Abstract
BACKGROUND AND OBJECTIVE: Retrospective studies show that most parents prefer to share in decisions to forgo life-sustaining treatment (LST) from their children. We do not yet know how physicians and parents communicate about these decisions and to what extent parents share in the decision-making process.
METHODS: We conducted a prospective exploratory study in 2 Dutch University Medical Centers.
RESULTS: Overall, 27 physicians participated, along with 37 parents of 19 children for whom a decision to withhold or withdraw LST was being considered. Forty-seven conversations were audio recorded, ranging from 1 to 8 meetings per patient. By means of a coding instrument we quantitatively and qualitatively analyzed physicians’ and parents’ communicative behaviors. On average, physicians spoke 67% of the time, parents 30%, and nurses 3%. All physicians focused primarily on providing medical information, explaining their preferred course of action, and informing parents about the decision being reached by the team. Only in 2 cases were parents asked to share in the decision-making. Despite their intense emotions, most parents made great effort to actively participate in the conversation. They did this by asking for clarifications, offering their preferences, and reacting to the decision being proposed (mostly by expressing their assent). In the few cases where parents strongly preferred LST to be continued, the physicians either gave parents more time or revised the decision.
CONCLUSIONS: We conclude that parents are able to handle a more active role than they are currently being given. Parents’ greatest concern is that their child might suffer.
What’s Known on This Subject:
Retrospective studies have shown that the majority of parents, independent of their country of origin, prefer a shared approach over a paternalistic approach or an informed approach when an end-of-life decision must be made for their children.
What This Study Adds:
In actual conversations parents act in line with their preference for a shared approach. This behavior contrasts with the “some sharing” approach of physicians who carefully prepare parents for an end-of-life decision already being made by the medical team.
After the death of a child, parents’ lives are never the same again. Despite their immense grief, most parents find a way to continue living as best as they can. The factors that support them in coping with the loss of their child have received growing attention.1–15 Recent studies show that most parents find it helpful to share in their child’s daily care during the dying process. Most of them also find it helpful to share in decisions to withhold or withdraw life-sustaining treatment (LST), decisions that often precede a child’s death.2,3,5,9,11,16–20 Parents’ perception of a shared approach may even lower their grief in the long term.21 The extent to which parents want to share in the decision-making process differs from parent to parent. Some want to share in all stages of the decision-making process, including exchanging information, deliberating about preferred options, and making the final decision together (Table 1). Others want to share information and preferences with their child’s physician but feel it is too great a burden to make the final decision. Therefore, recent studies have concluded that the way parents are involved should be highly tailored to their individual preferences and needs.2,16,22–31 What we do not yet know is how parents are actually being involved in end-of-life decision-making. Little information is available about how these decisions are made in PICUs. We therefore performed a prospective exploratory study aimed at answering 2 questions: How do physicians and parents communicate about decisions to withhold or withdraw LST, and to what extent do parents share in the decision-making process?
Definition and Stages of Shared Decision-Making
Methods
Study Setting and Inclusion of Parents
We conducted our study in the PICUs of 2 Dutch university medical centers. Both PICUs are combined medical and surgical tertiary facilities, including all specialty care except that involving prematurely born infants. The institutional review boards gave their approval for the study. To identify eligible conversations, the main researcher (M.A.V.) contacted the supervising pediatric intensivist twice a week. Upon being informed about an intended conversation in which a decision to withhold or withdraw LST would be discussed with parents, she contacted the physician who would be speaking with the parents. This physician then informed the parents about the study and asked for their consent. Parents were informed about the main purpose of the study: to help physicians to improve their communication with parents about treatment choices in critically ill children. It was stressed that parents had full liberty to refuse and could end their study participation at any time. We strived to achieve a large variation of cases according to the following patient-related characteristics: age, gender, prognosis, preexisting neurologic damage, total duration of care, and length of PICU stay. Similarly, we sought variation according to the following parent-related characteristics: gender, ethnic background, first language, religion, marital status, and length of physician–family relationship. The first inclusion period lasted from April 2008 to April 2009, the second from April 2010 to April 2011 (see Table 2 for details, including the consent rates of parents and physicians).
Quantitative Context of the Study
Coding and Analysis
The recorded conversations were transcribed verbatim and anonymized. Transcripts were then uploaded to MAXQDA (VERBI GmbH, Berlin, Germany).32 We performed a search in PubMed for schemes used to codify physicians’ and parents’ communication in end-of-life decisions. In the pediatric literature we did not find any such schemes. In the adult medical literature our search eventually led to 2 relevant and useful publications. First, Charles et al33 described a conceptual framework for decision-making in the medical encounter, consisting of different analytic stages: information exchange, deliberation, and deciding on treatment to implement. Second, White at al34 described a coding instrument to assess the extent of shared decision-making in relation to specific physician communication behaviors.
In the first draft of our coding scheme we incorporated the 3 analytical stages and the coded physician behaviors. After having closely read the first 10 transcripts, we elaborated the coding elements describing physician behaviors and added coding elements describing parents’ communication behaviors. After additional refinement of this extended coding scheme, transcripts were coded by 2 researchers (M.A.V. and B.M.G.). This included a recoding of the first 10 transcripts. Conflicting views about the applied codes were resolved by discussion. Next, we composed worksheets containing all retrieved conversation segments per applied code over all cases. We qualitatively and quantitatively analyzed this material and thoroughly discussed the results in several discussion rounds with all researchers. Coding scheme and worksheets are available on request.
Results
Characteristics of Participants and of the Decision-Making Process
Overall, 37 parents of 19 patients and 27 physicians participated. Table 3 lists their demographic characteristics. In Supplemental Table 8 we present the main characteristics of the decision-making process for each patient.
Main Characteristics of Included Patients, Parents, and Physicians
Speaking Time and Contribution of Nurses
Forty-seven conversations were audio recorded. The number of meetings ranged from 1 to 8 per patient, with a median of 2 meetings. The mean duration of meetings was 30 minutes. On average, physicians spoke 67% of the time, parents 30%, and nurses 3%.
Whenever a nurse was present, his or her contribution to the conversation always occurred at the end of the meeting and consisted of a combination of emotional and practical support, as indicated by remarks such as “This must be difficult to hear” or “We’ll arrange for you both to stay with your child tonight.” The nurse then accompanied parents back to their child’s room or got parents coffee and stayed in the conference room with them for some time.
Communication Behaviors Throughout the Decision-Making Process
In all but 1 case, physicians and parents went through all stages of the decision making process: providing and receiving information, deliberating about whether to withhold or withdraw LST, and reaching a final decision. In none of the cases this was a linear chronological process. Instead, within a conversation physicians and parents moved back and forth between the stages of information exchange and deliberating before reaching a decision to withhold or withdraw LST at the end of the meeting or postponing this decision to a subsequent meeting. The next meeting typically started with an extensive update about the child’s situation and a renewed deliberation about the pros and cons of treatment continuation or discontinuation.
In 6 cases it took 1 meeting to reach a final decision, in 7 cases 2 meetings, and in the 5 most complex cases ≥3 meetings. In the remaining case the child died just after the first meeting. In this meeting the physician cautiously announced that there could come a time when continuation of LST had to be considered futile.
Stage I. Providing and Receiving Information
Table 4 gives an overview of the specific communication behaviors physicians and parents showed in the first stage of decision-making, the number of cases in which we identified these behaviors, and illustrative quotes extracted from the transcripts.
Stage I: Providing and Receiving Information
Although physicians occasionally started the meeting by explaining its goal and presenting a short agenda, in most meetings he or she immediately started with an extensive and detailed update about the child’s current serious condition, the grave prognosis, and the remaining treatment options. Much of this information was repeated several times throughout the meeting. Most physicians emphasized that everything was being tried to overcome the patient’s life-threatening problems but that treatment was becoming increasingly burdensome for the child.
In one third of cases, the physicians asked parents to provide their perspectives about their child’s situation, especially their observations about symptoms of discomfort and their child’s ability to make any form of contact. None of the parents were invited to share their views about their child’s prognosis or the remaining treatment options.
In half of the cases, parents took the initiative to ask clarifying questions throughout the meeting. These questions showed that they wanted to understand exactly what was happening to their child. Furthermore, they specifically asked which treatments were being giving to their child, whether they were working, and whether everything was being done to protect their child from pain and other symptoms. Toward the end of the meeting most physicians asked the parents whether they had any additional questions. Occasionally physicians asked them to summarize their understanding of the situation. In these cases, the parents typically demonstrated good understanding of the key issues.
Stage II. Deliberating
In Table 5 we present the specific communication behaviors physicians and parents showed in the second stage of decision-making, the number of cases in which we identified these behaviors, and illustrative quotes.
Stage II: Deliberating
In none of the cases did the physicians ask the parents how they preferred to be involved in the decision-making process. In 1 case, the parents spontaneously expressed their preference and explained that they wanted to fully share in the decision-making process, including making decisions about whether to discontinue LST together with their physician.
Almost all physicians at some point explained which decision they would prefer. They presented these preferences as team preferences rather than personal preferences.
In most cases, physicians stated that according to the team’s shared opinion it would be best if LST was withheld or withdrawn, whether in the short term or longer term. Depending on the child’s specific condition and prognosis, physicians explained this preference as follows: Despite full support it was becoming clear that the child was deteriorating very fast and would die soon (in 7 of 19 cases), continued use of LST could not overcome the underlying problems and was only causing the child suffering (in 6 of 19 cases), or reinstituting LST would put the child’s already fragile quality of life under even more pressure (in 4 of 19 cases).
In the remaining 2 cases, the physicians explained in the first meetings that they preferred LST to be continued. In their opinion, more time should be given to await the child’s possible recovery. This preference contrasted with the parents’ preference to withdraw LST to spare their already severely disabled child more suffering. Over time the child’s condition deteriorated, and in later meetings the physicians emphasized that LST had clearly become inappropriate and therefore should be withdrawn.
In the majority of cases (in 16 of 19 cases) parents presented their preferences about additional treatment, often without being invited to do so. One third of these parents (5 of 16 cases) expressed a preference for treatment to be continued or restarted, in contrast to the preference presented by the physician to withhold or withdraw LST in the near future. In these cases, the parents explained that they were not yet certain that all had been tried and that they still thought and hoped their child could recover. Several parents added that their religious beliefs strengthened this hope. Two thirds of the parents (11 of 16 cases) emphasized that they did not want treatment to be continued at all cost. In most cases this preference was in line with that of the physician; in 2 cases (mentioned earlier) the parents’ preference that LST would be withdrawn was not yet shared by the physicians. The parents of both children emphasized that their wish to try everything to save their child by now was outweighed by their wish to protect their child from pointless suffering and from living a life they considered inhumane.
All parents reacted emotionally when the physicians indicated that a decision to withhold or withdraw LST could become inevitable or had become inevitable. These emotions varied between anxiety, dismay, grief, confusion, and disbelief. The parents of the 2 children mentioned earlier described how they felt a strange combination of deep sadness and relief for their child’s sake. Several other parents described how they felt trapped in the dilemma of wanting their child to be near them as long as possible and wanting to spare their child more suffering. Parents’ often intense emotions did not prevent them from taking the initiative to ask clarifying questions and present their preferences. Although most physicians acknowledged the parents’ emotions and were empathic, none of them explored these feelings further. Occasionally, physicians addressed parents’ needs for emotional and social support.
Stage III. Reaching a Decision
Table 6 gives an overview of the specific communication behaviors physicians and parents showed in this third stage, the number of cases in which we identified these behaviors, and illustrative quotes.
Stage III: Reaching a Decision
In 2 cases the physicians asked the parents which decision they thought should be made. In these conversations, a differentiated “package of decisions” was formulated, including withholding of future resuscitation, withholding of LST in case of severe, lasting problems, and initiating LST in case of “bridgeable” problems such as an epileptic seizure or pneumonia.
In the remaining cases, the physicians did not invite parents to share in making the final decision. Instead, during the (last) meeting physicians explained that the medical team had reached the conclusion that LST had become “futile” and should therefore be forgone. In almost half of these cases, physicians added that this decision was a medical decision and not a decision parents should make. In the majority of cases parents’ reactions made it clear that they did not feel taken by surprise and understood the inevitable need to withhold or withdraw LST. Nevertheless, in 4 cases the parents repeated their strong wish that LST be continued. After acknowledging how difficult this process was for the parents, all physicians suggested a follow-up meeting for additional discussion. In this meeting the physicians again emphasized the child’s critical condition and the duty of all caregivers to not prolong the child’s suffering. Yet they also gave the parents the opportunity to clarify their points of view. The parents of 2 children eventually agreed that there remained no other choice than to withdraw LST (in both children death was clearly imminent). In the other 2 children, the physicians and the parents agreed that not all life-sustaining treatments would be withheld.
At the end of the (last) meeting, all parents expressed their great concern that their child might suffer in the process of dying. In turn, the physicians promised that everything would be done to ensure the child’s comfort and peace. Moreover, several parents asked whether it would be possible to let their child regain consciousness so they could speak with him or her for the last time.
Extent of Shared Decision-Making
In Table 7 we provide an estimate of the extent of sharing which we identified in our study. In most cases the interaction between the physicians and the parents could best be described as “some sharing.”35 This implied that information and preferences were shared, mostly at the parents’ initiative, but the decision-making was not.
Extent of Sharing Between Physicians and Parents
Discussion
When an end-of-life decision has to be made for a child, most parents, independent of their country of origin, seem to prefer a shared approach by their physicians over a paternalistic approach or an informed approach.10,12,21–23,26,27,31,36–38 Our findings indicate that in actual conversations parents act in line with this preference. Most parents in our study made an effort to actively participate in the decision-making process, especially when their initial preferences regarding the course of treatment did not correspond with those of the medical team. By contrast, most physicians were focused primarily on carefully preparing the parents for the decision to withhold or withdraw LST. This approach seemed motivated by their dual wish to not overburden parents with too much responsibility while letting them grow toward accepting the inevitability of this decision.
Studies investigating end-of-life communication with families in adult ICUs have reported a comparable focus by physicians on providing medical technical information.39–44 What our study adds is the insight that parents appreciated receiving this detailed information, as evidenced by their reactions and their questions in return. It may be postulated that the provision of relevant information helps parents to grasp what is happening to their child and regain some feeling of control. Although in most cases there seemed to be the right balance between parents’ and physicians’ communication behaviors in the first decision-making stage (exchanging information), our data also suggest that physicians’ and parents’ behaviors in the second stage (deliberating) and third stage (reaching a final decision) were often out of balance. In these cases a better balance could have been reached if the physicians had actively invited the parents to explain their points of view. Moreover, although the physicians firmly believed that they had the final responsibility for whatever decision was reached, the parents would have appreciated and preferred an affirmation that the decisions were being reached in collaboration with them and with respect for their views and values.
Parents’ striving to fully share in all decision-making stages and physicians’ striving to act as the final decision-maker both seemed strongly motivated by an urge to protect. Parents wanted to guard their child’s life and, even more important, to protect their child from suffering. Physicians wanted to guard the child’s life and well-being, but they also wanted to protect the parents from (later) worries and guilt. Yet, this protection was not in line with what parents asked for, as indicated by their efforts to actively participate in all decision-making stages. In this respect our findings have important practical implications. First, they may help physicians become more alert to parents’ communication behaviors and to how these behaviors interact with their own. This alertness will help them tailor their approach to parents’ approaches in the different decision-making stages. Second, our findings may also prove helpful in the ongoing debate, at least in the Netherlands, regarding the role of minors and their parents in end-of-life decision-making. According to Dutch law, the judgment about the appropriateness of treatment continuation or discontinuation is a medical professional judgment in which the views of minors and their parents should be taken into account.45 It could seriously be questioned whether this law is still suited to the actual practice in which parents want to share in deciding whether treatment has become inappropriate.
In our study, most parents eventually expressed their assent to the decision being presented to them, convinced that there remained no other choice. In a minority of the cases the parents kept their resistance, mainly because they still hoped and believed that their child’s situation would improve. In these cases the physicians did not put pressure on the parents. Instead they either gave them more time to be with their child and to witness the ongoing deterioration despite full support or revised their decision to meet the parents’ wishes. This outcome confirms the finding of earlier studies that physicians make considerable efforts to prevent potential conflicts from escalating.46–48 Seen in this light, the choice of physicians to come forward to the wishes of the parents who disagreed could be characterized as a conflict-solving strategy. At the same time it could be considered unfair that parents who express their strong objections may influence the decision-making process to a greater degree than parents who comply with the decision being proposed. We do not yet know which emotional effects compliance or noncompliance may have for parents in the long term. On one hand, it can be hypothesized that parents who comply to the recommendation to forgo LST may later feel that they have given up too soon and blame themselves for not expressing their doubts and concerns. On the other hand, parents who resist the recommendation to forgo LST may later feel guilty about not allowing their child to die peacefully. Follow-up studies of these long-term effects are needed to better support individual parents throughout the decision-making process and in their later bereavement after their child’s death.
Our study suggests that in most cases parents’ intense emotions of anxiety, grief, and distress did not hinder them from asking relevant questions and from clearly explaining their considerations and preferences. It could even be argued that these emotions, especially the parents’ fear of losing their child and their anxiety of seeing their child suffer, heightened their alertness and awareness. In this respect there was no apparent difference between parents whose child had acutely fallen ill and parents whose child had a congenital disorder. This finding implies that the capacity of parents to understand complex issues should not be underestimated, despite their stress and grief. At the same time, it remains important that physicians use clear, unambiguous words, do not lose themselves in too many details and complex hypothetical scenarios, and keep encouraging parents to ask questions and present their points of view.40,43,49–53
Our study has several limitations. An important limitation is that our study was restricted to 2 medical centers, which was unavoidable given the logistical demands of the study. Moreover, we cannot rule out that the presence of a researcher influenced the course of the conversations. The greatest strength of our study is that we were able to explore in detail how physicians and parents communicate about end-of-life decisions. To do this we developed a coding instrument to systematically explore their communication behaviors. This instrument needs greater validation but may prove useful in international comparative research of actual end-of-life discussions by parents, physicians, and nurses. Although our study did not find evidence that nurses contributed to the decision-making during the “key conversations” we observed, this does not imply that nurses did not play an important role in other settings. Sometimes it was the nurses who first raised the issue of redirecting care in multidisciplinary team meetings. In addition, at the bedside nurses often spoke with parents about the information they had received earlier that day and explored parents’ reactions. It may be that the views that parents expressed in follow-up meetings were views that they had come to develop in these conversations at their child’s bedside. In future research it would be of interest to observe both formal and informal discussions between parents and caregivers and compare their impact on the decision-making process.
Conclusions
Most parents in our study took considerable effort to actively participate in the decision-making process, especially when their initial preferences did not correspond with those of the medical team. By contrast, most physicians were focused primarily on carefully informing the parents and not on actively inviting them to share in the decision-making. In most cases parents’ intense emotions of anxiety, grief, and distress did not hinder them from asking relevant questions and explaining their considerations and preferences. Physicians made considerable efforts to prevent potential conflicts from escalating. Our study reveals that parents feel the same double duty as physicians, making certain that everything is tried to save their child’s life while protecting their child from suffering, without sufficient quality of life. What parents fear most is that their child will suffer in the process of dying.
Acknowledgments
We are very grateful for the trust of the participating parents, physicians, and nurses. Moreover, we thank Hanneke Swinkels for her valuable assistance in the data collection phase of this study and Prof Hugo Heymans, Prof Hans van Goudoever, Prof Johan Legemaate, Prof Ellen Smets, Henk Galekamp, and Petra Hollak for their helpful comments and advice.
Footnotes
- Accepted November 7, 2014.
- Address correspondence to Mirjam A. de Vos, MSc, Section of Medical Ethics, Department of General Practice, Academic Medical Center, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, Netherlands. E-mail: m.a.devos{at}amc.nl
Ms de Vos conceptualized and designed the study, did the data acquisition, developed the coding scheme, supervised and carried out the analyses, and drafted the initial manuscript; Drs Bos, Plötz, van Heerde, Truog, and Willems conceptualized and designed the study, contributed to the analyses, and reviewed and revised the manuscript; Mr de Graaff developed the coding scheme, carried out the initial analyses, and reviewed and revised the manuscript; Dr Tates conceptualized and designed the study, carried out the initial analyses, and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: This study is part of a research project titled “Tailor-Made Decision-Making Within Dutch Paediatrics. How Parents Can Be Better Involved in End-of-Life Decisions for Their Children.” This project is funded by the Netherlands Organization for Health Research and Development (ZonMw). ZonMw did not in any way influence the data acquisition, data analysis, or drafting of this manuscript.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
References
- Copyright © 2015 by the American Academy of Pediatrics