Introduction
For parents, making an end-of-life decision for their child is a profound, consequential experience. Considerable research has examined parents’ roles (Brinchmann et al., Reference Brinchmann, Førde and Nortvedt2002; Caeymaex & Speranza, Reference Caeymaex, Speranza and Vasilescu2011; Carnevale et al., Reference Carnevale, Canoui and Hubert2006, Reference Carnevale, Canoui and Cremer2007; Einarsdóttir, Reference Einarsdóttir2009; McHaffie et al., Reference McHaffie, Laing and Parker2001a; Orfali & Gordon, Reference Orfali and Gordon2004; Sullivan et al., Reference Sullivan, Gillam and Monagle2014). It has also examined what is of concern to them in decision-making (Armentrout, Reference Armentrout2007; Caeymaex & Speranza, Reference Caeymaex, Speranza and Vasilescu2011; Einarsdóttir, Reference Einarsdóttir2009; Hinds & Kelly, Reference Hinds and Kelly2010; Moro et al., Reference Moro, Kavanaugh and Savage2011) and how they make decisions (Armentrout, Reference Armentrout2007; Kavanaugh et al., Reference Kavanaugh, Savage and Kilpatrick2005; Sharman et al., Reference Sharman, Meert and Sarnait2005). Additionally, some research is available on the longer term psychological aftermath for parents of deciding to forego life-sustaining medical treatment (LMT) (Armentrout, Reference Armentrout2007; Brosig & Pierucci, Reference Brosig, Pierucci and Kupst2007; Caeymaex & Speranza, Reference Caeymaex, Speranza and Vasilescu2011; Carnevale et al., Reference Carnevale, Canoui and Cremer2007; Hinds et al., Reference Hinds, Oakes and Hicks2009; Orfali & Gordon, Reference Orfali and Gordon2004; Sharman et al., Reference Sharman, Meert and Sarnait2005; Walwork & Ellison, Reference Walwork and Ellison1985; Wocial, Reference Wocial2000).
Previous palliative care research has considered the support needs of families in the period before the child's death (Himelstein, Reference Himelstein2006; Monterosso et al., Reference Monterosso, Kristjanson and Phillips2009; Weidner et al., Reference Weidner, Cameron and Lee2011), with arguments made for research about the lived experience of parents of children with life-limiting conditions (Beecham et al., Reference Beecham, Hudson and Oostendorp2016; Collins et al., Reference Collins, Hennessy-Anderson and Hosking2016). Further, the pediatric oncology literature has identified that the period between making an end-of-life decision and death can be an extended stage with possible attendant psychosocial issues for families (Hinds et al., Reference Hinds, Schum and Baker2005; Weidner et al., Reference Weidner, Cameron and Lee2011). Although there are some data on the time intervals between first discussion of withdrawal of LMT and the child's death in a pediatric intensive care unit (ICU) (Oberender & Tibballs, Reference Oberender and Tibballs2011), these do not provide insight into parents’ experiences during this interval. Typically, such studies have examined what decisions are made and how, but not what happens after a decision is made to forego LMT. Consequently, there are limitations to the knowledge base about end-of-life decision-making (ELD) in relation to how the whole process is experienced by parents over time, and, in particular, the period after the decision is made and before the child dies.
This paper discusses parents’ experiences of the period after an EOL decision has been, up until the death of the child and is based on findings from an exploratory descriptive study examining bereaved parents’ views and experiences of an ELD. Other aspects of the findings from the study, relating to parents’ perceptions of their role in the making the end-of-life decision and what they want from doctors in the process leading up to the decision, have been previously reported (Sullivan et al., Reference Sullivan, Monagle and Gillam2013, Reference Sullivan, Gillam and Monagle2014).
This paper reports the intricacies of putting the EOL decision into practice (“decision enactment”), including differing enactment pathways for withdrawing and withholding LMT, the need to recognize the moment when not to escalate or continue treatment, resultant decisions, and the timing of, and location for, enactment. It also indicates how unexpected events may affect parents’ decision enactment. It concludes by discussing ways in which healthcare professionals can support parents through the challenges of this process.
Patients and method
A qualitative method of semistructured interviews was used. This methodology is ethically appropriate for the potential vulnerabilities of participants and is particularly suited to illuminating a complex and understudied phenomenon (Gillam, Reference Gillam2008; Stroebe et al., Reference Stroebe, Stroebe and Schut2003).
With the approval (30061A) of the Human Research Ethics Committee of The Royal Children's Hospital Melbourne, an invitation to participate in the study was placed in newsletters of three organizations caring for parents whose children had died from various causes. The criteria for parents to participate were: (1) their child had a life-limiting condition; (2) their child was aged between 3 months and 12 years at the time of their death; and (3) the parents had discussed end-of-life decisions and care for their child. The third criterion was deliberately broad, so as not to assume discussing end-of-life decisions had only occurred with medical professionals. In contrast to many previous studies, the sample included not only parents of children who were patients in neonatal or pediatric ICUs but also in other settings, including wards or outpatient clinics. There was no restriction based on the child's diagnosis, other than it be a life-limiting condition. The purpose of this was to allow for an open investigation of the topic. The upper age limit of the child was chosen with the aim of limiting the sample to cases in which the child would not have input into the ELD because child input would have significantly altered the nature of the decision-making process. This constituted a purposive limit on the heterogeneity of sample. The aim was to reduce the variability in the sample to promote meaningful analysis.
Interviews were conducted by the principal researcher (JS). The audio-recorded interviews were transcribed and thematically analyzed using a four-stage analysis method (Liamputtong & Sherry, Reference Liamputtong, Sherry and Liamputtong2010), which allowed interpretations to emerge from the data, rather than being framed by pre-existing concepts or theories. Rigor of analysis was enhanced by use of a thematic analysis good practice guide (Braun & Clark, Reference Braun and Clarke2006). Preliminary coding of the first group of transcripts was conducted by JS and reviewed by LG. Differences in interpretation were discussed and resolved, and an agreed-on coding schema was created. JS then coded all remaining transcripts. Grouping of codes into themes was reviewed by both JS and LG to ensure that claims were supported by the data. The study exhausted the recruitment strategy; no other parents could be recruited. At this point, theoretical saturation was being approached.
Overall, six major themes were identified: (1) the nature of an end-of-life decision; (2) decision-making process; (3) factors in decision-making; (4) parents’ roles; (5) doctors’ roles and (6) the aftermath of an end-of-life decision. Findings reported here relate to the enactment stage, a subtheme of the nature of an end-of-life decision theme. Quotations from transcripts to support the analysis are provided.
Results
Parents
Twenty-five parents (19 mothers and 6 fathers), aged between 30 years and the mid-50s, participated in the study. The parents were all English speaking and were broadly culturally homogeneous, despite an intention to recruit from all cultural backgrounds. Interviews were conducted with 15 individuals and five couples. One parent had two children who had died.
Children
Twenty-one children, aged from three months to 10.5 years, were represented. The average age was 3.5 years. The time since death ranged from four months to 16 years, with an average of 5.75 years. Ten children had incurable conditions (neurological disorder, cardiac disorder, metabolic disease, chromosomal abnormality, neuromuscular disorder, muscular disorder), six had potentially curable conditions (cancer [neuroblastoma, neurofibrosarcoma]), and five had severe disability (spina bifida, brain damage, cerebral palsy).
Contextualizing end-of-life decisions: types and timing
The majority of parents’ decisions related to withholding rather than withdrawing LMT. Most parents had made decisions about more than one form of LMT. For example, when noninvasive ventilation was ineffective, parents had to decide about invasive mechanical ventilation:
…that's when the CPAP [continuous positive airway pressure] machine wasn't working for him and we had the intensive care doctors talk to us. And say now is the time you need to make a decision… It was if the CPAP machine doesn't work, we may need to intubate. (Parent 14)
Treatments withheld were: forms of ventilation (endotracheal intubation, CPAP, bilevel positive airway pressure, bag-valve mask), cardiopulmonary resuscitation, antibiotics, cure-directed therapy (radiotherapy/chemotherapy) or surgery, drug trial, experimental medication, and enzyme treatment. Withholding LMT happened more frequently out of the hospital, at home.
The treatment most typically withdrawn was invasive mechanical ventilation. Noninvasive forms of ventilation, cardiopulmonary resuscitation, and curative treatments were also discontinued. For one child, the parents stopped bag-mask ventilation after a time and then did not reinstigate it. LMT withdrawal mostly occurred in a hospital, and for the majority of children in the ICU. The location of treatment withdrawal was not discussed by doctors with all parents. One parent was told withdrawal had to occur in the ICU. One couple was given the option of withdrawal taking place at home.
The timing of decision-making was affected by the child's medical condition. For nearly all children with an incurable condition, decision-making about end-of-life started close to the time of diagnosis. For children with a potentially curable condition, the majority of parents decided at the time of disease recurrence not to continue cure-directed treatment. This was months to several years after diagnosis. Although diagnosis for most children with disabilities was made in early infancy, end-of-life decisions were made years later, when there was a medical crisis or hospital admission. Times when an end-of-life decision was made are demonstrated in Figure 1.
Fig. 1. Times when an end-of-life decision was made.
Enacting an end-of-life decision: variations in time
Once a decision to withdraw or withhold treatment had been made, the time interval before actively implementing the decision ranged from almost immediately after the decision was expressed to several years later. The time between decision implementation and the child's death also varied, from a matter of minutes, or hours, to several weeks or many months. Withdrawal of LMT was usually close in time to the child's death compared with withholding LMT.
Despite the variations of the period between making and enacting an end-of-life decision, the most important thing for all parents was to give their child the best life possible. Whatever time there was between making the decision to withdraw or withhold, then enacting and finally it having an effect, it gave parents the opportunity to attend to a number of things that were important to them. This included giving their child a good life, a “non-medicalized life”; saying what they needed to say to their child; and preparing themselves for the child's death. It is significant that this period, even if brief, seemed to enable a time for family life and the building of parent-child relationships, experiences, and memories.
The types of variations in the time of enacting an end-of-life decision and the child's death are illustrated in Figure 2 Part A.
Fig. 2. Parental experiences of end-of-life decision enactment.
Parents did not articulate that these periods in themselves were difficult, whether long or short. They did not report regret about or rethinking the decision they had made. However, what distressed parents was when their child suffered: “We sort of got to the point where towards the end as she was getting bigger she really was suffering . . . We probably should have let her go before we did but we weren't ready.” (Parent 3)
Pathways to end-of-life decision enactment
Parents’ accounts of their experiences revealed two main pathways to decision enactment.
Pathway 1: withholding LMT
Parents who had decided to withhold LMT had planned to do this either when their child had become acutely unwell in some form or when a particular cure-directed treatment was not having any effect. In these situations, treatment was not to be escalated. This meant that parents had to identify this as the circumstance and/or moment they had in mind when they first made their decision. When two partners were present, both agreed to withhold treatment. One parent used information provided by the paramedics to decide that the situation her child was in was what she had anticipated. These processes are illustrated in Figure 2 Part B.
Pathway 2: withdrawing LMT
Withdrawal of LMT meant parents had to decide to stop an active medical process. Mostly, there was no obvious physical point in the child's condition for parents to use as a marker for decision enactment. Parents had to discern an appropriate time to implement their decision. In hospital settings, parents were helped by staff to determine the time. Several parents reported that they had asked hospital staff when would be the “right time.” Staff advised them not to worry, that they would “just know.” One parent was told when withdrawal would occur. Exemplars of discerning the moment to withdraw LMT are found in Figure 2 Part C.
Several parents decided to implement their decision soon after it had been made. Others chose a time with symbolic meaning such as the hour of their child's birth, or after family and friends had gathered to say goodbye, or following the completion of the child's usual bedtime ritual. To illustrate:
Talked about how it would happen. We were able to choose the day, the time, all of this. … He was born at ten-thirty on a Saturday and we thought that would probably be the best time. (Parent 1)
So they started to take things away … I wanted her to have the story and the nursery rhymes. … When she closed her eyes to go sleep, we always did Twinkle Twinkle Little Star … I gave her a kiss and told her I loved her … And this would be no different. I sat with her and I was doing Twinkle Twinkle Little Star [sings] … She opened her eyes and I was smiling and singing Twinkle Twinkle Little Star to her … What a moment. How beautiful is that to know that she left this earth to go on to eternal life the way that she went to sleep. (Parent 17)
Unexpected events
In most cases, the circumstances of the child's death were consistent with the end-of-life care plan, resulting in the child's death. However, for several parents, enactment did not unfold as anticipated. Several children recovered from critical health events without medical intervention deemed necessary to preserve life. Of these children, a few recovered from a medical crisis more than once. Parents did not explicitly express concern about these events. Such episodes meant parents saw healthcare professionals following their child's end-of-life care plan, which these parents found comforting. For example:
He was really sick and couldn't breathe, was turning blue and they came and that's all they [paramedics] did [gave him oxygen] … They didn't do anything else. They didn't intubate him and they took him to hospital for observation and he was home in the afternoon. So we knew that it [the end-of-life care plan and document] worked then. (Parent 11)
Other people's actions, in several cases, unexpectedly affected parents’ plans for enacting their decision. One couple had to restate their original decision to forego resuscitation when paramedics initiated cardiopulmonary resuscitation on their daughter. In another instance, a friend called paramedics who tried to revive the child, despite this not being the decision that the parents had made:
Basically she [a friend] called the ambulance … And they [the paramedics] took him into his room. I don't know [if] they [administered] adrenalin to his heart … Just unbelievable …I couldn't stop them. No, I couldn't stop them and I was trying to. (Parent 9)
These events happened at home and when the decision related to withholding treatments.
Two children died in their sleep, so their parents did not need to enact their decision to forego LMT. There had been no physical indications of imminent death. “But sadly the third day he was home he passed away. Just in his sleep, no reason … No warning.” (Parent 11)
These parents viewed this nonenactment of their decision without regret. They felt that their child had died peacefully. What they did regret not being with their child at the time of death.
Discussion
The critical finding of this study is that ELD was not completed at the point when parents had reached their decision to withdraw or withhold treatment. There was more to come in implementing the decision. The literature to date appears to regard making a decision as synonymous with implementing a decision and the decision's outcome or effects. This suggests some conceptual confusion or conflation of what are actually separate individual elements of the overall ELD process. Our findings show a stage after that of making a decision to withdraw or withhold LMT, which we call “enactment.” Enactment should be seen another stage in ELD, and not simply as a waiting time, because, as our data show, parents are required to make more decisions that affect the timing and circumstances of their child's death.
Enactment was a crucial stage for our participants. Decision enactment has particular complexities related to the circumstances in which it occurs, the potential for intervening events, and the length of time before the child dies. The significance of this stage lies, partly, in that it moves an ELD from “thought into action,” making it real. Until enactment happens, the decision-making can be seen as being in “principle” rather than in practice.
One specific new finding was the need for parents to discern or decide the moment to actually withdraw or withhold treatment. The enactment stage is not merely a passive waiting for death, but rather is an active process requiring further deliberation and more decisions. One aspect of this was that parents needed to appraise their child's physical condition, to know if this was the right time to do what they had decided. Typically, this discernment or decision occurred at home without the immediate availability of healthcare professionals. Several times, partners helped to recognize the moment. They drew attention to and spoke about what was happening for the child and what they had decided to do when this moment occurred. Recognition required a composite judgment with weighty consequences. Like the moment of death itself, it can be thought of as a profound, intimate phenomenon.
Deciding on the moment of death has been noted as an increasingly common occurrence in contemporary society (Machado, Reference Machado2005; Moore et al., Reference Moore, Kerridge and Gillis2008). This study's findings are consistent with developments in the knowledge, philosophy, and technology of modern healthcare, which, in many circumstances, enable management of the timing, location, and nature of patients’ dying and death, with the goal of a “good death” (Beckstrand et al., Reference Beckstrand, Callister and Kirchhoff2006; Leichtentritt & Rettig, Reference Leichtentritt and Rettig2002; Machado, Reference Machado2005; Smith, Reference Smith2000).
It is also consistent with current Western social views about parental rights, generally and in healthcare in particular (Gillam, Reference Gillam2008; Gillam & Sullivan, Reference Gillam and Sullivan2011). Parents regularly make judgments on their child's behalf based on their intuition, values, knowledge, and observation. However, knowing the “right” time to forego LMT is an additional modern-day layer of parental responsibility. It may be empowering for parents to have this decision in their control. The importance of a sense of control for parents’ well-being in end-of-life matters has been established (Dowden, Reference Dowden, Twycross, Dowden and Bruce2009; Hinds & Kelly, Reference Hinds and Kelly2010; Mack & Wolfe, Reference Mack and Wolfe2006; Orfali & Gordon, Reference Orfali and Gordon2004; Wocial, Reference Wocial2000). This was reflected in parents’ accounts. Conversely, having this particular responsibility of enactment could be perceived as an unwelcome burden, which may complicate grief. This was not borne out in this study. Our participants reported that they wanted to be the decision-maker for their child (Sullivan et al, Reference Sullivan, Monagle and Gillam2013), and none regretted the decision they had made. No parent expressed a desire to avoid decision enactment or have others decide the moment for them. Some did seek healthcare professionals’ guidance to help discern the moment to forego LMT. There was no indication in the data that being the enactor as well as the maker of the decision gave parents any ongoing worry or regret.
End-of-life decision enactment was not universally simultaneous with death. This has been noted in previous research (McHaffie et al., Reference McHaffie, Lyon and Fowlie2001b; Oberender & Tibballs, Reference Oberender and Tibballs2011). Although mostly relatively short periods occurred between decision-making, enactment, and death, the time created a space for family and friends to visit and complete bedtime, religious, and leave-taking rituals. Parents described these expressions of care for their child as meaningful. The time between making and enacting an end-of-life decision, whatever its duration, was precious to parents. From a positive perspective, this interval may offer a period for psychological and practical preparation, fulfilling family goals and the creation of memories. It generally did not involve an extended period of physical suffering for the child. The significance for parents of being able to “choreograph” their baby's dying has been argued elsewhere (Milstein & Raingruber, Reference Milstein and Raingruber2007).
Equally, decisions made when the child is well and some time before their death could generate an emotional tension for parents. Previous research has found that a lengthy dying process following an end-of-life decision can be difficult for parents, provoking a sense of uncertainty about the decision, whereas shorter periods were seen as affirming the decision (McHaffie et al., Reference McHaffie, Lyon and Fowlie2001b).
Significantly, the study demonstrates how unanticipated events can affect parents’ experiences. The intervention of other people in ways that interfered with end-of-life decisions and care plans for a child does not appear to have been reported previously in the literature. Interference was not anticipated by parents, who mostly found these actions unhelpful. It has the potential to be distressing, disempowering, and undermining of parents’ authority and choices, although the parents in our study who had these experiences reported coping with them.
Another unanticipated event was the need for parents to reimplement their decision when their child recovered from a health crisis without medical assistance. If enacting a decision is emotionally demanding, then the cumulative impact on parents of re-enacting is likely to be substantial. It is possible that parents may feel that by recovering, their child is showing their will to live and hence that they have not made the right decision. Such experiences could contribute to feelings of uncertainty and distress. However, a sense of making the wrong decision was not expressed by any parents in this study. Rather, parents accepted recovery as a reflection of the nature of their child's diagnosis or character. The child's recovery did not unsettle their decision. These episodes meant several parents could test out their child's end-of-life care plan and know that their wishes would be respected. No parents in this study experienced the ethical dilemmas recently reported by Kutzche (Reference Kutzsche, Partridge and Leuthner2013) in situations where a child survives following ventilation withdrawal.
Study limitations
The study was based on a voluntary self-selected sample of parents, which is open to selection bias, but is only the ethically acceptable way of conducting such a study. Parents were retrospectively self-reporting their recollections of ELD, decisions, and experiences with which they now live. This may be viewed as a limitation. However, this is the exact perspective that is needed for a better understanding of how parents make meaning of their decision-making in the years after their child's death.
The sample had considerable variation in the child's medical condition, circumstances of death, and time since death. The specifics of parents’ experiences varied; however, a number of significant commonalities of experience emerged from this variation. The existence of, and the challenges associated with, the enactment stage of ELD was a key one of these. Enactment has been given scant attention in the published literature to date, which has focused very much on the factors leading to the decision to limit. Future research should examine this concept of enactment further, and investigate in more details parents’ experiences in the time after they make a decision to withdraw or withhold. Attention to cross-cultural factors with culturally diverse samples, which this study was unable to do, would be another important focus in future research.
Implications for clinical practice
ELD by parents for their children is recognized as a psychologically demanding responsibility (Armentrout, Reference Armentrout2007; Caeymaex & Speranza, Reference Caeymaex, Speranza and Vasilescu2011; Einarsdóttir, Reference Einarsdóttir2009; Hinds & Kelly, Reference Hinds and Kelly2010; Orfali & Gordon, Reference Orfali and Gordon2004). This study confirms this view, drawing attention specifically to the underrecognized enactment stage of decision-making.
This is important because healthcare professionals should inform and prepare parents, emotionally and practically, for the challenges of enacting their decision. Preparation should include information about the possible physical indicators and symptoms in the child so they can recognize the moment to enact their decision. This may be especially important for parents whose children are not hospitalized because they will need to identify when to enact their decision without the presence of healthcare professionals. Although community palliative care teams may be able to help with recognizing this time, not all families are referred to palliative care. The healthcare team needs to advise parents of the likely duration of time between decision enactment and their child's death and counsel parents about what they can do during this stage to make it as meaningful and regret-free as possible. They should encourage and allow parents to fulfill their parental role in whatever ways are possible, including deciding on the time and location of decision enactment and who should be present.
Additionally, parents should be prepared for the possibility of unexpected events affecting their decision enactment. It could reasonably be hypothesized that such events are more likely to occur outside of the hospital setting, when the child is at home and withholding LMT is being considered. Our findings highlight the need for care-planning documentation to be completed and readily accessible for parents to use should others interfere. If parents do have these kinds of experiences, having opportunities to review the event with their child's doctor and integrate what has happened would be helpful.
Conclusion
This study identifies the enactment stage of ELD as a significant, complex part of parents’ experience. This stage has not been explicitly recognized in the literature as part of ELD, and the implications for providing support and care for parents have not been comprehensively considered. This study is potentially relevant for all clinical areas where children die after a decision to limit treatment. It will be important to test and expand findings in further research with larger numbers and in different settings, including palliative care units and community services. Attention to these findings should prevent parents being left without guidance and support during the enactment stage of ELD.
Acknowledgments
The authors thank the parents who participated in the study and those who reviewed the study material for their generosity of spirit, their concern for other parents, and commitment to the professional development of healthcare professionals. The support and practical assistance given to this research study by The Compassionate Friends Victoria, The Social Work Department, The Royal Children's Hospital Melbourne, and Very Special Kids and the aid of the Australian Centre for Grief and Bereavement are gratefully acknowledged by the authors.
Conflicts of interest
LG's position is partly funded by the Royal Children's Hospital Foundation. JS was the holder of an Australian Postgraduate Award. The authors have no financial relationships relevant to this article to disclose.
