Keeping all options open: Parents’ approaches to advance care planning

Abstract Background Early engagement in advance care planning (ACP) is seen as fundamental for ensuring the highest standard of care for children and young people with a life‐limiting condition (LLC). However, most families have little knowledge or experience of ACP. Objective To investigate how parents of children and young people with LLCs approach and experience ACP. Methods Open‐ended, semi‐structured interviews were conducted with parents of 18 children; nine children who were currently receiving palliative care services, and nine children who had received palliative care and died. Verbatim transcripts of audiotaped interviews were analysed following principles of grounded theory while acknowledging the use of deductive strategies, taking account of both the child's condition, and the timing and nature of decisions made. Results Parents reported having discussions and making decisions about the place of care, place of death and the limitation of treatment. Most decisions were made relatively late in the illness and by parents who wished to keep their options open. Parents reported different levels of involvement in a range of decisions; many wished to be involved in decision making but did not always feel able to do so. Discussion This study highlights that parents’ approaches to decision making vary by the type of decision required. Their views may change over time, and it is important to allow them to keep their options open. We recommend that clinicians have regular discussions over the course of the illness in an effort to understand parents’ approaches to particular decisions rather than to drive to closure prematurely.

Most studies on ACP for children have taken place in the United States. Some have shown that parents who have participated in ACP may perceive these discussions as helpful in several ways. 7 Parents report that the process can be valuable in ensuring the best care and quality of life; it can allow adequate time and information to make decisions, communication about desired care, and offer peace of mind. 8 In some instances, ACP may help parents to acknowledge where they would like their child to die and achieve this preference, so that the death may occur more peacefully. 9 However, other studies have shown that most families have little knowledge or experience of ACP discussions for their child. [10][11][12][13][14][15][16][17] For example, parents of sons with Duchenne muscular dystrophy reported that they were not familiar with the concept of ACP and they had not had any discussions of plans for future care within the family or with HCPs. 11 In this article, we report the findings of a pilot study to explore parents' experiences in the United Kingdom (UK) of ACP discussions for children for whom cure is not likely. We aimed to increase our understanding of how parents approach and experience planning for their child's future care. We included: (i) parents of children with life-limiting and life-threatening illnesses who were currently receiving specialist palliative care and (ii) parents of children who had recently received such care and had died.

| Sample
Our sample was drawn from a large metropolitan specialist palliative care service in the UK. The size and scope of the caseload provides a population aged 0-19 years with a diverse range of life-limiting conditions, ethnic and socio-economic backgrounds.
The invitation, recruitment and sampling methods have been described in detail elsewhere. 18 Two groups were invited to participate: (i) parents whose child was currently receiving services from the palliative care team (Group A) and (ii) bereaved parents whose child had received care from the palliative care team and had died 6-10 months previously (Group B). Parents were excluded if they were currently or had recently been participating in other psychosocial research, were unable to communicate in English or unable to give informed consent.
Eligible parents were invited to take part by the specialist palliative care clinician under whose care they were registered.

| Interviews
Interviews were conducted by two experienced researchers (EB, JC) at a time and place of the parent's choice. Interviews were audiorecorded, and field notes were taken to provide context. All parents were invited for a second interview, 12 weeks later, to explore further any issues raised but not fully addressed in the first interview, informed by an iterative process following the reading of transcripts.
We used an open-ended semi-structured topic guide, developed from the literature 19 and designed to allow participants to provide a narrative of their own experiences of ACP. This included: (i) recounting of the child's illness, (ii) their involvement in decisions made, (iii) how they experienced the decision making, (iv) when decisions were made; and (v) how decisions were documented. A copy of the interview guides is available from the authors on request. At the end of each interview, every parent was given a leaflet containing details of referral services should they need further support; the day after the interview, the researcher contacted each person via the phone or email to see how they were.

| Data analysis
All interviews were transcribed verbatim by an independent transcription service and checked for accuracy by the interviewers (EB, JC).
The follow-up interview was conducted after preliminary analysis of the first interview. Following each interview, the transcripts and field notes were read and reread independently by the two interviewers (EB, JC). They added non-verbal behaviour recorded in their fieldnotes to the transcripts. Meetings were held post-interviews to facilitate reflection and reflexivity in relation to the interview process. After first interviews, each interviewer prepared, and the other members of the research team (MBL, PK) reviewed, the timelines of the parents' experiences of ACP including at what point in the trajectory particular decisions were made or discussed (eg diagnosis, end of life, crisis) and the child's condition (eg stable, unstable). They also prepared and the research team reviewed summaries of the key issues raised by parents.
These findings then informed the content of second interviews.
We sought to develop accounts of parents' experiences of ACP and to uncover individual approaches. To do this, we used principles of  21 Definitions were developed for each code, and subsequently, the whole dataset was coded using the codebook. All transcripts were then consistently coded by members of the research team (EB, JR) for (i) type of decision made, (ii) periods in the illness and child's condition when decisions were made, (iii) who was involved in the decision, (iv) factors parents identified as contributing to the decisions made, (v) parents' advice for other parents, (iv) reflections on participating in this or forthcoming studies. Our process is illustrated in Figure 1.
Apropos NVivo queries were conducted to explore relationships among the aforementioned codes/nodes in particular where the child was in the illness trajectory, the child's condition and who was involved (See Table 2).

| Participants
Between December 2011 and December 2012, parents of 519 living children (Group A) and 73 deceased children (Group B) met the inclusion criteria. Clinicians introduced the study to the parents of 28 (5%) children in Group A and 21 (29%) children in Group B and invited them to take part. These unexpectedly low invitation rates were explored in a separate study which revealed multiple barriers to invitation by clinicians. Barriers included infrequent contact with families and clinician concerns about families' well-being, and anticipated reaction to invitation 18 and were confirmed in a study by Siden et al. 22 Parents of nine children in each of Group A and Group B (total sample 18) consented to take part in an interview. Table 1 shows the characteristics of participants. Interviews were conducted between February 2012 and May 2013 and lasted a median of 80 minutes (range 19-168 minutes). All parents were offered a follow-up interview but for various reasons (eg death of the child or deterioration of their condition) not all parents could undertake a second interview. In six cases, parents completed one interview, in 11 cases parents completed two interviews and in one case parents completed three interviews at the family's request.  Table 2).

| Findings from interviews
In 15 cases, parents recalled actively deciding where their child would be cared for, while in the other three cases parents either did not mention actively deciding or mentioned being happy with the place in which their child was cared for and/or contemplation of changes was not necessary. In 13 cases, parents reported discussions about using the hospice for respite and/or end-of-life care; however, most cases (n=13, eight Group A and five Group B) parents expressed a preference to care for their child at home as much as possible.
Discussions about preferred place of death were reported by parents of all deceased children in Group B and parents of four children in Group A (including parents of one child who had died before the follow-up interview took place). Parents preferred their child to die either at home or in a hospice (n=12 cases, four Group A, eight Group B). The other five cases were Group A, and the place of death was not mentioned in the interview. Decisions were not necessarily straightforward. As one mother explained: "We wanted, well when she died she was at [the tertiary hospital], but we wanted to get her into the hospice, but we could…they, she wasn't stable enough to transfer her. That's where she wanted to be and where we wanted her to be, but we couldn't get her there." (ii) Periods in the illness and child's condition when decisions were made Parents mentioned different periods in the illness in which discussions were held and decisions made ( Table 2). As shown in Table 2, decisions (n=10 decisions, 12% of all 83 decisions) were recalled as being made at the time of diagnosis and these were recalled in only eight cases.
Most decisions (n=32, 39%), including those reported in 15 cases, were made at times when the child was unstable, and a change in intervention was required to address problems (eg inserting a feeding tube, or deciding about high dose chemotherapy). In addition, decisions (n=11, 13%) were reported to have been made in six cases when there was a crisis requiring urgent medical attention because of increasing symptom burden with worsening family distress, or as reported by parents in five cases at the end of life when death was thought to be imminent (n=14, 17%). Other decisions were not limited to a particular time period and seemed to be made repeatedly throughout the child's illness. For example, parents in two cases mentioned that written plans were revisited 6 monthly or annually.
Of note, many parents' narratives indicated a desire to keep options open, by stating that they would decide at the time or by agreeing to limit treatment with the knowledge that they could change their mind later. Parents of only three children agreed to limit all active treatment; however, all with the proviso that they could change their minds, as illustrated in these remarks by one mother: "There's been many milestones. We sat in this very room some time ago where… it was deciding whether T A B L E 1 Characteristics of parents who participated in interviews and their children Characteristics Group A Parents whose child was currently receiving palliative care (9 cases) Group B Bereaved parents whose child had received palliative care (

Not specified
Subtotal decisions regarding resuscitation HCP, healthcare professional; NS, not specified; SCP, social care professional; NG tube, nasogastric tube (a tube that provides access to the stomach via the nasal passage); TPN, total parenteral nutrition (intravenous nutrition); PEG, percutaneous endoscopic gastrostomy (an endoscopic medical procedure in which a feeding tube is placed through the abdominal wall and into the stomach).
a Number of discussions/decisions in which these stakeholders were involved. Bag-valve-mask ventilation is a basic airway management technique that allows for oxygenation and ventilation of patients while avoiding more aggressive endotracheal intubation.
T A B L E 2 (continued) document with your daughter still not showing the symptoms they will show when they're going to die."

-Father of child 10 (Group B) (iii) Involvement in decision making
Parents reported a range of people involved in discussions and decisions about children's care and treatment, including the parents themselves, an ill child, a well sibling, an extended family member and a family friend, and health and social care professionals (see Table 2). However, most decisions (n=68, 82%) reported by parents did not involve anyone else apart from the parent(s) or the HCP(s).
Notably, parents reported that there were some decisions made within the family without consulting a HCP. Parents commented that sometimes HCPs asked parents to make a particular decision. Yet, as one mother explained parents did not always want the HCP to involve them in decision making: "The first major one, yeah. And decisions like when she has a chest infection, whether to treat or not to treat.  Parents in eight cases reported being given strong advice by clinicians to limit treatment and why they accepted that advice despite misgivings. As one mother explained: "That's the difficulty is you're in a position where you know nothing and other people know a lot more than you. That doesn't always mean that they're right, but you have to kind of make a decision based on the information that you're given and therefore I don't knowwas that the right decision or not? I don't know, but the fact is that [the palliative care nurse] thought it was. And therefore we accepted it as being the right thing to do."

-Mother of child 16 (Group B)
Parents in eight cases felt they did not have much choice with regard to feeding options (eg because their child had an NG tube fitted directly after birth).
(v) Helpful ways to support parents when making decisions about the child's care and treatment Parents discussed several ways that they thought would be helpful in supporting parents in decision making. All parents prominently mentioned the interaction between clinicians and parents, including the need for clinicians to understand the bigger picture of the life of the child and the life of the wider family rather than simply focusing on treating a particular symptom. Parents spoke of the importance of clinicians understanding the need for them to take professional control at certain times and provide practical help. Their suggestions also included the need for clinicians to give parents sufficient time to make decisions, allowing them time to adjust to the child's diagnosis or prognosis. In addition, parents mentioned it would be helpful to have more information about treatment options and likely outcomes, as one mother described: Recording and sharing the plan can help to provide practical guidance for HCPs and ensure that care and treatment follow the families' expressed preferences. 23,29 As in other studies, parents reported a range of people having been involved in decision making at some point, including family members and friends, various HCPs and psychosocial professionals. 7,23,30 While parents in this and previous studies indicated a preference to be involved in discussions and decisions, 10 Intensive Care Unit has indicated an association between parents who perceived that they had shared in decisions and lower grief scores. 32 We identified several ways to support parents in making decisions.
Examples include, as also reported by parents in the United States and Australia, the sharing of information by clinicians in a trusting relationship, 34 information about the potential outcomes of treatment options and the consequences of refraining from certain options. 10,17,35 It might help parents to visualize the various treatment options or outcomes, if they are informed simply and shown pictures or videos or offered opportunities to see similar cases first hand. 36 Our parents, and those in a previous study, appreciated clinicians providing a clear recommendation while still allowing parents to be involved in decisions at the level they prefer. 35 In addition, we identified a need to provide parents with more information about the aims and process of ACP, including procedures for updating their preferences, and how written plans might be used.

| Strengths and limitations
A primary strength of our work was the inclusion of perspectives from the parents of children with a range of LLCs, both deceased and alive, and not restricting the study to decisions about predefined options. The inclusion of a follow-up interview for participants allowed researchers, guided by emerging data, to explore and understand the decision making process in more depth. Parents appreciated the chance to speak for a second time, as one mother put it: "it's almost like exploring it all, you know, like finding out different things about how it was. So it's been quite, quite good, like quite reflective." However, our sample was limited to the families of 18 children, and in most cases, only the child's mother participated in the inter- views. An analysis of selection bias due to non-invitation of eligible families has shown that clinicians were more likely to invite families they knew well, with whom they felt they had a "good" relationship. 18 We appreciate that our sample was drawn from the caseload of a specialist paediatric palliative care team for whom ACP is a recognized aim of their practice; in other settings this may not be so.
That said, however, running throughout all of the interviews were Our findings suggest that to understand and take account of the nuanced nature of decision making for parents whose children face a limited future, further work should involve observation and recording of discussions between parents and clinicians as they occur in real time. We also note the need to capture the experiences of families for whom English is not their first language. This group is often excluded from research yet makes up a large proportion of the caseload of the palliative care team in a metropolitan hospital. We have now begun a longitudinal, prospective ethnographic study of children with high-risk brain tumours to understand decision making in this context.

| CONCLUSIONS
This study highlights that the views of parents of a child with a LLC change over time that their approach to decision making is influenced by the type of decision required and that most parents wish to keep their options open for as long as possible. As one mother told us: "Well, I don't know for how long she's going to live, but I have to live with the decisions I make". We recommend that clinicians have regular discussions over the course of the illness in an effort to understand parents' approaches rather than to arrive at a particular decision.