Even though the majority of the doctors were opposed to permitting assisted dying, it is worth noting that as many as 30.7 % of the doctors agreed partially or strongly that assisted dying ought to be permitted for persons who are fatally ill and have a short remaining life expectancy. This is as large a proportion (31 %) as among fifth- and sixth-year medical students who were questioned about this issue in 2012 (2). A considerably higher number of doctors (12.7 %) than medical students (5 %) were in favour of assisted dying in cases of chronic illness. However, the medical students did not have the opportunity to be more specific in their response of 'agree' or 'disagree'.
Our study provided the respondents with the opportunity to be more specific, but this also opens room for interpretation: what does it mean to 'partially agree' or 'partially disagree' that assisted dying ought to permitted? Would all these respondents have answered either 'agree' or 'disagree' if these had been the only response alternatives? In our analyses we have assumed that the respondents were either in favour of or opposed to legalisation, but it is reasonable to assume that the specific design of a law on assisted dying would influence these doctors' opinions. They would probably be more receptive to arguments in favour of or against legalisation than those who chose the extreme responses. Viewed in this perspective, a large proportion of Norwegian doctors appear not to have had an unshakeable opinion, but might be willing to change their views on the issue of assisted dying.
Comparisons with previous surveys are made difficult by the fact that definitions and formulations of the questions differed. However, to the extent that the 1993 survey Box 1) was biased, our interpretation is that it tended to sway the respondents towards taking a positive view of legalisation. In our view, the distinction between assisted dying and decisions to withhold treatment seemed somewhat unclear, and assisted dying was portrayed in a positive light by charged phrases such as 'help to die', 'well-considered' and 'painless manner' ((3). Since there was more support for assisted dying in 2016 than in 1993, despite the possibility that the formulations in 1993 may have swayed the respondents towards a positive answer, we believe that there is reason to assume that there has been a real change of opinion among doctors.
Nonetheless, doctors are still considerably less in favour of legalisation than the population in general. In 2015, the three first questions in Table 1 were also posed to a sample of the population (1). To the assertion that 'physician-assisted suicide should be permitted for persons suffering from a fatal disease with a short remaining life expectancy', 37.5 % responded 'strongly agree', 35.6 % 'partially agree', 7.3 % 'neither agree, nor disagree, 6.7 % 'partially disagree' and 12.9 % 'strongly disagree'. The majority (66.5 %) also agreed strongly or partially to the assertion regarding euthanasia in case of fatal illness, while 37.9 % agreed strongly/partially to the assertion concerning assisted dying in cases of chronic illness. Please note, however, that the survey had a low response rate.
With the assertion that 'there are cases in which it may be right/morally defensible for a doctor to provide assisted dying, even though it is illegal' we wished to investigate whether doctors who were opposed to legalisation would nevertheless think that there might be situations in which such an act could be appropriate. While one-quarter of all respondents believed that assisted dying may sometimes be appropriate in spite of being illegal, this view was also shared by 12 % of those who were opposed to legalisation. On the other hand, the majority of the respondents report that they would not have provided assisted dying, even in a situation where this was permitted.
These findings probably reveal something both about the way in which doctors perceive assisted dying as a dilemma in terms of professional ethics and about their attitude to healthcare legislation. Although an act may be illegal and generally regarded as unethical, there might be extreme situations where healthcare personnel feel that the law and the ethical norm have unreasonable consequences and that there are good grounds to violate them in order to do what is perceived as appropriate for the individual patient. Doctors face dilemmas that are caused not only by value conflicts, but also by role conflicts (11, 12). Doctors fill at least four – potentially conflicting – roles: as society's gatekeeper with responsibility for ensuring that laws are abided by and resources fairly distributed, as the patient's spokesperson, as a professional and as a private individual. In special situations, a small minority of doctors might be willing to break the law or refuse to implement statutory practices (11). A study from 2014 indicated, however, that very few Norwegian doctors have provided assisted dying and thus transgressed the boundaries of the law (4).
The strength of this study is that a representative sample of Norwegian doctors have been asked, and that the response rate is high. The challenge is that although we have sought to use precise definitions, our questions and response alternatives nevertheless leave room for interpretation. For example, we cannot know how the respondents interpreted 'short remaining life expectancy'.