Clinical ethics committees have been established in many hospitals in many countries during recent decades.1^–9 One of the main tasks for such committees is to discuss concrete clinical cases that raise important ethical problems. Empirical research observing such deliberations is scarce, with some exceptions.10

The purposes of this pilot study were to explore the content and dynamics of the committee deliberations and to identify areas for improvement. This study is part of a national research project initiated in 2004 to assess the quality of the work of Norwegian hospital ethics committees and to facilitate exchange of experiences. The other parts of the national project have been published elsewhere.11^–14

Norwegian hospitals are organised into well over 30 hospital trusts serving a population of 4.7 million people. The healthcare services are by and large publicly funded and most of the hospital trusts provide comprehensive services. The first clinical ethics committees (CECs) in Norway were established in 1996. By 2008, 38 such committees had been established in the hospitals.

METHOD

An invitation was sent by mail to all 24 CECs that were established by May, 2004. Initially our intention was to observe real case deliberations. However, we soon realised that this was not possible, because too few cases were brought to the committees within the 8 months dedicated for data gathering. (Possible reasons for this are explored in another article.12) Therefore we decided to observe the committees deliberating a paper case (box 1).

The committees were asked to approach the case as if this was a written submission to the committee and to deliberate for 25 minutes based on the information in the paper case. After 20 minutes, the committees were asked to use the last 5 minutes to sum up the deliberations and to focus on what the committee’s response and follow-up would be if this were a real submission. The committees were asked to indicate any lack of information or to make explicit assumptions, if these were needed to continue the deliberations.

The committees’ deliberations on the paper case were followed by a 25-minute semistructured group interview. The interview guide was developed to clarify and sum up key aspects of the deliberations (box 2).

The observations and interviews were performed by the first and second authors together, and the whole session was recorded and transcribed verbatim. Then the written texts were studied by all authors independently and themes that could be used to describe the content of the texts were agreed on. The first and second authors then analysed the texts through the chosen themes, first independently and then together. Finally, the analysed material was condensed and the article was drafted by the first author. All authors read through earlier drafts of the article in order to scrutinise the findings and control for haphazard or biased subjectivity in the analysis.15 Preliminary results and a final draft of this article were presented to all the participating committees to validate the analyses and interpretations. The study did not include patients or patient data and therefore was not within the mandate of the Norwegian regional ethics committees.

RESULTS

The first paragraphs of this Results section describe the participating committees and our observations of their deliberations; the subsequent semistructured interviews with the committees are presented in the last paragraphs.

The participating committees

Nine committees, representing three of the five health regions, were included in the study. All the committees were interdisciplinary. For further details on the committees’ composition, see table 1.

Of the nine committees that participated, the newest had been created a year earlier and the oldest, 10 years earlier. On average the committees interviewed had more experience than Norwegian CECs in general. Seven of the nine committees had previously done case consultations involving a particular patient. All committees considered this to be part of their mandate. Formal ethics training is not required for members or the chair of the CECs, but it is recommended that every committee should have some ethics competence among the members.

Deliberation procedures and chairing

Seven of the nine committees claimed to have a specific procedure or method for case deliberations, but two of these hardly used the procedure. Three different deliberation procedures were used. In five of the committees, the procedure included the following five steps:

• identifying the facts;

• identifying the involved parties;

• identifying the values, principles and virtues at stake;

• describing possible courses of action;

• balancing the different interests and considerations to attain an acceptable conclusion.

One of the committee members chaired the deliberations—most often the committee chair. Three committees used a blackboard during the deliberations. All the chairs attempted to structure the deliberations—for example, by use of a deliberation procedure or by inviting everyone to express their view before summing up the deliberations.

The use of a well-known, plain, simple deliberation procedure and a blackboard seemed to make the deliberations more encompassing, explicit, transparent and systematic. On the other hand, if such structuring was too rigid, it seemed to hamper the dialogue—for example, the possibility of clarifying arguments and disagreement. One committee chair used a rather complex procedure that was new for him. In this deliberation, the procedure seemed to hamper the flow, coherence and intelligibility of the discussion.

The content and dynamics of the deliberations

In our analysis of the transcribed deliberations, 13 dimensions that describe key elements of the committees’ case deliberations emerged (box 3).

In all the deliberations, the elements in Box 3 were closely interwoven and the elements did not appear in succession. Rather, the discussions typically leapt from one element to another.

All the committees explored relevant moral values. However, moral principles and theories were not used extensively, and when they were used, it was to pinpoint important values such as patient autonomy, rather than to deduce a certain conclusion. No committee made use of only one moral theory. Instead, pluralistic approaches dominated.

Various interpretations

The participants’ experience, knowledge, and interpretations seemed to influence the committees’ analysis of the ethical problem and to some extent also their conclusions. For example, one committee assumed that the patient’s general condition was not very poor and that the patient’s vital organs had still not failed, and thus one could not know for certain that cardiopulmonary resuscitation (CPR) would be futile or more harmful than beneficial. Accordingly, CPR could certainly be appropriate if the cause of heart arrest was cardiac rather than pulmonary cancer and lung failure. In other committees, it was assumed that the patient’s general condition was very poor and that cardiac arrest would be caused by lung failure due to the cancer. Consequently, the patient would either die or become an intensive care patient who would be impossible to wean off the respirator. From this perspective, one committee member argued:

It is after all very bad ethics to start CPR when the patient’s condition is so poor, and I thought that it would be terribly frightening for the relatives if the patient develops pulmonary oedema or dies, and then we should call the CPR team, and when they start it is quite frightening for the relatives; completely appalling, it is hard enough to watch your husband lying there dying; this [CPR] I think is an undignified end of life.

The importance of clinical experience and knowledge

Initially, quite a few members saw no reason why one could not comply with the patient’s wish. Many regarded CPR as a rather limited procedure and not very costly, and considered that the patient would die anyway. Where clinicians with relevant experience were present, it was emphasised that if CPR was performed, the patient might end up in an intensive care unit, unconscious and completely dependent on respiratory treatment. Such sharing of clinical knowledge and experience seemed to make questions of resources, professional integrity and dignified death more relevant. In two committees, relevant clinical experience also seemed to contribute to the identification of a possible compromise, that is, to avoid chest compressions but to include defibrillation, ventilation support and medication.

Relevant clinical experience and knowledge, as of oncology or intensive care, also seemed to facilitate the exploration of missing information and uncertainties—for example, the prognosis and the patient’s general condition. The committees in which members with relevant experience were absent during the deliberations seemed to take clinical facts and judgments presented in the case more for granted—for example:

The therapists do certainly have valid reasons for their opinions and conclusions, both when it comes to inflicting pain and injuries to the patient with fragile ribs, and considering that this is a terminal patient.

CEC members with clinical experience also helped to make the patient’s perspective more concrete—for example, to consider that the patient and his wife might regard the therapists’ judgments of the prognosis to be too pessimistic if the patient’s general condition and symptoms were still not very serious. Clinical experience was also important to evaluate the quality of the information given. For example, as one experienced physician with a relevant subspecialty commented that

… the physicians say that there are only a few weeks left … this I have never told a patient, because this you cannot know. They can die tomorrow, or in several weeks … such information about number of weeks I think is totally inappropriate.

Quite a few CEC members emphasised how the therapists’ attitudes and experience might influence their judgments about incurable disease, death and futile treatment. One physician member expressed it like this:

One perspective that we have not touched upon in our deliberation is how to deal with futile treatment and our [ie, clinicians’] paternalistic tendencies—to know better than the patients—we should at least not forget to scrutinise this in our own thinking in such a situation.

Some committee members did share their experience from analogous situations, but this was done less frequently than we anticipated, considering that most members were experienced clinicians. Furthermore, the normative relevance or implications from the experiences mentioned was rarely explicated. However, when such concrete and practical experience was introduced, it did seem to enrich the deliberations—for example, pointing to a possible solution: one physician with experience of treating similar patients commented on the possibility of denial, and added that when the patient gets more symptoms and understands the possibility of being choked slowly,

… then my experience is that the patients realise that it may be better to die from heart arrest. I think one should let some days pass and then raise the issue again, rather than saying “this is how we do it”.

The same physician added that in one analogous situation in which CPR had been performed, the patient ended up unconscious on a respirator. In this instance, it took a week of intensive care treatment before the relatives reconciled themselves to the inevitability of death and accepted withdrawal of the respirator treatment. Experience from analogous situations also helped to make alternative goals of life-prolonging treatment more explicit—for example, to stay alive until a child’s birthday, to give time for the relatives to understand, to maintain hope or to avoid giving the feeling of being abandoned. One physician said that he recently had used ventilator treatment for a limited time to buy time for the relatives:

Yesterday I put a child on a respirator; I was convinced that the child would not survive, but the mother and father wanted to baptise their child and the hospital chaplain had still not arrived … the parents needed me to ventilate the child for 2 hours; so here there are grey zones—not every medically futile treatment is objectionable.

Exploring the clinical communication

All the committees emphasised the communication process and the need to clarify the perspectives of all the involved parties and how well they had understood each other’s perspectives. As one committee member stated, “How the dialogue has taken place and progressed is really always the crux of the matter.” The exploration of the clinical communication was often interwoven with the committee members’ interpretation of the medical situation. For example, some committee members considered that the issue of CPR had been raised too early, since the patient’s general condition was presumed not to be very poor. Others presumed that the patient was in a too advanced stage of the disease to raise the issue of CPR (since the patient had no real option). Although the timing of the communication about CPR was quite often criticised, the committees rarely explored what would be an appropriate timing.

One committee member wanted to openly discuss with the patient how many resources the hospital could expend just to comply with the patient’s wish or to sustain hope. From his perspective, the treatment of other patients could suffer if no limits were drawn, and these limits should not be concealed. Another committee member commented, “I would certainly not raise the question about resources in such a situation; this is not of interest for a family in deep crisis.” Most of the other members of this committee agreed that focusing on costs would be devastating to the communication in a delicate situation like this. Costs were a peripheral subject in most of the deliberations.

Clarifying key concepts

Most of the committees did not attempt to clarify key concepts in the case (eg, “CPR”, “DNAR” (do not attempt resuscitation), “futile treatment”, and “dying”). However, when this was done, it often triggered relevant discussions. For example, in one committee a clarification of what CPR means led to a discussion about possible respirator treatment and about resource-intensive treatment lacking a clear-cut medical indication.

Clarifying legal regulation

In most of the committees, some members asked what kind of actions would be lawful—for example, is it lawful for the healthcare team to make a DNAR decision unilaterally if resuscitation is considered futile? However, this kind of question was often not answered or was sometimes answered erroneously. Although only one of the participating committee members present in this study was a jurist, only three committees mentioned that they would have considered seeking legal advice. The committees’ legal considerations are explored in depth in another paper.11

Answers to the interview guide

The main ethical problem

The main ethical problems were often identified not during the committees’ deliberations, but rather as a response to the interview guide. Some participants initially did not see any ethical problems in the case. One committee member commented that she often found it difficult to identify moral problems in her work on the ward, and added that “if this [the paper case] represents an ethical problem, then we’ve got a load of moral problems in our department.” However, during the deliberations and in particular during the following interview, all the committees agreed that there was not just one ethical problem but several moral challenges in addition to “CPR or not”—for example, concerning futility, professional integrity, not to harm, dignified death, clinical communication and resource allocation.

Identification of the involved parties and how to involve them in the deliberations

All committees identified the same involved parties. However, there was a wide variety of opinions concerning how to involve them, and the committees seemed relatively offhand when they attempted to answer this question. Most committees were ambivalent or reluctant to invite the patient or his wife to a deliberation, for various reasons, such as that that it was too uncomfortable for the patient, that the committee had too little experience with patient participation in the committee deliberations, that the committee’s primary role is to guide or support the clinicians or that the presence of the patient or his wife could cast a damper on the dialogue. One committee chair commented:

We have not been in the habit of inviting the involved parties in such acute cases … But I thought that if we were to support someone, then the committee’s task should be to support the treatment team here and possibly to meet them for a dialogue and a discussion to attempt to guide them. Our assistance here should not affect the authority or the confidence in the treatment team, but rather reinforce it so that they can appear in a proper way towards the relatives.

However, after a while a committee member added that the information from the relatives and patient might not be satisfactory if “we only get information from the treatment team”.

In two committees, there were participants who considered that it would be absurd (due to serious illness) for the patient or his wife to meet up with the committee members. On the other hand, some committee members did find it important to invite the patient, his wife or someone not directly involved in the treatment who could represent the patient. Some reasons stated for this were the need to carefully explore communication aspects also through the patient’s perspective, and, in case of conflict, it would be dubious if one of the adversaries should represent both sides. One committee member asserted that

… although we have defined ourselves as a committee mainly for the employees, we have to secure that we do not neglect the obligation to attend to any party, and not only the perspective of the employees.

Possible conclusion and follow-up

Not surprisingly, given the limited information in the paper case, the committees reached various conclusions, at least partly dependent on their interpretations of the medical facts. One committee concluded that resuscitation seemed to be futile and thus ought to be avoided. Another concluded that one should avoid the “DNAR” code, but only perform defibrillation. A third committee concluded that resuscitation should be attempted, and that the discussion of DNAR had been taken too early and should be postponed until the patient got worse. A fourth committee concluded that it was impossible to conclude due to lack of information—for example, about organ status and the involved parties’ perspectives. Four committees disagreed on the recommended code (DNAR or no code) and some also disagreed about what one ought to do to in the case of cardiac arrest if no code was written. One committee did not discuss whether to write “DNAR” or not.

All the committees emphasised the need for adequate clinical communication.

Although most committees in general wanted to make a written report, many of the committees were unsure of how to do this—for example, who should get the minutes and who should prepare the documentation (the committee or the clinicians). One committee explained that the main reason why the committee sometimes does not make a written report is that it is too time consuming.

Four of the committees stressed that they would communicate verbally with the involved parties after the deliberations, since they did not perceive written feedback as sufficient.

The committees’ evaluative comments

Six of the nine committees mentioned that the experience with this deliberation had uncovered a need to strengthen the committee in breadth of competence through a new member with clinical expertise, a jurist, a patient representative, or an ethicist. Three of the committees using a deliberation procedure felt that they needed more practice applying the procedure to structure their deliberations in a better way.

In all the committees, some of the members were absent because of other duties, and many of the physician members present had to leave during the deliberations. Four of the committees felt that this had negatively influenced their deliberations, making the exploration of medical facts or legal aspects inadequate.

Four committees commented that in a real case consultation they would have used more time and gathered more information and that it would have been possible to contact the involved parties or relevant expertise. However, five of the committees considered that their deliberations had been quite representative of what a real clinical ethics case consultations could be like.

DISCUSSION

Because our study was a qualitative pilot study including nine committees deliberating a particular paper case within restricted time limits, there is limited potential to generalise to other contexts. Future studies might benefit by including a case with richer descriptions, giving more time for the committee deliberations and, if feasible, observing real deliberations. However, using a paper case made it easier to compare the committees’ deliberation strategies, and our observations of the committee deliberations did provide some tentative insights.

First, we identified some closely intertwined key elements in the committee deliberations (see box 3). Second, our observations indicate that the composition of the committees, the members’ attendance and the members’ experience and interpretations are likely to influence the content and results of the deliberations. Third, we observed variations in interpretations, conclusions and procedures. Such variations are typical of clinical ethics challenges.10 16^–24 This is an important reason why multidisciplinary, systematic, transparent discussions are needed in the face of complex clinical ethics challenges. Variability and the need to clarify standards in clinical ethics consultation services have also been underscored in other studies.9–10 12 14 22 Fourth, the skilled use of a deliberation procedure seemed to make the deliberations more transparent and comprehensive. With many people and professions around the table, it is a challenge to perform the deliberations in a limited time. Thus, skilled facilitation and structuring are important in clinical ethics deliberations.

Our study did clearly identify possible areas of improvement for the committees involved. Examples are documentation routines and inclusion of the involved parties. There may be valid reasons for the patient not to be present during the deliberations, but all the involved parties’ perspectives and interests should be appropriately represented. Good routines in clinical ethics deliberations are important to include key elements, such as identification of the main ethical problems, scrutinising key concepts and legal regulations and exploring analogous clinical situations. Some of the challenges identified may be best dealt with through better procedures and targeted training, while some also indicate a need for broader representation in some of the committees—for example, inclusion of an ethicist, a jurist or a patient representative. These possible amendments are supported by the other parts of this research project.12^–14