Decision-making at the limit of viability: differing perceptions and opinions between neonatal physicians and nurses

A large majority of the respondents indicated a range of problems when decisions about limiting intensive care for an extremely preterm baby must be made (Table 2). 94% of all HCPs found it difficult to foresee the patient’s future quality of life, and 90% named the difficulty to make an accurate long-term prognosis. The following difficulties with end-of-life decision-making were reported more frequently by nurses than physicians: difficulty of interpreting parents’ attitudes precisely (92% vs. 82%), insufficient time for decision-making (81% vs. 54%), legal constraints (80% vs. 54%), lack of a consistent unit policy (73% vs. 36%) and conflict between your own principles and unit policy (60% vs. 40%). In comparison to physicians, nurses also mentioned significantly more often a lack of solidarity in our society (46% vs. 22%) and a shortage of services for disabled (46% vs. 12%). Lack of a consistent unit policy was more often indicated in the French speaking area (81%) than in the German speaking area (58%).

Administering sedatives and/or analgesics to suppress pain even if this might cause respiratory depression and death was acceptable to 95% of all respondents (Table 3). 24% accepted administering drugs with the explicit purpose to hasten death. Compared to physicians, nurses significantly less often indicated that withholding intensive care (83% vs. 100%), refraining from increasing respiratory support (66% vs. 80%) and withholding full parenteral nutrition (50% vs. 75%), tube feeding (28% vs. 45%) or antibiotics (67% vs. 80%) would be acceptable options when it comes to limiting intensive care.

More than half of the respondents (60%) thought that parents should have the opportunity to take part in the decision-making process; this opinion was shared more often in the German speaking than in the French speaking area (64% vs. 49%). Moreover, 15% thought that parents should always have the opportunity to decide the course of action for their infant. Nurses (25%), as compared to physicians (12%), were more in favour of not directly involving parents in decision-making; they indicated that parental wishes and attitudes should be explored indirectly and considered by the decision-making health care team (25% vs. 12%). Only 0.5% of all HCPs thought that parents should not be involved at all, but merely be informed about the decision (see Additional file 2).

HCPs gave several reasons for not directly involving parents in decisions on treatment limitations (Table 4). Firstly, 95% considered that parents should not be involved because they might change their minds later and experience feelings of guilt. Secondly, they felt that parents should be spared the burden of such decisions (90%), with significantly more nurses taking that standpoint than neonatologists (94% vs. 71%). Thirdly, parents might not be in the right state of mind to take such decisions (76%), again with more nurses agreeing with that view (81% vs. 50%). Finally, nurses more often than physicians (82% vs. 36%) stated that parents canot fully understand the possible options and consequences (74%). Interestingly, 47% of HCPs took up the view that the responsibility for such decisions belongs solely to the physicians (nurses 47% vs. 46% physicians, ns); this position was significantly more often cited in the French speaking than in the German speaking area (60% vs. 37%).

In the event that parents would request limitation of intensive care, while HCPs recommend continuation of treatment, 43% of respondents considered hospital ethics committees to be the ultimate decision-makers and 31% felt that this would be the right of the parents. Another 10% indicated the medical staff and only very few (3%) chose the court as ultimate decision-makers (Fig 1a). In the opposite situation, namely if the parents request continuation of intensive care, while HCPs think that life-sustaining therapies should be withdrawn, 36% of all respondents considered the ethics committee to be the ultimate decision-maker, 20% named the parents, 16% the medical staff and only 1.5% the court (Fig. 1b). Nurses differed significantly from physicians in both scenarios.

The role of hospital’s ethics committee was seen as follows: 80% of respondents indicated to give advice in individual cases, 12% being responsible for ultimate decision-making in individual cases - the latter was favoured more often by nurses (14%) than by physicians (6%) - and 5% only to set general guidelines; this view was taken more often in the German speaking area (6% vs. 0.9% in the French speaking area) (see Additional file 3).

A majority of neonatologists argued that decisions to withhold life-sustaining therapies from infants born at the limit of viability should not be made prenatally because prognosis is not sufficiently accurate (52%) and individual assessment after birth allows a more nuanced approach (54%) (Fig 2). There was no consensus among physicians whether parental wishes and the child’s best interest should be viewed differently before than after birth: 25% of physicians agreed that, prenatally, parental wishes and values should be given more weight than the child’s best interest, whereas this no longer is true after birth; 51% disagreed with this statement.

This survey reveals important differences between nurses and physicians regarding end-of-life decision-making in extremely preterm infants. These differences concern the perceived importance of specific challenges in the decision-making process, of acceptable approaches to limiting intensive care and of the degree of parental involvement in the decision-making process. For instance, nurses considered insufficient time for decision-making, legal constraints and lack of a consistent unit policy significantly more often problematic than physicians did. Moreover and although the large majority of neonatologists and neonatal nurses agree that withholding intensive care is acceptable in futile situations, nurses were more reluctant to withdraw respiratory support and tube feeding. Several speculations can be put forward when trying to explain these discrepancies between neonatal nurses and physicians.

First, differences may exist concerning education in medical ethics and ethical arguing. In Switzerland, physicians traditionally had a longer and perhaps more intensive training and exposure to ethical dilemmas, starting at medical school and continuing into professional life. Over the last years, medical ethics have also been incorporated into the curriculum of neonatal nurses and most units provide a continued training including ethical topics such as futility, over- and undertreatment, withholding and withdrawing lifesaving interventions.

Second, the differing views may be due to the different roles physicians and nurses play within the end-of-life decision process of extremely preterm infants [21]. Physicians are expected to have an understanding of the prognostic and outcome data of infants born extremely premature. Often considered less emotionally attached to the patient as nurses, they provide ‘an expert’ point of view to the team and parents. This medical judgement however is encumbered with significant medical prognostic uncertainty. It is therefore no surprise that physicians rather than nurses indicate prognostic issues as challenging. The active involvement of neonatal nurses in end-of-life decisions in NICUs was implemented only in recent years [22, 23]. Prior to that, the opinion of the nursing staff was not, or not explicitly, taken into account [24].

Third, the discomfort reported by nurses regarding some forms of therapy limitations (e.g. withdraw tube feeding, refraining from increasing respiratory parameters) may stem from the fact that the nurses will be more directly exposed to the patients and the parents. In some Swiss NICUs, a framework for decision-making has been introduced and has shown to reduce stress for both, physicians and nurses [25].

The results of our survey indicate that today the Swiss HCPs are more willing to involve parents in the decision-making process than 10 years ago; however, they still do not give them full authority [26]. The trend towards increasing parental involvement in end-of-life decisions has also been reported in other countries [27‐29].

The Swiss Civil Code for the protection of adults and children, enacted in 2013, emphasises parental authority when a child is incapable of making his or her own decisions. Shared decision-making between HCPs and parents is proposed by many national guidelines [7, 16]. It is a process aiming at a collaborative decision. This process may not be easy and disagreement, rarely reported in the past, has become more frequent in recent times [30]. Alandagady et al. observed that parents quite often do not agree with HCPs on limitations of life-sustaining therapies [31]. In this context, it is noteworthy, that in our study, HCPs give more weight to parental opinions in cases of disagreement when parents request limitation of intensive care compared to situations where they ask for continuation of intensive care. Although the difference between withholding and withdrawing intensive care measures are morally equivalent, these differences might be considered relevant in practice [32]. In both cases, a large percentage of respondents think that a hospital’s ethics committee should be the ultimate decision-maker. This stands in contrast to a Swiss population survey, where a large majority of lay-people expressed that parents should be the final decision-makers [12].

There are differences between nurses and physicians in how and to which extent parents should be involved in the decision-making process. Nurses were more in favour of integrating the views of the parents indirectly rather than to directly confront them with the choice between several options and to embark in a shared decision-making. In Switzerland, neonatal nurses are fully and directly involved in the decision-making process, together with the physicians and parents. However, it could be that they are more prone to embrace the ‘expert opinion’ of physicians regarding the prognosis and outcome for a given infant than to incorporate parental values, which often include aspects of family autonomy, parental obligations towards other children and own legitimate self-interests. As stated by Leuthner, the concept of best interest and medical expertise on the course of neonatal diseases and outcome data is a too restrictive concept of decision-making [32]. Only the addition of parental values gives meaning to the prognosis. Moreover, within the triad physicians, nurses and parents, it is often the case that over the first days and weeks when it may come to a decision-making process, the nurses have the most close and intimate contact with the premature infant. This fact may add to the greater difficulty for neonatal nurses to acknowledge and integrate parental values and attitudes within the decision-making process.

HCPs who believed parents should not be (directly) involved in the decision-making on whether or not to limit intensive care, gave several reasons for doing so. These reasons, however, are empirically unfounded or proven false. For example, Caymeax et al. showed that greater involvement of parents in end-of-life decisions was associated with lower levels of grief [33].

Prenatal decisions differ in several aspects from decisions made after birth not only because the legal status of an unborn child differs from that of a liveborn infant, but also because the prognosis is much more uncertain [34, 35]. Therefore, a prenatal decision not to resuscitate an infant born at the limit of viability should be based on strict criteria and take into account the wishes and preferences of the parents [6, 36].