Design

We conducted a descriptive and comparative field survey employing both quantitative and qualitative methods. The results presented here form part of the larger observational Northern European evaluation project on MCD,1 6 17–19 studying different existing MCD practices.

Sampling

A convenience sampling method according to observational design was applied, recruiting workplaces in Northern Europe: the Netherlands, Norway and Sweden. These workplaces, to our knowledge, had planned to implement MCD in the near future due to an expressed need for reflection. Heads of departments or teams were approached first by phone, then through a formal invitation letter. In the Netherlands, heads of institutions or MCD facilitators planning to implement MCD contacted us (Molewijk AC, VUmc) . In Sweden, managers of workplaces in provinces in Middle Sweden with access to MCD facilitators that had communicated a need for ethical reflection were approached. In Norway, managers in provinces included in a governmental project to implement ethics reflection in community care were also approached. In addition, one care unit in somatic care was included.

In total, 73 workplaces in 16 provinces within four healthcare settings were recruited (table 1). Healthcare professionals in these workplaces who had no prior MCD experience were invited to participate.

Data collection and measures

Data were collected through a survey, distributing the Euro-MCD instrument6 to either all healthcare professionals on the workplace or the professionals selected to participate in MCD, prior to the start of MCD being set up. First, researchers provided verbal information about the study during workplace meetings. Second, the instrument was distributed to individual professionals, either on paper in their pigeonholes or electronically by email or through a web-based questionnaire, depending on the preferences of each workplace. Two reminders were sent. When distributed, the instrument was accompanied by an information letter about the voluntary nature of responding, and informed consent was obtained by virtue of them having responded. Responses were handled confidentially. The professionals were also briefly informed about the common denominators for MCD (see the Introduction section) and were given the following definition for an ethically difficult situation: ‘a situation in which you experience unease or uncertainty about what is right or good to do or there is disagreement about what should be done’.6

The Euro-MCD instrument6 contains 26 possible MCD outcomes, sorted into the following six domains: ‘enhanced emotional support’, ‘enhanced collaboration’, ‘improved moral reflexivity’, ‘improved moral attitude’, ‘impact at organisational level’ and ‘concrete results’.

In this study, the instrument was administered before the professionals participated in MCD, and asked about their perceived importance of the outcomes. The instrument was also distributed after their participation in a series of MCDs and the results of that survey is published  elsewhere 20 (asking also about experienced outcomes). In the present study, the following three questions were used:

1. Open-ended question: ‘Please formulate in your own words 3 to 5 outcomes that you consider important to reach in order to support you and your co-workers in managing ethically difficult situations in everyday clinical practice’ (instructed not to read ahead).

2. Closed questions for each of the 26 predefined outcomes: ‘How important is the outcome to you?’ A four-point adjective response scale was used: ‘not important’, ‘somewhat important’, ‘quite important’ and ‘very important’. The option ‘cannot take stand’ was also offered.

3. Fixed-choice question: ‘Finally, please list 5 of the above outcomes that you consider as most important (of the 26 outcomes)’.

The instrument was translated into Dutch, Norwegian and Swedish.6

Analysis

Outcomes perceived to be the most important

Based on the quantitative analysis, all 26 outcomes in the Euro-MCD instrument were perceived as either quite or very important by 76%–97% of the respondents (table 3). There were missing responses (including the option cannot take stand) averaging 14 missing responses for each item (2%) (Table 3).

Outcomes in the domain enhanced collaboration were rated as most important, comprising more open communication, better mutual understanding and mutual respect among coworkers. The other prominently important outcomes concerned the domain concrete results, covering items about enabling decisions on concrete actions and finding more courses of actions in order to manage the ethically difficult situation. Outcomes assessed as least important comprised mostly outcomes in the domain improved moral attitude, such as listening more seriously to others’ opinions, and having the courage to express an ethical standpoint (table 3). The results of the fixed-choice question about the five most important outcomes (perceived from the list of 26) are also presented in table 3 (bold items) and these correspond with the above-mentioned results concerning most important outcomes.

Differences in perceptions among respondents

The scandinavians perceived 23 of the 26 outcomes as significantly more important compared with the Dutch respondents (table 3). Professionals working in community or disabled care services, nurse assistants, women, older respondents and those with more years of professional experience, were significantly more likely to perceive most of the outcomes as quite or very important. Respondents working in psychiatry, physicians and men, perceived most of the 26 outcomes as significantly less important as the other groups (but still found most outcomes quite important) (table 4).

The multivariable analysis appeared to provide better explanations and showed that differences († in tables 3 and 4) could mostly be explained by the variable ‘country’ in 16/26 items, but also indicated that many of the differences could be explained by the variable ‘gender’ (14/26), and some by age or being a physician (or both, in 8/26 items). Regarding differences between healthcare settings, it appeared that, after adjustment for the variables of country, gender and ‘professional group’, none of these differences were statistically significant. See the online supplementary file for fuller description of the analyses (OR).

Further subgroup analyses of healthcare settings and healthcare professions represented in more than one country, that is, within the group of registered nurses and within somatic hospital care, were conducted. This also showed country differences. In somatic hospital care, 21 outcomes were perceived as statistically significantly more important by the Swedes compared with the Dutch. The Scandinavian nurses perceived 18 outcomes as significantly more important as did the Dutch nurses.

The largest statistically significant differences of perceptions of importance between various subgroups concerned the items ‘greater opportunity to have a say’; ‘I listen more seriously to others’ opinions’; ‘strengthens my self-confidence when managing ethical difficult situations’; ‘enhances possibility to share difficult emotions and thoughts with coworkers’; and ‘enables me to better manage stress caused by ethically difficult situations’. These items mainly belong to the domains enhanced emotional support, enhanced collaboration and improved moral attitude (tables 3 and 4). Considering these findings in light of the domains (tables 3 and 4), multivariable analysis showed that the differences in perception of importance of items in the domain enhanced emotional support could especially be explained by the variable country (Scandinavia vs the Netherlands (p<0.001)). However, these differences could also be explained by the variable gender (p<0.01 to p<0.001 for these items). Being Dutch was also an explanation for scores of less importance in the domain improved moral attitude (p<0.001), but this could also be explained by being a physician (p<0.05 to p<0.001). The domain enhanced collaboration was significantly more highly valued in Scandinavia, while some of the differences among the items within this domain could also be explained by being a woman or older. Outcomes in the domain concrete results revealed the least differences between all subgroups (tables 3 and 4).

Old and new outcomes based on the open-ended responses

The qualitative analysis of the responses to the open-ended question, produced, in total, 82 different kinds of outcomes.

Reflection on perceived important outcomes in relation to goals of MCD

The top outcomes of ‘collaboration’ and concrete results fit well with the theoretical background of MCD (ie, hermeneutics, pragmatism and dialogical ethics) in which mutual dialogue and practical usefulness are import key values of MCD.23 24 The main outcomes considered as important were apparently the need to communicate and understand each other better, as well as to determine concrete actions to take. This finding is consistent with previous MCD evaluation literature.8 14 The above-mentioned theories presuppose that, in order to learn what to do in an ethically difficult situation, a joint learning process is needed, in which everyone expresses and shares their viewpoint on what is morally right. The MCD participants become open towards each other’s viewpoints and they get to know and understand each other better.14 25 Hence, openness towards one another and better collaboration are both important preconditions for and results of moral learning.12 25

Furthermore, according to the theoretical background of MCD, MCD always starts with a moral challenge that is experienced in a concrete situation. It does not primarily aim at a theoretical insight or a final conceptual definition.23 25 Rather, MCD aims at learning to deal with ethically difficult situations,6 improving the quality of care and learning about what is morally right, based on moral reflections and reasoning.23 Hence, the focus on reaching concrete results as an outcome of MCD fits well with MCD’s normative aim of improving practices and learning through reflection about concrete situations.23 25

Besides the top domains, the new domain revealed in the responses to the open-ended questions, better interaction with patient/family, was an important reminder to not forget to focus on ethics support outcomes for the patient and for improving the quality of care as the basic goal of and justification for ethics support.14 25 26 The main reason why this domain was not included in the original six domains of the Euro-MCD was that these outcomes were not found in the extensive literature search and were not suggested in the Delphi panel as the basis for the development of the instrument.6 This is supported by the recent publication regarding the content of MCD in the Swedish component of the Euro-MCD project: establishing a responsible relationship with the vulnerable patient formed the basis for the participants’ moral reasoning and can be understood as relational autonomy.18 Furthermore, this study showed how relational-oriented ethics may form a foundation for principle-based moral reasoning during MCD. This element, and paying more attention to the direct impact of MCD on patient care, is something that we will consider when revising the Euro-MCD instrument.

Discrepancy between MCD goal and a priori perceived importance of outcome

An essential element of MCD is reflecting on moral questions emerging from concrete experiences by means of moral reasoning and engaging in a joint critical moral inquiry.23 MCD has been described as aiming to improve moral competencies.27 It is therefore remarkable that the outcomes deriving from the domains of moral reflexivity (eg, analysis skills) and moral attitude (eg, courage) were not perceived as the most important outcomes. Perhaps the respondents did not explicitly think about improving their moral competencies in the first place. In fact, if this explanation is accurate, this assumption fits well with the pragmatist approach of ethics teaching, that is, that moral competencies are learnt by doing (eg, while reflecting on concrete cases).

Considering the differences between subgroups

Most of the differences in perceived importance between the subgroups (profession, healthcare setting and so on) can be explained by the variables, country and gender. However, some of the differences might also be explained by the variables ‘age’ and ‘professional background’. Regarding professional background, the nurse assistants, who dominate Scandinavian community care, perceived most of the outcomes as significantly more important than the other professions. An explanation for this could be that nurse assistants in general have fewer opportunities for attending team meetings or educational activities, while at the same time being confronted with many ethically difficult situations in their daily work. The physicians found many outcomes significantly less important, but with large variation (56%–96%). This may be interpreted as their having a better confidence to discriminate between outcomes and/or simply valuing MCD less than other professions.

The finding that female respondents rated so many items higher than male respondents is surprising. It might be due to differences in perceived moral distress, as it could be assumed that experiencing a higher level of moral distress would contribute to a higher need for ethical reflection, and perceiving outcomes such as better stress management or feeling more self-confident as more important. In the literature, we found some evidence for gender differences in moral distress. Possible explanations have been provided by, for instance, Lutzky and Knight,28 who suggested that men and women experience similar levels of moral distress, but that men may be reluctant to acknowledge their distress or may not even be aware of it, leading to biased results when assessing moral distress by use of self-reporting questionnaires.28 More recently, this gender difference was found again in a study about experiencing moral distress among critical care nurses in the US.29 We could therefore say that the possible influence of gender differences in experiencing moral distress, or in their ways of completing questionnaires, was also observed in our study. However, the female respondents form the majority of the sample (81%) and the male respondents were mainly drawn from those who work in the Netherlands , and who worked in psychiatry, as physicians or as therapists. Therefore, the differences between gender might overlap with the differences between countries. But, because of the low sample size of male respondents, we were not able to further disentangle this possible influence.

There are several possible explanations for why the variable country showed large differences in ratings. First, there might be cultural differences regarding the rating across the countries, and one can only speculate about the reasons. One explanation might be that Scandinavian yearn for a forum for exchange and reflection, while in Dutch healthcare, various forums are more established (eg, in psychiatry, where 53% of the Dutch respondents worked). A second explanation could be the different approaches to responding to self-reported questionnaires in the three countries. Jürges30 found that the Swedes are more likely to report good or better health than respondents in all other countries.30 This tendency of Swedes, and perhaps all Scandinavian respondents, might also have occurred in our study. Third, the mode of administration of the questionnaire might have caused some differences between countries. However, no major differences in answering questions have been found in recent overviews.31 Therefore, we think that, with regard to ratings of importance of outcomes of MCD, this might be less of an issue here.

Another possible explanation for these variations relates to the differences in performing MCD in the different countries. That the Swedes and Norwegians valued outcomes related to the domain enhanced emotional support to a higher degree than the Dutch is in line with results obtained from a previous Swedish study about what MCD participants talked about during the MCDs linked to this project. A median of 29% of the spoken time was spent on reflections on the psychosocial work environment.17 This raises the normative question as to what degree emotional support and psychosocial reflection should be a core component of MCD outcomes. Within the theoretical understanding of MCD that adheres to an Aristotelian view on emotions, emotions can be seen as part of moral wisdom and should therefore be an element of MCD.11

Finally, the data suggest that some of the differences could be explained by age. It seems that older respondents (>50) perceived many outcomes as being more important than the younger ones. An explanation might be that these older respondents have had more experience with difficult ethical situations and thus express a stronger need for engaging in ethical reflection.

Weighing empirical results versus normative thinking about MCD outcomes

It is only after collecting the perspectives of those who have engaged in MCD that the overall normative discussion on determining the appropriateness of MCD outcomes can begin. In this discussion, we, as authors, take a middle position in that we assume that neither theoretical viewpoints nor empirical results alone can determine what ‘the’ right MCD outcomes are. This means that, although respondents found outcomes relating to moral competencies (ie, moral reflexivity and moral attitude) somewhat less important as compared with other Euro-MCD outcomes, they could still be considered as important, given the fact that ethicists and MCD facilitators argue that MCD aims at, among other aims, fostering moral competencies.32 Given the limited scope of this paper, we will elaborate on the integration of empirical findings from all Euro-MCD field studies and our normative reasoning about appropriate MCD outcomes in a future paper. Finally, we should not conflate the findings related to the importance of MCD outcomes with the aims of MCD; the outcomes and aims of clinical ethics support are not the same. Different groups and different countries seem to prefer different outcomes and different aims. For example, although not studied explicitly yet, we know anecdotally that ethicists state a more limited number of aims of clinical ethics support. Furthermore, their aims are usually focusing more on the moral question and ethical analyses of the reasoning and arguments used. Future research on these different ranges of aims and preferred outcomes of MCD may have implications regarding how to introduce MCD within healthcare institutions, how to train the future MCD facilitators and on how to structure and steer the MCD sessions.

Strengths and weaknesses

A major strength in the study was the large number of responses enabling multivariable analysis. But a weakness was the heterogenic sampling of healthcare settings between the countries, which complicated comparisons between countries. However, the multivariable analysis provided evidence for healthcare setting not being associated with differences in responses. Furthermore, our main goal of the Euro-MCD project was to further develop the Euro-MCD instrument and to find out whether MCD makes a difference at all. The heterogeneity of inclusion is in line with the observational design, meaning not interfering with the real world, that is, the organisation of the MCD practices. However, in order to make a better generalisation, a larger field study is needed with more even distribution of subgroups in the different countries as well as including countries outside Northern Europe. This will, however, be postponed until the instrument is revised.

The survey was organised differently in the three countries with regard to recruitment of potential respondents and to the format of the questionnaire (paper, web-based, email). This might have affected the response rate in Scandinavia, as the Norwegian was web-based without personal contact and in Sweden, the questionnaires were distributed besides in pigeon holes also on information meetings and reminders on the coffee room tables. Another reason of low motivation to respond might be that the Norwegian part was associated with the governmental project. However, as the results of perceived important outcomes were similar between Sweden and Norway, we interpret that the differences in response rate may not have influenced the result. In Sweden, there were more respondents but fewer workplaces included and in Norway the vice versa, which complement each other. Unfortunately, we do not know the exact response rate for the Netherlands, but the estimated response rate is in line with other questionnaire studies.

Finding that almost all of the outcomes were perceived as quite or very important might indicate both a weakness and a strength of the Euro-MCD instrument. A weakness is the lack of discrimination between items and a ceiling effect in the Scandinavian results. A strength would be the good validation of the instrument, particularly as both the qualitative and the quantitative analysis revealed more or less the same important outcomes. We are surprised by this result, as we purposively included all possible MCD outcomes with few normative preferences.6 The lower ratings of some items, such as those relating to ‘courage to express my ethical standpoint’, might imply a need for reformulation instead of deletion. A further weakness is the nature of open-ended questions, which cannot contribute with the same richness of information as qualitative interviews can.