   decisions compared with dialysis   spouses.
Ekelund et al. [41]	35	Home HD  spouses and  patients	Not defined	(3) 100 (5) 8 (1), (2), (4), (6) not reported	Sweden	Longitudinal	(a) self construct (b), (c), (d), (e), (f), (g) N/A	Spouses report major sexual problems.
Kaye et al. [35]	21	HD and CAPD  family  members  and patients	Closest person with  whom the patients  lived with	(5) 1.5 (1), (2), (3), (4), (6) not reported	USA	Longitudinal.  follow-up at  7–14 months	(c) FES (f) PAF (physical); PAIS (psychosocial) (a), (b), (d), (e), (g) N/A	Family members show a loss in independence  and a reduction of recreational activities  on follow-up.
Lindqvist et al. [33]	55	HD, CAPD  and  transplant  spouses	Not defined	(1) 56 (2) 62 (3) 100 (4), (5), (6) not reported	Sweden	Cross-sectional  (mixed)	(d) SwQoL (e) VAS (coping) (a), (b), (c), (f), (g) N/A	HD spouses less able than the other two  groups in handling physical, social and  existential aspects of illness CAPD and  HD spouses had lower QoL than  age-matched general population.
Page and Weisberg [43]	37	Family  members  and patients	Not defined	(1) 42 (3) 43 (4) 35 (5) 1.5 (2), (6) not reported	USA	Cross-sectional	(c) FES; self-construct (family cohesiveness), MAES (d) self-construct (mental) (a), (b), (e), (f), (g) N/A	Home partners encouraged more direct  expression of feeling, although less active  in recreational pursuits Home partners  were more sensitive towards patient needs  to feel in control and to be given more  attention.
Shimoyama et al. [28]	34	CAPD family  carers	Family defined as a  unit of two or more  carers, are mutually  related through  emotional intimacy  and conscious of  being a family	(2) 59 (3) 56 (4) 68 (5) 6 (6) 3.9 (1) not reported	Japan	Cross-sectional	(a) ZBI (d) SF-36 (b), (c), (e), (f), (g) N/A	Carers had poorer mental health and social  functioning than general population ‘carer  burden’ associated with increasing age and  decreasing health-related QoL.
Wagner [44]	10	Family  members  and nurses	Not defined	(1) 32–49 (2) 90 (3) 60 (4), (5), (6) not reported	USA	Cross-sectional	(g) modified NGQ (family needs) (a), (b), (c), (d), (e), (f) N/A	Families identified most important needs as  information about patient care and  assurances that this contributed to patient  comfort.
Wicks et al. [27]	96	Transplant  family  carers	Person who patients  could depend on  to care for them  if they could no  longer care for  themselves	(1) 47 (2) 76 (3) 62 (4) 67 (5) 1.6 (6) 1–46	USA	Cross-sectional	(a) CBI (d) General QoL (g) self-construct (general health) (b), (c), (e), (f) N/A	Carer QoL associated with ‘carer burden’,  their self-reported health and employment  status. Family carers reported both little  ‘burden’ and excellent QoL.

CAPD = Continuous Ambulatory Peritoneal Dialysis; CBI = Caregivers Burden Interview; FES = Family Environment Scale; HD = Haemodialysis; IIRS = Illness Intrusiveness Ratings Scale; MAES = Marital Attitudes Evaluation Scale; NGQ = Norris & Groves Questionnaire; PAF = Physicians’ Assessment Form; PAIS = Psychosocial Adjustment to Illness Scale; SF-36 = Short-Form 36; SwQoL = Swedish Health-Related Quality of Life Survey; VAS = Visual Analogue Scale; ZBI = Zarit Burden Interview.

Table 2.

Impact on family life (qualitative studies)

Authors	No. of carers	Population involved in the study	Carer definition	Demographic details (1) mean age (years) (2) female (%) (3) spouses (%) (4) in employment (%) (5) mean patient dialysis duration (years) (6) mean length of caring (years)	Country of origin	Study design	Themes explored	Main findings
Beanlands et al. [11]	37	Home PD, home HD, hospital HD and in-centre self care HD carers	Family member or friend responsible for providing most of dialysis and health care at home	(1) 58 (2) 68 (4) 49 (5) 3.9 (3), (6) not reported	Canada	Longitudinal follow-up at 6 months	(1) exploration of carers’ abilities and activities	Carers ‘grow into the job’ with their caring skills developing over time The amount of care that carers’ need to give patients is inversely related to patients’ ability to self-care for themselves. Carers play pivotal role in supporting patients in the management of their disease.
Friesen [32]	8	Home HD spouses	Not defined	(1) 33–66 (2) 100 (3) 100 (4) 63 (5), (6) not reported	Canada	Cross-sectional	(1) experience of being a carer (2) factors influencing response to living with home HD partner	Lack of socialization opportunities due to the lack of time due to patient poor health Marital relationship strengthened if spouse considered their involvement as cooperative.
Luk [39]	30	Home HD carers	Person responsible for providing the majority of the carer tasks without pay	(1) 68 (2) 70 (3) 57 (6) 1.9 (4), (5) not reported	China (HK)	Cross-sectional	(1) impact of home dialysis on carers’ life and relationships with extended family, friends and community (2) carers’ health status	Carers feel exhausted and withdraw from job market and social life to accommodate caring duties. Carers want opportunities to discuss problems with staff, but feel informational support is inadequate.
Pelletier-Hibbert and Sohi [29]	41	HD/CAPD family members	Not defined	(1) 55 (2) 85 (3) 73 (4) 56 (5), (6) not reported	Canada	Cross-sectional	(1) stressful situations around renal failure and dialysis treatment (2) management of stress sources	Main source of carer stress identified as uncertainty about patients’ declining health, the lack of spontaneity pursuing family activities and the after-effect of dialysis treatment. Carers cope by living each day as it comes and by finding meaning in illness.
Wellard and Street (1999) [36]	3 families	Home HD family members and patients	Not defined	(1), (2), (3), (4), (5), (6) not reported	Australia	Cross-sectional	(1) exploration of families’ experiences of living with home-based dialysis (2) issues around restructuring family life around treatment	Family members report social isolation due to strict timetabling of treatments. Female family members not asked about their willingness to assume the caring role. Family members report a lack of effective support from health services.
White and Grenyer [30]	22	HD and CAPD (home and hospital) partners and patients	Support member for someone on dialysis	(1) 62 (2) 55 (4) 23 (5) 3.5 (3), (6) not reported	Australia	Cross-sectional	(1) relationship between patient and partner dyad (2) perception on life situation	Partners expressed positive aspects of relationship, though recognizing its negative impact.
White, Richter, et al. [38]	Number involved unclear	HD carers, patients and additional family members	Not defined	(5) 2–5 (1), (2), (3), (4), (6) not reported	USA	Longitudinal	(1) impact of illness stressors on the family (2) resources associated with the dialysis experience	Family resilience achieved through social support and by maintaining a sense of “normalcy” through open communication between family members, keeping busy, maintaining a sense of control and finding meaning in illness.
Wright and Kirby [37]	5	CAPD family members, patients and health professionals	Not defined	(1) 56 (3) 80 (5) 0.7 (2), (4), (6) not reported	UK	Cross-sectional	(1) patient adjustment to ESKD	Family members agreed that patients had adjusted to their illness but were frustrated by the life style restrictions.
Ziegert et al. [31]	13	HD spouses	Not defined	(1) 61 (2) 85 (3) 100 (6) 2.7 (weighted) (4), (5) not reported	Sweden	Cross-sectional	(1) health experiences in everyday life (2) how spouses maintain good health	Three types of experiences identified: (1) arduousness experienced when spouses’ lives dominated by caring; (2) stamina reflects spouses’ willingness to help at the expense of own health and (3) independence acknowledges the importance of spouse caring for themselves which in turn benefited their own health.
Ziegert and Fridlund [34]	12	HD family members	Not defined	(1) 46–65 (2) 75 (3) 42 (5) 4 (4), (6) not reported	Sweden	Cross-sectional	(1) impact of ESKD on emotional, behavioural, work, biophysical, existential and family situation	Family members’ daily life affected as result of feeling confined (limited social activity) and fear of patients’ death Family members recognize the potential impact that they have in promoting patients’ well-being.

CAPD = continuous ambulatory peritoneal dialysis; HD = haemodialysis; PD = peritoneal dialysis.

In only one study, close persons rated their quality of life as excellent and reported few pressures resulting from their carer responsibilities [27], whilst in all others, ESKD and dialysis were shown to increase the close person's sense of responsibility and lead to a poorer quality of life when compared with age-matched controls [28]. Close persons found living with an ESKD patient on dialysis stressful [29] and experienced increased fatigue [30]. The dominating effect of caring for an ESKD patient often led close persons to neglect their own health. For those who took time to have a break from their carer responsibilities, there were health benefits [31]. Other issues that close persons reported included isolation through the loss of social activity [30–36], life restrictions [36–38], increased workload, negative economic consequences [39,40], changed relationship with the patient [30,34] and sexual problems for spouses [41].

The treatment modality may also have an impact on family members. Studies have highlighted that spouses of transplant patients were more assertive, self-sufficient and able to handle the physical, social and existential aspects of the illness better than dialysis spouses [33,42]. Despite these pressures, close persons recognized that they play a positive role in promoting patients’ well-being [11,43]. They recognized that health care professionals were important in providing support to discuss their problems [39] and would like to have more information about the care being provided to patients [44]. However, they also reported poor communication with professionals, felt that their needs were not always addressed [30,39,45] and felt themselves uneducated and poorly equipped to deal with the regimented lifestyle associated with the dialysis regimen [14].

Caring and psychological Health (Table 3)

Table 3.

Caring and psychological health

Authors	No. of carers	Population involved in the study	Carer definition	Demographic details (1) mean age (years) (2) female (%) (3) spouses (%) (4) in employment (%) (5) mean patient dialysis duration (years) (6) mean length of caring (years)	Country of origin	Study design	Outcome (carer) (a) psychological morbidity (b) psychological (others) (c) ‘carer burden’ (d) quality of life (e) social support (f) patient outcomes (g) miscellaneous	Main findings
Alvarez-Ude et al. [46]	221	HD and PD carers and patients	Family member identified as mainly responsible for looking after the patient and most closely involved in their care	(1) 57 (2) 77 (4) 34 (6) 5 (3), (5) not reported	Spain	Cross-sectional, quantitative	(c) ZBI (d) SF-36 (e) DFSS (f) BI (functioning) (a), (b), (g) N/A	Carers’ mental health was worse in those with lower social support, higher subjective ‘burden’ of work on those caring for patients with poor mental health.
Asti et al. [49]	65	CAPD carers and patients	Not defined	(1) 44 (2) 82 (3), (4), (5), (6) not reported	Turkey	Cross-sectional, quantitative	(a) revised UCLALS, BDI (e) PSSFFS (b), (c), (d), (f), (g) N/A	Most carers were not depressed and had high levels of perceived social support. Depression was inversely associated with the level of perceived family social support.
Belasco et al. [57]	201	In-centre HD and home PD carers	Person mainly responsible for both looking after patient during course of disease and caring for them	(1) 42—58 (2) 80 (4) 32 (5) 1.8–3.8 (6) 1.5–3.9 (3) not reported	Brazil	Cross-sectional	(a) CBS (d) SF-36 (e) CID (f) KPS (b), (c), (g) N/A	PD carers had worse mental health than HD carers. Carers of younger patients had potentially poorer access to transport and health services. Burden for PD carers associated with their mental composite scores (SF-36), whereas burden for HD carers associated with vitality, pain and social aspect.
Belasco and Sesso [50]	100	HD carers and patients	Person responsible for looking after the patient during the course of the disease and most closely involved in caring for the patient for a period of >3 months	(1) 47 (2) 84 (3) 49 (4) 34 (5) 2.8 (6) 2.5 (median)	Brazil	Cross-sectional, quantitative	(c) CBS (d) SF-36 (a), (b), (e), (f), (g) N/A	‘Carer burden’ correlated with the number of patient comorbidities and the length of time spent as carers. Perceived ‘carer burden’ was explained by carers’ mental health, patient vitality, relationship type and carer pain.
Binik et al. [56]	89	Pre-dialysis, HD and post-transplant spouses	Not defined	(3) 100 (1), (2), (4), (5), (6) not reported	Canada	Cross-sectional, quantitative	(a) GSI, ABS, SEI (b) MRQ, KDS-15 (marital role strain), L-WMAT (marital adjustment) (c) self-construct (objective and perceived intrusiveness) (d), (e), (f), (g) N/A	Perceived intrusiveness of ESKD significantly associated with greater ‘marital role’ strain, poorer marital adjustment and decreased individual well-being of both the spouse and the patient. Increased symptoms of psychopathology significantly associated with perceived intrusiveness into marital and non-marital aspects of life and increased marital strain.
Blogg et al. [53]	61	Home HD carers	Person who cares for a patient at home	(1) 45–49 (2) 69 (3) 93 (5) 3.5 (4), (6) not reported	Australia	Cross-sectional, quantitative	(a) GHQ-28 (c) RSS (b, (d), (e), (f), (g) N/A	Main variables associated with carer distress were younger age (<45 years) and having a low involvement in the dialysis procedure Other associated variables include no prior experience of the dialysis process and high involvement levels in a rural setting.
Courts [47]	14	HD partners and patients	Not defined	(1) 47 (median) (2) 93 (3) 71 (4), (5), (6) not reported	USA	Cross-sectional, quantitative and qualitative	(a) CAS, STAI, GCS (b) PAIS-SR (psycho-social adjustment), HSS (HD stressor) (c) CBS (d) SF-36 (g) description of HD experience (interview) (e), (f) N/A	Most home dialysis partners had little psychological distress. However, partners perceived home HD to be stressful but preferred HD at home.
Daneker et al. [55]	55	HD carers and patients	Not defined	(1) 52 (2) 76 (4) 34 (5) 2.9 (3), (6) not reported	USA	Cross-sectional, quantitative	(a) BDI, CBI (b) JCS (coping); DAS, DAS-S (marital adjustment) (d) QoLI (e) MSPSS (f) KPS (functioning), ESRDSC (comorbidity) (c), (g) N/A	Spousal depression inversely correlated with their marital dissatisfaction but correlated with patient depression. Spouses’ perceived social support inversely related to both spousal depression and spousal marital dissatisfaction. Spouses’ psychosocial variables not correlated with patients’ disease severity.
Dunn et al. [54]	38	CAPD spouses	Not defined	(1) 58 (2) 58 (3) 100 (5) 2.6 (4), (6) not reported	USA	Cross-sectional, quantitative	(b) JCS (coping); DAS (marital adjustment) (d) QoLI (f) ESRDSI (disease severity) (a), (c), (e), (g) N/A	QoL influenced by marital adjustment and income. Emotional coping was negatively correlated with marital adjustment but positively correlated with years on dialysis.
Ferrario et al. [51]	50	Carers and patients	The main person responsible for patient care outside the hospital	(1) 54 (2) 80 (3) 70 (4) 26 (5) 4.7 (6) not reported	Italy	Cross-sectional, quantitative	(a) STAI; DQ (b) EPQ (personality) (c) FSQ (f) SWLS (life satisfaction) (d), (e), (g) N/A	Carers’ sense of burden and satisfaction correlated with their neuroticism, anxiety and depression.
Harris et al. [48]	78	Transplant family members	Person identified as providing assistance to patient who is unable to self-care	(1) 21–88 (2) 76 (3) 53 (4) 70 (5) 2.3 (6) not reported	USA	Cross-sectional, quantitative	(c) ZBI (a), (b), (d), (e), (f), (g) N/A	Most carers reported little ‘burden’. Both personal strain and role strain were lowest in carers with patients who were independent in ADL.
Piira et al. [52]	38	HD and PD carers and patients	Person who plays a significant role in the dialysis process and in caring for the patient	(1) 54 (2) 56 (3) 68 (5) 2 (4), (6) not reported	Australia	Cross-sectional, quantitative	(a) DASS (b) JCS (coping); LCB (f) ESRDSI (comorbidity), SIP (functioning) (c), (d), (e), (g) N/A	Carers’ psychological morbidity was positively associated with the use of external locus of control and emotion-focused coping strategies.
Rideout et al. [40]	40	CAPD spouses and patients	Not defined	(1) 51 (2) 65 (3) 100 (5) 0.2 (4), (6) not reported	Canada	Cross-sectional, quantitative	(a) CES-D (b) IFS (family impact); DAS, DAS-S (marital adjustment) (e) PSS (f) SIP (functioning) (c), (d), (g) N/A	Whilst most spouses were not depressed, the lack of perceived social support from their ill partner and social/financial changes were major significant predictors of spouse depression.

ABS = Affect Balance Scale; BDI = Beck Depression Inventory; BI = Barthel Index; CAPD = Continuous Ambulatory Peritoneal Dialysis; CAS = Clinical Anxiety Scale; CBI = Caregivers Burden Interview; CBS = Caregiver Burden Scale; CES-D = Center for Epidemiologic Studies Depression Scale; CID = Cognitive Index of Depression; DAS = Dyadic Adjustment Scale; DAS-S = Dyadic Satisfaction Sub-scale; DASS = Depression, Anxiety and Stressor Scale; DFSS = Duke-UNC Functional Social Support Questionnaire; DQ = Depression Questionnaire; EPQ = Eysenck Personality Questionnaire; ESRDSI = End-Stage Renal Disease Severity Index; ESRDSC = End-Stage Renal Disease Severity Coefficient; FSQ = Family Strain Questionnaire; GCS = Generalized Contentment Scale; GHQ-28 = General Health Questionnaire 28; GSI = Global Symptom Index; HD = haemodialysis; HSS = Haemodialysis Stressor Scale; IFS = Impact on Family Scale; JCS = Jalowiec Coping Scale; KPS = Karnofsky Performance Status Scale; LCB = locus of control of behaviour; L-WMAT = Locke–Wallace Marital Adjustment Test; MRQ = Marital Role Questionnaire; MSPSS = Multidimensional Scale of Perceived Social Support; PAIS-SR = Psychosocial Adjustment to Illness Scale—Self Report; PD = peritoneal dialysis; PSS = Perceived Social Support Scale; PSSFFS = Perceived Social Support for Friends and Family Scale; QoLI = Quality of Life Index; RSS = Relatives’ Stress Scale; SEI = Self Esteem Inventory; SF-36 = Short-Form 36; SIP = Sickness Inventory Profile; STAI = State Trait Anxiety Inventory; SWLS = Satisfaction with Life Scale; UCLALS = UCLA Loneliness; ZBI = Zarit Burden Interview.

Whilst some studies have shown that close persons display few signs of psychological distress [40,46–49], others have identified the following:

1. A negative association between close persons’ psychological health and their sense of carer responsibility [28,46,50,51], their use of emotionally focused coping strategies (i.e. strategies that reduce the symptoms of stress without addressing the source of the stress) [33,52], the close person's age [53] and the social [40] and financial changes [40,54] imposed as a result of ESKD onset.

2. A positive association between good mental health and the following factors: low marital strain [55,56], a lack of perceived intrusiveness of dialysis [56], the type of dialysis patients are on [57], availability of social support [40,46,49] and reduced sense of carer responsibility [46].

3. The sense of carer responsibilities are lower if patients are independent in activities of daily living (ADL) [48], have less severe dialysis-linked complaints [41] or lower comorbidity [29,50]. This was further emphasized in studies in patients receiving home haemodialysis or awaiting transplant [33,47,48] where close persons not experiencing anxiety or depression [47] felt less troubled by their carer responsibility [48] and had a comparable quality of life to the age-matched population [33]. Further associations were found between higher levels of responsibility and other outcomes such as quality of life [27] and neuroticism [51].

4. The use of emotionally focused coping was found to have a negative correlation with marital adjustment, but a positive correlation with the number of years on dialysis [54].

5. Close persons were more likely to have negative feelings towards patients if they had no prior experience of the dialysis process and had a high level of involvement with the caring process whilst living in a rural environment [53].

End of life and end-stage kidney disease (Table 4)

Table 4.

End of life care

Authors	No. of carers	Population involved in the study	Carer definition	Demographic details (1) mean age (years) (2) female (%) (3) spouses (%) (4) in employment (%) (5) mean patient dialysis duration (years) (6) mean length of caring (years) (7) mean post-dialysis survival time (days)	Country of origin	Study design	Main study aims for carers	Main findings
Ashby et al. [14]	5	Family members, patients and health professionals	Not defined	(3) 100 (1), (2), (4), (5), (6), (7) not reported	Australia	Qualitative, cross-sectional	Reasons on decisions to stop dialysis	Family members did not feel adequately educated or equipped by hospital to prepare for lifestyle associated with dialysis regimen.
Cohen et al. [58]	86	HD carers	Not defined	(3) 49 (1), (2), (4), (5), (6), (7) not reported	USA	Quantitative, cross-sectional	To explore the perceptions of bereaved family members about patients’ terminal symptoms and the quality of their end of life care	Family members perceived that most patients were in pain in last week of life. Fatigue was reported as second most disturbing symptom Death at home was associated with less pain, more peace and completion of a living will. Those who completed a health care proxy felt it had helped them ensure that patients’ wishes were followed.
Miura et al. [59]	398	Family members, patients and health professionals	Not defined	Not reported	Japan	Quantitative, cross-sectional	To assess how accurately family members predict patients’ wishes about medical care and treatment under various medical scenarios	Family members were poor at assessing both patients’ current and future preferences for cardiopulmonary resuscitation or dialysis continuation.
Perry et al. [60]	46	Carers	Not defined	(1) 53 (2) 63 (5) 5.4 (3), (4), (6), (7) not reported	Canada	Quantitative, cross-sectional	To describe the opinions of their family members/friends regarding advance directives	Relatives/friends felt that routine visit to the health professional was the best time to discuss advance directives, though over half felt that the final decisions about major treatments should be left to physicians.
Phillips et al. [45]	26	Family members	Not defined	(1) 47 (3) 27 (5) 3 (7) 8 (2), (4), (6) not reported	USA	Quantitative, cross-sectional	To explore the long-term impact on families of deaths preceded by cessation of dialysis	Most family members showed little long-term psychological distress, though principal carers and spouses had significantly higher intrusive thoughts. Good death was perceived as being pain free, with the patient being mentally alert, able to remain at home and dying while asleep. Many felt that health professionals had not provided sufficient information.

HD = haemodialysis.

We identified five studies that explored end of life issues, but only one focused specifically on close persons. This study examined the long-term impact of death occurring on families when dialysis was discontinued. It found that most families felt that patients had a good death (defined as dying at home, pain free and with close people present) and most family members showed only low levels of distress. However, principal carers and spouses, i.e. those with potentially higher levels of caring responsibility for the patient, did report significantly higher intrusive thoughts [45].

The remaining four studies focused on patient outcomes such as reasons for withdrawing from dialysis [14], the quality of death [58], the accuracy of close persons in predicting ESKD patients’ preference to present and future medical care [59] and use of advance directives (defined as a legal document, prepared in advance by patients, which specifies the course of treatment to be taken by health care providers once a patient is unable to provide informed consent due to that person's lack of capacity) [60]. These studies found that close persons thought that whilst most patients had a peaceful death, many patients were perceived to be in pain and suffering from fatigue [58]. Health care proxies supported family members in ensuring that patients’ last wishes were followed and most relatives recognized the importance of living wills, which they felt were best discussed during routine health care professional visits [60]. Close persons were poorly prepared for caring for a dying dialysis patient, and patients did not want to be a burden on their families [14]. They were also poor at assessing both patients’ current and future preferences for cardiopulmonary resuscitation or dialysis continuation [59]. No studies were identified for carers of patients with supportive care needs, e.g. those on maximum conservative management programmes.

Methodological critique

General

The main criticism of most studies included in this review is the lack of detail and analysis of the demographic information. Just over half of the studies give demographic details of close persons’ ages and gender or details of their relationship to patients (19/36). Reporting of other relevant demographic variables such as employment status (13/36), ethnicity, educational level or social class is sporadic and only seven studies report the time that close persons had spent as carers.

Whilst all studies recorded the type of dialysis received, half (18/36) had reported the length of time a patient has been on dialysis, with very few recording patients’ functional status, although carer studies in other specialities suggest that this has an impact on carers’ quality of life [61]. In total, only 18 quantitative, 9 qualitative and 1 mixed methods study gave clear details of their sample sizes.

Quantitative studies

Nineteen of the 20 quantitative studies included in the analysis used cross-sectional designs and were predominantly descriptive. Such studies are useful in exploring associations between variables but are limited as they do not allow causal relationships to be established. As only five studies provide the full details of sample recruitment, it is therefore not possible to assess response rates or how representative the sample was. The sample sizes were generally small with a median of 55 participants, and details of statistical power were reported in only one study. Finally, the diversity of the tools used to assess the different outcomes makes it difficult to do any meaningful comparison between the different studies.

Qualitative studies

We identified that most studies gave poor descriptions of their sampling methods and a lack of detail of their data analysis process. Most used opportunistic samples of family members (6/11), with five studies using purposive strategies, but only one gave specific details of their sampling method. Seven studies gave detailed descriptions consistent with their analysis plan. Two studies gave vague descriptions consistent with their plan; one gave a description of their analysis but without stating their analysis approach. One study gave no details of their analysis.

Overall, the studies were poorly written, meaning that the writing style was unclear and overly complicated. Several authors were unsure as to how to use qualitative data effectively, with the depth and richness of the qualitative data lost amongst under-analysed demographic detail, long explanations of data collection scales and poor use of supporting quotes.

Conclusion

This study reviewed the current literature exploring the experience of close persons of patients with ESKD, from which several points can be made regarding the quality of past research and recommendations for the direction of future research in this area.

Firstly, studies exploring end of life issues are limited, with only four identified. Although these four studies are mostly patient-centred, they do suggest that health professionals’ routine visits are the best time for informal carers to discuss patients’ preferences about end of life care and, additionally, report little long-term distress of close persons resulting from the patients’ deaths. However, little remains known about how such end of life issues affect close persons, and we recommend future studies in this area.

Secondly, definitions of close persons were problematic since there were some subtle, but fundamental differences identified between the different groups. While the current literature supports the impression that most close persons are closely related to patients, the situations in relation to patients who were single and those with alternative family and relationship arrangements were not explored. Furthermore, not all close persons decide to take on the role of the informal carer, and the development of this role can be a gradual process as illness progresses. Although our review did not identify any differences between informal carers and family carers in the outcome measures used (we included both groups under the umbrella term of ‘close person’), differences may exist in their perceived role, with ‘informal carers’ having a more active role, whereas ‘family members’ may be perceived as being more passive. Again, this needs further investigation.

Thirdly, serious issues that need to be addressed concerning the use of methodology were identified in the literature. Most quantitative studies were exploratory and descriptive, with no intervention studies being identified. Studies had sporadic reporting of demographic details and response rates and all were cross-sectional with small sample sizes that used standardized measures. Whilst the use of standardized measures increases the robustness of the results, the findings from cross-sectional studies are limited in that they can only establish associations between variables. We recommend the use of longitudinal methods in future research that would ensure that causality between key variables could be explored in more depth. This would also enhance insight into the long-term experience of close persons. In those studies using qualitative methods, most used opportunistic samples, with poor reporting and understanding of demographic details, sampling methods, analysis and social and cultural context. Many authors used the term ‘burden’ or ‘carer burden’ without questioning how such terms may influence our understanding of the relationship between close person and patient. It was not clear whether they were prompted by close persons to use the term ‘burden’ or whether it was a term coined by the authors to describe the responsibilities of caring. Such methodological and theoretical considerations affect the confidence of applying study findings beyond the original setting.

Fourthly, half of the studies originated from either the USA (11/36) or Canada (7/36). This North American dominance makes it difficult to generalize the results of these studies outside their context, due to variations in the organization of health care provision in other countries.

Lastly, while there is emerging evidence that suggests maintaining the psychological health of close persons enables them to continue to care effectively, which in turn can benefit the mental health of patients, we failed to identify studies in our review that looked specifically at how health services supported close persons of patients with ESKD. Our review has clearly identified shortcomings in the published literature that need to be filled. Future research is therefore needed to explore the relationship between health services and close persons in order to develop practical empowering interventions.