Background
Palliative care is recognised to have an important role to play in addressing the support needs of bereaved families [
1‐
5]. Although grief is a natural process, in which most people learn to adjust, around nine percent of adults experiencing a loss develop complicated grief reactions [
6]. These reactions have been described as painful and persistent responses associated with mental and physical health problems [
6], and for bereaved caregivers, estimations of the proportion of the population who experience complicated grief range from between 10 and 20 % [
1]. However, the relationship between caregiving and bereavement remains under-researched [
7,
8], and the process of adaptation to loss often decontextualized, without consideration of the experience of caregiving or of the death [
8].
A number of quantitative studies have demonstrated relationships between caregiving experiences and long term health and quality of life outcomes for bereaved caregivers [
9‐
13], while the available qualitative evidence also illustrates some of the ways in which caregiving impacts upon bereavement. Although caregivers in some studies described positive outcomes such as feelings of privilege, accomplishment, expressions of love and improved family relationships [
8,
14‐
19], these and other study findings have also pointed to caregiver burden and the physical and emotional exhaustion experienced by carers at the time of death [
7,
14,
20,
21]. Former caregivers have also reported feelings of failure, guilt and regret in relation to unfulfilled place of care preferences and decision-making responsibilities and outcomes [
15,
19,
20,
22,
23]. These accompany haunting images of physical and emotional suffering associated with the illness [
8,
12,
24,
25], the physical and cognitive decline of the patient [
8,
15,
25] and the trauma of the death itself [
17,
25]. Other factors that have been found to impact upon bereavement and complicate the grieving process for family caregivers include missing the death, lack of preparedness for death [
12,
20,
25], inadequate terminal support [
20,
24,
25] and insufficient knowledge of patient history amongst treating healthcare professionals [
20].
This paper reports on a supplementary qualitative analysis of ‘free text’ (open ended) survey data relating to bereavement, which was collected from a recent research priority setting exercise for palliative and end of life care (EoLC) research [
26]. The results reported in this paper provide further insight into how carer specific experiences of EoLC and the death of their loved ones impact upon their bereavement experiences. Support needs and clinical implications are also identified.
Methods
In 2014 the Palliative and end of life care Priority Setting Partnership (PeolcPSP) carried out a public survey of 1403 people with the aim of identifying unanswered questions in palliative and EoLC (see Table
1 for breakdown of respondents). Using an electronic questionnaire made available on Survey Monkey and a paper version made available in hospices and via the Marie Curie Nursing Service, respondents were invited to ask questions about palliative and EoLC, as displayed in Table
2. A purposeful approach to sampling was used; the survey was disseminated to stakeholder networks via email, newsletters, social media, web posts, presentations and stands at conferences, and blogs, as described elsewhere [
26]. The process followed a method developed by the James Lind Alliance for setting research priorities, which has typically been used to identify research questions concerned with evaluating clinical interventions [
27]. Following this approach and completion of the exercise a ‘top ten’ list of research priorities for Palliative and EoLC in the UK was produced, and is available in an on-line report [
26].
Table 1
Numbers of different groups responding to the PeolcPSP survey; multiple answers were possible
I am in the last few years of my life (Patient) | 59 | 6 |
I am a carer or family member or partner or friend of someone in the last few years of their life (Current carer) | 176 | 20 |
I am a bereaved carer or family member or friend (Bereaved Carer) | 494 | 76 |
I am a professional working with people in the last few years of life (Professional) | 680 | 68 |
I am a volunteer working with people in the last few years of life (Volunteer) | 43 | 3 |
I am a member of the public who has an interest in the subject (Member of Public) | 181 | 11 |
Other | 142 | 4 |
Table 2
Survey questions on palliative and EoLC
Q. What questions do you have about care, support and treatment of people who are in the last few years of their lives that could help them to live as well as possible? This could also include question(s) about care and support for current carers or families. |
Q. What questions do you have about care, support and treatment of people for those rapidly approaching the end of their lives? This could also include question(s) about care and support for current or bereaved carers or families looking after someone at the end of life. |
However, it became clear whilst analysing the survey responses that not all responses could be translated into questions that might be answered by intervention based studies. Many people described their experiences and raised more general questions, for example about the purpose of palliative care. The PeolcPSP Steering Group felt strongly that these responses (which were considered ‘out of scope’ of the original survey analysis) should not be lost and that the voices of everyone who took part in the survey be heard. In particular, it was felt that the detailed, experiential nature of many of the ‘free text’ responses to the survey questions could provide valuable insight into many topics of interest in current palliative and end of life care research, if analysed and reported using a thematic approach.
Once the prioritisation exercise was completed all survey responses were subjected to a supplementary thematic analysis. The initial coding framework was developed using an inductive approach, after two researchers reviewed 200 printed copies of the survey responses. The coding framework was applied to 50 survey responses to confirm that it reflected the data. Researchers entered the coding framework into NVivo 10 software and systematically coded the survey responses into broad thematic areas using established techniques of coding and comparison [
28]. More descriptive codes were then created and applied to the ‘bereavement’ data set by the author (EH) following the same techniques, and were checked by another member of the research team (AN).
Discussion
This paper adds new insights to the limited evidence available on the specific effects that caring for a loved one at the end of life can have on bereavement experiences. These are based on the first hand experiences of family members, the second hand observations of health and social care professionals, or in many cases a combined perspective brought by bereaved professionals, which were articulated in ‘free text’ written responses to a self-completed survey. The data reported here demonstrates two core bereavement issues for family caregivers; the consequences of traumatic deathbed experiences on caregiver grief and feelings of guilt; and the ‘void’ effect caused by the withdrawal of professional support immediately after death, compounding feelings of loss for some recently bereaved family members. Support needs relating to these experiences are also identified with practical suggestions made for how such needs might be addressed.
One of the strongest themes to emerge in this data set concerned the impact of traumatic death experiences on family members’ abilities to grieve and adjust in the months and years that followed. As in Sanderson et al’s study reporting on death experiences and bereavement, the language of trauma was evident as respondents recounted their death bed experiences and appeared haunted by memories of the death [
17]. It seems that some of our respondents also experienced an added critical dimension to their trauma and adjustment difficulties; living with guilt and regret. Whereas most participants in Sanderson et al’ s study engaged in positive self-reappraisal and were able to transform their trauma into more ‘bearable’ stories [
17], these respondents not only perceived physical suffering in their loved ones, but also experienced a sense of powerlessness, exclusion and lack of understanding over what was happening. As a result, many also experienced feelings of lasting guilt and self-doubt over whether the right decisions were made and whether they could have done more. Although positive impacts of the LCP for families have been reported in other European countries [
31,
32], these negative feelings were directly connected by some of our respondents to their first hand experiences of the implementation of the Pathway. Further, it seems that their doubts and sense of regret may have been compounded by media reactions, and the recent withdrawal of the LCP in the UK.
These kinds of guilt experiences are also evident in some of the literature reporting on the bereavement impacts of caregiving prior to the death, such as failure to provide EoLC at home, concerns and regrets over healthcare decisions that were made [
15,
20,
22,
23], and increased, long term psychological morbidity amongst widowers who perceived un-relieved symptoms in their loved ones in the last three months of life [
9,
10]. Moreover, the ‘good death’ experiences reported in this and other papers [
25,
33,
34] and more positive stories of satisfaction and accomplishment reported by participants in other studies [
8,
14‐
19], underline the importance of knowing that everything went as well as it could have done for the process of adjustment post-death.
Experiences of poor communication, resulting in limited understandings of the dying process and medical interventions amongst families, was a common factor in many of the troubled death and bereavement experiences reported here. Following this, several communication and support needs relating to death and dying were articulated. These included better information on the signs of death in order to facilitate awareness of and preparedness for death [
12,
20,
25,
34], along with real time updates on what is happening during the dying process, and what relatives could be doing to help their loved ones at this time. Caregivers and health professionals also described the need for more effective communication and emotional support around EoLC planning and decisions, with detailed discussion at points where the focus of care shifts to prioritising EoLC goals, as evidenced elsewhere [
30,
31]. Post death consultations with health professionals who were directly involved at the end of life were also recommended to alleviate some of the confusion, doubts and concerns of bereaved family members, the importance of which has been identified in other studies [
35,
36].
Another grief experience specific to family carers was the reported “void” created by the withdrawal of services after death. Models of palliative care provision include extended support for the carers and families of patients during the period of illness, as well as intensive multi-professional support for the patient [
3‐
5]. The ‘double loss’ experiences of carers has been reported elsewhere in terms of the ‘vacuum’ effect created by the loss of the caring role after death [
16]. This ‘loss of relationships’ thus indicates a further loss experience for some carers, along with that of their caring role and the major loss of their loved one. It follows that many respondents would have liked continued contact between families and the healthcare organisations and professionals involved in the EoLC of their loved ones, as highlighted elsewhere [
35‐
38]. A lack of information on, and access to, appropriate support was also identified in this and other studies [
7]. Current recommendations for bereavement support in UK palliative care include targeted intervention based on formal risk assessment, coupled with universal information provision on grief responses and available support [
2,
4]. These findings suggest unmet need in relation to both information provision and follow up support. Although most hospices in the UK offer bereavement support and many make contact with bereaved families at around six weeks post-death, these services have been noted as ‘idiosyncratic’ [
38], while evidence guiding the timing, frequency, duration and nature of follow up support is lacking [
36,
38].
Clinical implications
Several implications for practice can be identified. First is the need for improved communication between health care professionals and families, in the period leading up to, and immediately following the patient’s death. As recommended elsewhere, family caregivers should be helped to recognise the signs of death in order to facilitate their preparedness for death [
1,
12,
20,
25,
34] and be made aware when death appears imminent (1). Further, these findings suggest that preparedness for death should incorporate broader aspects of caring for the dying patient, such as physiological changes (e.g. airway secretions, the need or not for parenteral fluids), how to physically care for the dying body and what to do after death. In order to address caregiver experiences of marginalisation and lack of understanding relating to decisions made at the end of life, there also needs to be more effective engagement and discussion with families on end of life care planning and decisions, in particular when the focus of care shifts to prioritising end of life care goals [
1,
29,
32]. As recommended in recent guidelines developed in Australia, it seems that many relatives would benefit from immediate post-death contact from a member of the care team, not only to offer condolences but also answer those potentially ‘nagging’ questions and concerns relating to the death [
1].
These findings also indicate a need for better information provision on bereavement support prior to death and more effective follow up approaches, which in recent guidelines have been recommended to take place 3–6 weeks post death [
1]. Such contact should inform, assess and connect family members with appropriate, available support (if required) [
1,
35,
36], whilst also providing continuity and preventing carers from feeling “cut off” and experiencing the sense of multiple loss described above. As noted elsewhere, well conducted follow up visits also provide an opportunity to acknowledge the family member’s efforts during caregiving, thus also addressing feelings of guilt and supporting caregivers to find positive meaning in their experiences, as discussed above [
36].
Limitations and implications for research
The strength of this data set is the access which it gives to a wide range of perspectives on palliative care in the UK, and the breadth of detailed responses which were submitted on this and other topics of relevance to palliative care provision and research. These types of ‘free text’ survey responses can prove a rich, insightful source of data when analysed appropriately [
39,
40], but there are also limitations. Given the survey aim and design it is likely that the responses represent more ‘extreme’ cases, particularly in relation to bad death experiences which may have left the respondent feeling compelled to speak out, and which contrast with some of the more mixed, even positive experiences reported in other qualitative studies [
8,
14‐
19,
33,
34]. It is also more likely that respondents would have commented on service related problems or experiences than other aspects of their caregiving/ grief experiences which might be perceived as more inevitable (e.g., the decline of the patient). These more extreme and problematic experiences are nonetheless essential to capture, particularly in the context of complicated bereavement and grief issues. Because these responses were ‘fixed’ upon submission, it has also not been possible to unpack these experiences at an individual level or to pick up on points of interest as in interview based approaches, which are recommended for further in depth exploration of these issues. Future research which explores the characteristics and effectiveness of different communication and engagement approaches with families in the pre and post death periods is also recommended.
Acknowledgements
Our thanks go to all of our respondents, to everyone who helped disseminate the survey, to Dr Sabine Best and the PeolcPSP steering group for their support for this part of the project, and to Jordan Van Godwin, Kristen Thomas, Dr Stephanie Sivell and Dr Jessica Baillie for their work on the project.