Reported burden on informal caregivers of ICU survivors: a literature review

Background Critical illness and the problems faced after ICU discharge do not only affect the patient, it also negatively impacts patients’ informal caregivers. There is no review which summarizes all types of burden reported in informal caregivers of ICU survivors. It is important that the burdens these informal caregivers suffer are systematically assessed so the caregivers can receive the professional care they need. We aimed to provide a complete overview of the types of burdens reported in informal caregivers of adult ICU survivors, to make recommendations on which burdens should be assessed in this population, and which tools should be used to assess them. Method We performed a systematic search in PubMed and CINAHL from database inception until June 2014. All articles reporting on burdens in informal caregivers of adult ICU survivors were included. Two independent reviewers used a standardized form to extract characteristics of informal caregivers, types of burdens and instruments used to assess these burdens. The quality of the included studies was assessed using the Newcastle-Ottawa and the PEDro scales. Results The search yielded 2704 articles, of which we included 28 in our review. The most commonly reported outcomes were psychosocial burden. Six months after ICU discharge, the prevalence of anxiety was between 15 % and 24 %, depression between 4.7 % and 36.4 % and post-traumatic stress disorder (PTSD) between 35 % and 57.1 %. Loss of employment, financial burden, lifestyle interference and low health-related quality of life (HRQoL) were also frequently reported. The most commonly used tools were the Hospital Anxiety and Depression Scale (HADS), Centre for Epidemiological Studies-Depression questionnaire, and Impact of Event Scale (IES). The quality of observational studies was low and of randomized studies moderate to fair. Conclusions Informal caregivers of ICU survivors suffer a substantial variety of burdens. Although the quality of the included studies was poor, there is evidence that burden in the psychosocial field is most prevalent. We suggest screening informal caregivers of ICU survivors for anxiety, depression, PTSD, and HRQoL using respectively the HADS, IES and Short Form 36 and recommend a follow-up period of at least 6 months. Electronic supplementary material The online version of this article (doi:10.1186/s13054-016-1185-9) contains supplementary material, which is available to authorized users.


Background
Since 1992, the in-hospital mortality of intensive care unit (ICU) patients declined from 32 % [1,2] to 15-20 % [3, 4]. ICU survivors frequently suffer from psychological distress, reduced social well-being and long-term physical limitations which may result in a reduced quality of life [5]. This combination of complaints has been defined as post-intensive care syndrome (PICS).
PICS and other problems faced after ICU discharge do not only affect the patient, but also reduce the physical, mental, social, and financial position of patients' informal caregivers, often family members. The combination of psychological problems affecting informal caregivers is known as PICS-family (PICS-F) [6,7], though there is disagreement on what the term 'caregiver burden' entails and how it should be utilized [8].
Systematic reviews have been published on the burden on informal caregivers of ICU patients, but all have different definitions of caregiver burden. Some reviews only include quantitative literature [9], some only focus on the needs and satisfaction of informal caregivers [10,11] and others focus on specific burdens, such as PICS-F [7], post-traumatic stress disorder (PTSD) [12] or psychosocial burdens [13,14]. There is no review which summarizes all reported burdens informal caregivers of ICU survivors can suffer after discharge, and no clear overview of tools to asses these burdens. It is important that the burdens on these caregivers, in addition to PICS-F symptoms, are systematically assessed so the informal caregivers can receive professional care if necessary.
We performed a literature review to: (1) assess which burdens on informal caregivers of adult ICU survivors have been documented; (2) assess which assessment tools are used; and (3) make recommendations on which burden should be assessed and which tools could be used.

Materials and methods
We searched for articles describing burden on informal caregivers of adult ICU survivors, using PubMed and CINAHL from database inception to June 2014. The search strategy is presented in Table 1. Only English and Dutch articles were included.
Two authors (IvB and FBR) independently assessed the titles and abstracts of 50 randomly selected articles to ensure that the inclusion criteria were not ambiguous. For 47 (94 %) of these articles the inclusion criteria were applied identically. After discussing the differences, consensus was reached. We considered the consistency between the two authors sufficient and made no alterations to the inclusion criteria. We included original studies if: the subject of the study was an informal caregiver of an adult ICU patient; the ICU patient was discharged from hospital alive; at least one of the measurements of the burden took place after hospital discharge; and the burden on the informal caregiver was a main outcome of the study. We excluded studies on deceased ICU patients, studies on the needs or satisfaction of the informal caregiver, presence during cardiopulmonary resuscitation, and involvement in end-of-life decisions, because we hypothesized that informal caregivers of these groups would suffer different burdens.
One author (IvB) evaluated the titles and abstracts of all articles. The abstracts were either included, excluded or marked as doubtful. Another author (FBR) read the title and abstract of articles marked as doubtful and both authors discussed these articles to reach consensus on inclusion. We supplemented our searches by scanning the reference lists of previously included articles. The full text of all eligible articles was read by two authors (IvB and one of FBR, NdK, MvdS, or DAD). Both authors extracted data on the study type, characteristics of the informal caregivers, hospital and setting, type of burden and instruments used to assess the burden. If information could not be extracted from the article or online appendices, we e-mailed the corresponding author for additional information. We assessed the quality of the quantitative articles, using the Newcastle-Ottawa scale (NOS) [15] for observational studies and the PEDro scale [16] for randomized trials.

Results
We retrieved 2704 articles using the search strategy described in Table 1. After removing duplicates, we assessed the title and abstract of 2311 articles and excluded 2264 articles based on title and abstract. Figure 1 summarizes the inclusion process and provides the reasons for exclusion. We assessed the full text of 47 articles and excluded another 21 articles. We hand searched the references of the 26 included articles and included two additional studies. Nine authors were contacted to complete the data for 12 articles and six authors responded.
Most of the informal caregivers were female (47-100 %) and most of them were the partner/spouse of the ICU survivor (24-100 %).
We present the results of the quality assessment of the included articles in Table 2. The NOS scores for the 24 observational studies ranged from two to three on a scale of zero to nine, indicating low quality. The PEDro scale score for the three randomized controlled trials ranged from four to seven on a scale of zero to ten, indicating moderate to fair quality.

Burden
We found a large diversity in types of burden reported. Table 3 shows a summary of the main findings. A complete overview of all types of burden is presented in Table S2 in Additional file 2.
Thirteen studies described informal caregivers' employment status [18-20, 22-25, 27, 29, 31, 32, 43, 44] and at study enrolment between 25.4 % and 72.3 % were in paid employment. Four studies reported a reduction in employment around 2 months after enrolment [24,25,27,29] and two reported that almost 50 % of caregivers, who had been employed at enrolment, reduced their work hours, quit their job or were fired in order to provide informal care [24,29].
Of the informal caregivers, who were employed prior to the ICU admission, 84.6 % had returned to their previous work 12 months after enrolment [31]. Their mean sick leave was 11 days (range 4-42) for full-time employees and 9 days (range 0-44) for part-time employees during the patient's ICU stay and 17 days (range 0-124) for full-time employees and 21 days (range 0-106) for part-time employees during the 12 months after ICU discharge [31]. Thirty-eight percent of the informal caregivers reported that it was somewhat difficult to pay for basic needs such as food, housing, medical care and heating. Some of them even moved to a less expensive home, delayed educational plans or medical care for themselves or another family member, or filed for bankruptcy due to the financial burdens [29].  Of the seven studies which described health-related quality of life (HRQoL), four used the Short Form 36 (SF-36) [18,32,33,37], one used the Short Form 8 (SF-8) [23] and two used a single-measure item [22,24] (Table  S6 in Additional file 6). Two found no change in physical health scores [33,37], one reported that 36 % of informal caregivers experienced negative changes in their physical health [22], and one reported no statistically significant differences in changes in physical health between informal caregivers and controls over time [23]. Three studies reported major decreases in the mental health of informal caregivers [32,33,37], one reported that informal caregivers scored lower on all domains of the SF-36 than an age-and gender-matched population [18], and one reported a slight decrease in general health [32].
Eight studies assessed the lifestyle interference of informal caregivers (Table S7 in Additional file 7). Two used the Activity Restriction Scale [27,28], two the Changes in Role Function scale [19,25], one the Caregiving Impact Scale [18], one the "objective indicator" portion of the "objective and subjective burden" tool [22], one the Family Impact Survey [29] and one qualitative methods [42]. Lifestyle interference was high [19,27,28], the percentage of informal caregivers who had quit other activities in order to care for the ICU survivor was 84.5 % 1 month and 45.8 % 12 months after ICU admission [29]. One month after ICU discharge, 75 % had moderate or great restrictions in visiting friends and 48 % in practicing hobbies and recreation [19]. They provided about 5 hours of care a day [22,25,27,43] between hospital discharge [22] and 12 months after initiation of mechanical ventilation [27].

Qualitative research
Five studies had qualitative elements. One relied entirely on semi-structured interviews [21] and four had some qualitative components [31,38,42,44]. They mainly reported psychosocial burdens, such as sleep disorders, nightmares, sadness, distress, anxiety, exhaustion, crying for no apparent reason and keeping a distance from family and friends. Parents described it as 'emotionally draining' to explain the situation to the children [21] or were scared of leaving children alone with the ICU survivor at home. Children's involvement made it more complicated to balance the logistics of home life and work [21,31]. An ICU admission can also impact the relationship between the ICU survivor and the informal caregiver. Informal caregivers and ICU survivors can feel more irritated with each other, experience less freedom than before [42], experience a sense of increased distance in their relationship [21] or even attribute the end of their relationship to the ICU admission [31]. However, one couple stated that they showed each other more tenderness and respect and another reported that their life was more positive following the ICU admission [42].

Discussion
We performed a literature review to assess the burdens experienced by informal caregivers of adult ICU survivors have been documented, how they are assessed and to make recommendations on which burdens should be assessed. We have shown that informal caregivers of ICU survivors have extensive burdens following the patient's ICU admission. This is reflected in psychosocial status, quality of life, lifestyle, employment and financial status. The most frequently used assessment tools were the HADS, the CES-D, the IES and the SF-36.
Psychosocial burdens are most commonly reported and, in this review, we described these in depth. Generally, the prevalence was highest during and shortly after the ICU admission, decreased over time, but remained higher compared to control groups. In contrast, the prevalence of PTSD increased over time. Although different measurement tools were used, the prevalence of depression among informal caregivers of ICU survivors was higher than among informal caregivers of patients with colorectal cancer [45] and following coronary bypass surgery [46], stroke, hip fracture, congestive heart failure and myocardial infarction [46]. We found that, 3 months after ICU discharge, between a quarter and two-thirds of informal caregivers reported anxiety. This is similar to the prevalence reported in a systematic review on anxiety in informal caregivers of people with dementia [47]. Burdens such as insomnia, concentration problems, fear of death and spiritual problems were only described by few authors in low-quality, observational studies. However, these burdens can influence informal caregivers substantially. Further research on the scope of these problems and the appropriated assessment tools is necessary.
A range of assessment tools can be used to quantify the burdens on informal caregivers. However, these tools use different cut-off points to quantify the burden. For example, the HADS uses two different cut-off points. Scores of eight to ten on the anxiety or depression subscale potentially indicate pathology and scores of 11 or more are considered more definite [48]. However, these tools can be used as screening instruments, but are not valid methods for obtaining a clinical diagnosis and cannot predict which informal caregivers will need professional treatment to recover.
Correct use of the questionnaires is crucial, but not always found. For example in the article by McAdam [26] the IER-R is used for informal caregivers during ICU admission of their relative and refers to the outcome as PTSD. However, according to the definition of PTSD, PTSD cannot be evaluated during the event. Symptoms have to be present for at least 1 month after the event of interest in order to be diagnosed as PTSD [49].
Possible benefits of post-ICU clinics for ICU survivors are mentioned before [50]. However, we did not find any recommendations on screening informal caregivers in post-ICU care, though there are recommendations on inviting the informal caregiver to the patient's post-ICU care [51]. Considering the high prevalence of a wide range of burdens in informal caregivers, we highly recommend assessing the informal caregiver as part of the post-ICU care so they can be referred to the appropriate healthcare provider(s) if necessary.
There is a large resemblance between a recently published systematic review about the psychosocial outcomes informal caregivers of ICU patients can suffer [13] and our study, as 11 articles were included in both studies. However, a strength of our study is that we did not restrict our literature search to psychosocial outcomes and could include 17 additional articles [17,26,[29][30][31][32][33][34][36][37][38][39][40][41][42][43][44]. Consequently, we also report on other burdens such as anxiety, loss of employment, financial problems and healthcare consumption. Recognition of these additional types of burden is important for referral to the appropriated healthcare provider.
Another strength of our review is that we included both quantitative and qualitative studies describing burdens informal caregivers can suffer. This means that we could identify additional burdens such as sleeping disorders and negative impacts on social life and relationships [21,31,38,42,44].
Our study also has some limitations. Two pairs of articles describe the same data from samples of 57 [40,41] and 284 [32,37] informal caregivers. Both pairs of articles report on results obtained using the same instruments at the same time points. Since we did not perform a meta-analysis, we believe that the influence of these duplicate data is limited. In addition, the methodological quality of the 24 observational studies was low and the three randomized studies moderate to fair. Although all of the studies report similar results, more high-quality studies are needed to obtain accurate assessments of the prevalence and severity of burdens informal caregivers suffer. extraction for a selection of the eligible articles and helped to draft the manuscript. NdK filled out the data extraction for a selection of the eligible articles and participated in the design and coordination of the review. DAD filled out the data extraction for a selection of the eligible articles and helped in analysing the results. MvdS filled out the data extraction for a selection of the eligible articles and participated in the design and coordination of the review. All authors read and approved the final manuscript.