Support from healthcare professionals in empowering family carers to discuss advance care planning: A population-based survey

Background: Family carers have a prominent role in end-of-life care for seriously ill persons. However, most of the advance care planning literature is focused on the role of healthcare professionals. Aims: To investigate (1) what proportion of family carers discussed advance care planning with their relative and associated socio-demographic and clinical characteristics (2) what proportion received support from healthcare professionals for these conversations, (3) what type of support they received and (4) to what extent the type of support received was considered sufficient. Design/participants: Population-based cross-sectional survey in Belgium of bereaved family carers of persons with a serious chronic illness (N = 3000) who died 2–6 months before the sample was drawn, identified through three sickness funds. The survey explored support from healthcare professionals for family carers during the last 3 months of the patient’s life. Results: Response rate was 55%. The proportion of family carers that engaged in an advance care planning conversation with their relative was 46.9%. Of these family carers, 78.1% received support from a healthcare professional, mostly by doing the advance care planning conversation together (53.8%). Of family carers receiving support from a healthcare professional, 57.4% deemed the support sufficient. Conclusion: Many family carers engage in advance care planning conversations with their dying relative. Healthcare professionals often support them by performing the advance care planning conversations together. More insight into how family carers can be supported to conduct these advance care planning conversations, both with and without involvement of healthcare professionals, is necessary.


Introduction
Advance care planning is defined as an early-initiated dynamic communication process between patients, family carers, and healthcare professionals that enables individuals to define goals and preferences for future (end-of-life) care. 1,2 The goal of advance care planning is to have a timely exploration of a person's underlying values, wishes, and preferences for future end-of-life care in order to develop a shared narrative and understanding to inform patient-centered care, which becomes particularly important if a patient loses the ability to make their own decisions about care. 1,2 Involvement of family carers is crucial in this process since they often have a prominent role in care decision-making at the end of life, 3 and it has been shown that involvement in advance care planning improves family carers' confidence when making decisions at the end of life on behalf of their relative. 4 Recent interventions mostly focus on improving advance care planning between the patient and/or family carers and healthcare professionals, 5,6 however qualitative studies have shown that patients and family carers also want to discuss future care and preferences among each other. [7][8][9] Moreover, previous studies found that healthcare professionals experience many barriers in initiating advance care planning conversations, such as lack of time or not knowing when or how to start these conversations. 10,11 Patients also mostly turn to professional carers for physical and disease-related problems, but do not want to burden them with discussing their psychosocial or existential needs. 12 Family carers' time spent caring for their relative is often more substantial than the time spent by healthcare professionals, 13 which means that there are many opportunities to discuss advance care planning within the home setting. Although family carers are mostly willing and welcoming to have these conversations, they often feel unprepared, anxious, or uncomfortable to do so. [14][15][16][17] Healthcare professionals can play an important role, for instance by facilitating advance care planning conversations between patients and family carers in the home setting; performing the advance care planning conversation together with them or by empowering the family carers to increase their self-efficacy for initiating such conversations by themselves. Providing family carers with tools, such as conversations cards, to initiate advance care planning discussions are practical examples of such approaches. 18,19 The aim of this study was to examine the extent to which family carers of people with a serious illness performed advance care planning conversations with their relative at home in the last 3 months of life and how they were supported by healthcare professionals to conduct these conversations.
Specific research questions are: 1. What proportion of family carers discussed advance care planning with their seriously ill relative in the last 3 months of life and what sociodemographic and clinical characteristics are associated with having such discussion? 2. What proportion received support from healthcare professionals for these conversations? 3. What type of support did they receive? 4. To what extent was the type of support received from healthcare professionals considered as sufficient?

Study design and setting
A population-based nationwide cross-sectional survey in Flanders, Belgium was conducted among bereaved family carers. In Belgium, people with long-term extensive care needs can apply for a monthly care budget that helps cover What this paper adds: • • About half of the family carers have an advance care planning conversation with their relative in the final 3 months of life • • Family carers were more likely to engage in advance care planning in case they were 55 of age or younger, had a medical degree (e.g. nurse or doctor), and when specialist palliative care services were involved • • The majority of family carers received advance care planning conversation support from a healthcare professional by performing the advance care planning conversation together Implications for practice, theory or policy: • • Healthcare professionals can play an important role in introducing tools, website or information campaigns of advance care planning • • Healthcare professionals may consider strategies to support and empower family carers to conduct advance care planning conversations outside the clinical context non-medical care. We used registers of the three largest health insurers (accounting for 79% of the population) in Flanders to identify participants. The insurers maintain records people who apply for this care budget and the names of their family carers who are also registered during the application process. The Strengthening of the Reporting of Observational studies in Epidemiology (STROBE) 20 was used to describe all relevant aspects of the study.

Population
We randomly sampled 3000 deaths of people with a serious illness who applied for a care budget-indicative of increased informal care needs-from one of the participating health insurers. For these cases, the person registered in the database as their family carer was selected. We only included family carers who had provided care to someone who recently died between 2 and 6 months before inclusion to allow sufficient time for grieving while limiting recall bias. [21][22][23] Additional inclusion criteria were correct postal address and that both (deceased relative and family carer) were older than 18 years.

Sampling and selection of the participants
Random sampling with proportional stratification was used to obtain a distribution among the different health insurers reflecting the actual distribution of registered family carers in the total population (Christelijke Mutualiteit n = 2748; Socialistische Mutualiteit n = 157; Liberale Mutualiteit n = 95). To answer the research questions accurately, family carers were included for analysis in case (1) they had contact with the deceased person during the last 3 months of life (question 2 in questionnaire); and in case (2) the deceased person had at least one of the listed serious illnesses (question 3 in questionnaire).

Data collection
Questionnaires were sent to the family carers via post between November 2019 and January 2020 by an independent data collector, including an information letter with details of the study, the voluntary nature of the study and the option to complete the questionnaire online. Informed consent was provided as part of the returned the questionnaire. In case of non-response the family carers received two reminders by post (after 2 and 4 weeks of non-response). A non-response questionnaire was sent no longer than 2 weeks after the final reminder.

The questionnaire
As we could not identify an appropriate pre-existing validated instrument(s) to address our research questions we utilized a combination of validated 24 and self-developed items. The questionnaire was tested through two rounds of cognitive interviews with five family carers whom were recruited via convenience sampling. The cognitive testing resulted in two additional questions (questions 2 and 4), changing the sentence structure of certain items and clarifying certain concepts. The final version of the questionnaire can be found in Supplemental Appendix 1.
In this study, we focused on the psychosocial task "advance care planning discussions," which was defined in the questionnaire as follows: "Discussing with my relative about end-of-life care planning (e.g.) the care that my relative wanted or did not want at the end-of-life, advance directives, a living will, wishes about euthanasia or other end-of-life decisions such as palliative sedation." For each caregiving task, the family carer was asked: (1) Did you perform this task at least once during the last 3 months of the patient's life? (yes/no); (2) Which healthcare professionals have supported you in this task? (multiple answers possible including no support was received from healthcare professional); (3) How were you supported in this task by healthcare professionals (multiple answers possible), and (4) To what extent was the support of healthcare professionals sufficient to perform this task (4-point scale).

Statistical analysis
Data were analyzed using SPSS Statistics version 26. Descriptive statistics were used to summarize the data, therefore we adopted frequency tables and contingency tables with the χ 2 -test. To describe which factors were associated with advance care planning discussion, we used multivariable binary logistic regression (advance care planning discussion vs no discussion as dependent variable). The potential predictors included: bereaved family carer characteristics (gender, age in years, educational attainment, medical degree, work status); deceased patient characteristics (serious chronic illness, age in years); and care characteristics (involvement of other family carers, patient lived at home in the final 3 months, and palliative care services involved) were included in a hierarchical stepwise model selection. We started with a univariable binary logistic regression model and selected for the stepwise model only the variables which were (marginally, i.e. p < 0.1) statistically significant. Next, three multivariable binary logistic regression models were constructed: (1) deceased patient characteristics, (2) adding bereaved family carer characteristics, (3) adding care characteristics. Only variables with p < 0.05 were retained for a next step and for the final model. Decisional capacity was left out of the analysis due to multicollinearity with dementia.

Ethical approval
Ethical approval was obtained from the Ethical Review Board of Brussels University hospital of the Vrije Universiteit Brussel (B.U.N. 143201940562).

Results
Of the sampled 3000 people 2889 received the questionnaire (see Figure 1 for the reasons for exclusion). A nonresponse survey showed that another 89 participants of the sample were non-eligible, resulting in 2800 eligible cases. In total, 1539 participants completed the questionnaire, resulting in a response rate of 55%. For the analysis, a total of 1334 respondents (i.e. family carers of someone with at least one serious illness and had contact with them during the final 3 months of life) were included.

Sample characteristics
Most of the family carers were women (68.2%) and were either the child (51.7%) or partner (34.4%) of the deceased person. Family carers were on average 65.5 years. Approximately 13% of family carers had a medical degree in healthcare (e.g. physician, nurse) ( Table 1).
Almost 40% of the deceased persons were older than 85 years. The most common chronic illnesses were cancer (31.8%) and dementia (30.7%). In 40.2% of the cases, the deceased person could still make decisions for him/herself in the final 3 months of life and 36.7% of the family carers indicated that specialist palliative care services were involved in the care ( Table 1).

Profile of family carers who discussed advance care planning
In 46.9% of the cases, family carers had engaged in advance care planning conversations with their relative 3 months before bereavement (not shown in table). Advance care planning conversations were more likely in those family carers who were 55 years or younger (54.7%), had a medical degree (e.g. physician, nurse) (57.8%), were the partner of the deceased patient (52.2%) and were the only family carer (52.5%). As for the patient characteristics, advance care planning discussions were more likely when the deceased persons were aged between 66 and 75 (57.8%); had cancer (44%); had decisional capacity in the final 3 months (64.8%), lived at home in the final 3 months (57.4%) and was supported by specialist palliative care services (59.2%) ( Table 1).

Factors associated with having advance care planning conversations
Multivariable analysis confirmed a higher likelihood of having an advance care planning conversation when family carers were 55 years or younger (OR = 1.48, 1.03-2.13), had a medical degree (OR = 1.76, 1.20-2.58), and when specialist palliative care services were involved (OR = 2.07,    1.55-2.75). Family carers were less likely to have an advance care planning conversation when the deceased relative was diagnosed with dementia (OR = 0.25, 0.17-0.37) ( Table 2).

Support of healthcare professionals and type of support received to conduct advance care planning
Of those family carers who had an advance care planning conversation with their relative 78.1% received support for this from a healthcare professional. Of those that did not receive support (21.9%), 78.3% indicated that they did not need the support of healthcare professionals while 21.7% indicated that they did. Most often, support received was from the GP (66.6%), palliative care nurse or doctor (36.6%), home care nurse (33.2%) or nurse or care worker in hospital or residential care center (32%) ( Table 3). The type of support received from a healthcare professional was in 53.8% doing the advance care planning conversation together with the family carer and the deceased person. It was less often by explaining the family carers in how to do advance care planning conversations (17.9%) and rarely by referring them to other useful sources on how to do advance care planning conversations (9.4%) (not in table). The support received was deemed as sufficient by 57.4% of the family carers (see Figure 2).

Main findings
Our findings show that about half of the bereaved family carers indicated to have engaged in advance care planning conversations with their relative in the final 3 months of life. These conversations were more likely to have taken place when family carers were 55 years of age or younger, had a medical degree or when specialist palliative care services were involved, but less likely when the deceased person was diagnosed with dementia. Seventy-eight percent of the family carers received support from a healthcare professional, usually the GP. The type of support received was most often the healthcare professional doing the conversation together with the family carer and seriously ill person.

Strengths and limitations
This study has several strengths. First, the sampling framework of this study was unique because we used complete registers of health insurances and took a random sample from this. The study therefore suffers less from selection bias compared to most studies that select family carers, for instance, via professional palliative care or other healthcare service use; and statistical generalization to the target population of family carers is improved. Second, the questionnaire was available both on paper and online which seemed beneficial for the response rate (55%), as it was higher compared to other studies with similar populations. [25][26][27] This study has also some limitations. First, despite the effort of construing a complete sample frame there might still be some selection bias because we only could include registered family carers. As a result, we have no insights into the experience of family carers who were not registered by one of the health insurances. Second, the term "advance care planning" is not always wellknown in the general population and therefore, in this survey, we described advance care planning as conversations about end-of-life care and end-of-life decisions or completion of advanced directives or living wills. Recent recommendations highlight the need to approach advance care planning on a broader level, whereas the focus is more on the persons' values, wishes and preferences, and the need of advance care planning occurring over time with multiple conversations. 28,29 To reduce recall bias, which cannot be entirely excluded, we used a very specific description of advance care planning and the time between advance care planning and the survey completion was a maximum of 9 months, which also minimizes the impact of recall bias.

Interpretation of the main findings
Almost 50% of the family carers did have advance care planning conversations with their relative. This result is similar compared to another cross-sectional study about family relationships in advance care planning. 30 We found that family carers who were 55 years of age or younger were more likely to engage in advance care planning conversations with their relative. This finding is in line with other studies showing that in general younger patients  and family carers desire more detailed information regarding the illness, and want more involvement and autonomy in decision-making. 31,32 As attitudes toward discussing end-of-life issues might differ in younger family carers compared to older generations, younger family carers might be more open and willing to initiate advance care planning conversations with their relative themselves. To increase normalization of advance care planning, public health campaigns are developed 33 and more self-guided tools to facilitate advance care planning conversations (e.g. websites, conversation cards) are becoming available. 34 Older adults might find it difficult to find their way to these tools, websites or information campaigns of advance care planning, so healthcare professionals have a key role in introducing these tools to older adults. 35 As a lot of these tools has been developed in the recent years, it is highly recommended that these are tested upon the acceptability, comprehensibility, and utility in future studies. 36 Also, a recent review about advance care planning in the community revealed that the COVID-19 pandemic has worsened some known barriers to advance care planning (e.g. issues of trust given fears of rationing of healthcare resources), while others have improved (e.g. lack of consumer knowledge about advance care planning and perceiving advance care planning as irrelevant due to the increased media discussion of advance care planning). 36 Future research about the impact of the pandemic on the experiences of bereaved family carers now is warranted. This study focuses on the last 3 months of life, but it is highly recommended to initiate advance care planning in the earlier stages of the disease trajectory, 37 and to focus on the process of advance care planning, rather than on its product. 38 Our results showed that when the deceased person was diagnosed with dementia, the likelihood of family carers performing advance care planning conversations with their relative in the final 3 months decreased, possibly because the person with dementia did not have the mental capacity to participate anymore. Previous studies have shown that only a minority of people with dementia get the opportunity to engage in advance care planning themselves and more often the healthcare professional only involves the family carer. 39 It has been recommended to maximize the opportunity for people with dementia to participate in advance care planning as active agents, in order to highlight their capabilities and to have their voice heard. 38 Furthermore, advance care planning should be reviewed regularly and especially in case of change in health condition, and particularly in people with dementia. In general, it is advocated to consider advance care planning as a holistic, flexible and relational process that is initiated in a timely manner. 38 Our findings seem to indicate that the principles of advance care planning have yet to be adopted more outside the specialized palliative care context. The involvement of specialist palliative care services was strongly associated with engaging in advance care planning in the last 3 months of life. This is not surprising as one of the core tasks of these services is to inform patients and family carers about the severity of the illness and to prepare them for the approaching death. 40 Family carers with a medical degree were also significantly more likely to engage in an advance care planning conversation with their relative in the final 3 months of life. They might be more aware of the importance of engaging in advance care planning conversations, are more skilled to perform these conversations themselves and find their way more easily to specific professional support and services. 41 In 78%, family carers were supported by a healthcare professional (mostly by the GP) and most of the support by healthcare professionals was given by performing the advance care planning conversations together with the family carers and their relative. This can be considered as an empowering collaborative approach (versus providing information or explaining how to conduct these conversations which were less frequently reported in this study). However, previous studies have shown that family carers and seriously ill persons also are willing to have these conversations alone, without the involvement of a professional. [7][8][9] Perhaps family carers felt safer to perform these conversations together with a healthcare professional or they might have lacked the necessary tools or skills to perform these conversations themselves. Given that only 57.4% of the family carers perceived the support received as sufficient, more insight is needed into the specific support needs of family carers for conducting these advance care planning conversations with their relative during the disease trajectory, or on how family carers wish to be supported by healthcare professionals in engaging in advance care planning conversations, as openness for communication about end-of-life care and end-of-life decision making might grow during the disease trajectory. 15,35 A possibility for increasing empowering and self-efficacy in family carer is via educational progammes (e.g. FOCUS program). 42,43

Conclusion
This study shows that about half of the family carers had engaged in advance care planning conversations with their relative in the final 3 months of life. Most of the family carers who had done so, received support from a healthcare professional, mostly from a GP and mostly by performing the advance care planning conversation together with the family carer and the seriously ill person. Further insight is needed into how family carers can be supported to conduct these advance care planning conversations, both with and without involvement of healthcare professionals.