‘MAiD’ refers to a medically qualified provider ending a person’s life at their request in the setting of a terminal illness with a short life expectancy. The distinction between provider-administered MAiD (often termed euthanasia) and self-administered MAiD (often termed physician-assisted suicide) is significant, but the terminology varies by stance and region. We decided to use the term MAiD throughout and describe the administration type where relevant.

‘Canada’ includes Quebec, although the timing of legalisation and specifications of the statute were slightly different in Quebec than in the rest of Canada. MAiD was initially legalised in Quebec in 2014,9 10 and then a 2015 court decision required the Canadian legislature to legalise assisted dying for the rest of Canada by mid-2016.11 The statutes in Quebec and the rest of Canada were similar enough in the study period (2016–2019) to consider the two collectively.

‘Emotional impact’ can be positive or negative. While medical caregiving can be gratifying, many physicians undertake roles that include witnessing suffering and grief, coping with heartbreaking situations and making high-stakes decisions in conditions of uncertainty. Studies document that physicians frequently struggle emotionally when treating people with devastating illnesses and loss.12 Emotional distress in physicians working with suffering people affects their inner lives. It can lead to burnout and other negative consequences that have downstream effects on their professional lives and the quality of their future patient care.13–16 Physicians witnessing dying and caring for the seriously ill elicit ‘intense emotions arising from the care of patients’16 17 Unacknowledged emotions can result in behavioural changes and detract from ‘the quality of their medical performance as well as their personal growth in their profession.’ (Vicini et al,12 1192) The personal toll on physicians includes professional loneliness, loss of professional sense of meaning, cynicism, frustration and anger.14 Resulting in physician behaviours including avoidance, incoherent care goals, patient mistrust and failures to identify patient and family values that influence decisions. (Meier et al,14 3008) The inner lives of physicians at the front line of high-stakes clinical medicine and the related phenomenon of emotional regulation and burnout in these professionals have become an area of intense scrutiny.18

Animal care workers are described as challenged by the ‘killing-caring paradox’,1 the emotional and moral stress caused by euthanasia provision within veterinary care, which is found to contribute to provider burnout.19–22 The robust literature about the impact of the caring-killing paradox on animal workers was helpful for considering the challenges of providing euthanasia.23 24

‘Moral impact’ is intertwined with the emotional domain but can be defined and described separately, such as in the nursing literature about moral distress. Physician-centric literature describes that the stress (and distress) of caring for dying patients impacts physicians in both emotional and ‘spiritual’ domains.14 25 We use ‘moral domain’ rather than ethical or spiritual. Moral philosophy and moral psychology assist with understanding and describing the experience of MAiD providers. ‘Moral intuition’ and ‘moral justification’ are concepts described in moral psychology that other MAiD researchers have used to describe moral responses to MAiD.26 Haidt writes that human brains first experience non-verbal moral intuitions and then justify intuition through moral reasoning.27 ‘Moral responsibility’ describes a situation where a person has enough choice and control to have a ‘causal role’ and thus accountability for an outcome.28 29 MAiD providers have ‘agency’ in that they choose to provide MAiD for a specific patient in a specific circumstance. A person with moral responsibility may perceive a ‘moral burden’ if they later question the decision-making that led to their actions. We summarise the moral domain as including the intersection between a provider’s pre-existing ethical stance, the provider’s attempt to resolve presenting clinical ambiguities, their recognition of ethical tensions, and, finally, internally justifying one’s choices.

Key requirements of Canadian law that may influence emotional or moral impact include the manifest causal link between the provider’s role and the patient’s death. Additionally, the stipulation that the provider be present with the patient from administration of the lethal medications until after death ensures that the provider witnesses the entire process, including family grief. It also accentuates the provider’s experience of the differences between the duration of completion between the oral and intravenous routes. With the intravenous route, death occurs within 3–15 min in most cases.30 The oral route allows a possible prolonged period between administration and death. Since the law came into effect in Canada, the intravenous route has accounted for the vast majority of MAiD deaths in Canada.31 The shortened interval between administration and death makes the cause-and-effect relationship obvious, which may increase the provider’s sense of responsibility for the death process.

Canadian providers are also longitudinally responsible for the patient in the interval between qualification for MAiD and death. After assessment of eligibility, the provider has continuing obligations to ensure all procedural safeguards are met, to obtains, use, store and return the pharmacologic substance. The many roles of the MAiD provider are outlined in the statute32 33 and are described and analysed in detail elsewhere.34 This is different from the ‘assisted suicide’ model, which minimises the provider’s responsibilities after the qualification assessment. The Canadian provider’s role after qualification includes negotiating a time and place, responsibility for picking up medications and transporting them, injecting the lethal medication, proximally causing and witnessing death. After death, the Canadian provider has continued responsibilities including paperwork, disposing of unused medications and comforting and taking leave of a grieving family.

Newly implemented MAiD:. An international review of healthcare providers’ perspectives on implementing MAiD summarised 33 publications on the topic.35 Known stressors from providing MAiD include provider mistakes and unexpected harm, family conflict and the perceived departure from cultural norms.35–37 Departure from the traditional moral stand is an important but time-specific challenge for physician providers of newly legalised MAiD. As societal norms changed and public demand for MAiD grew, some physician groups articulated opposition to it coming under the umbrella of medicine38 because many thought that doctors hastening death could undermine the covenant of trust between the doctors and society.39 40 For many countries, including Canada, during the 2014–2016 legalisation process, the traditional ethical codes for physicians continued to predominate and medical professional organisations opposed MAiD. As these organisations set physician standards, acting in opposition to the traditional social contract posed tensions for physician providers. This was described in a recent Australian study as physicians having to ‘individually determine’ the way they participate in MAiD due to their ‘profession’s response and collective belief’ that MAiD ‘does not form part of medical practice (and) this tension underpins one of the most enduring professional challenges raised by participants’ (Rutherford et al,37 1064).

Canadian provider experience: The information available on provider experience in Canada has grown rapidly since we completed data collection in 2019. When our study was developed, very little had been published about the experiences and challenges of Canadian MAiD providers. In one study, conducted just a few months after legalisation of MAiD,41 found that providers reported being undercompensated financially for their time, sacrificing personal time due to increased workload and that MAiD provision affected some collegial relationships. There was concern that providers could not continue to provide the service over time without addressing these challenges to providers.41 Studies also indicate that initially there were few organised supports available for providers, including a paucity of precedents, guidance, mentorship or evidence-based educational support for their individual circumstances.42

Oregon provider experience: Before Canadian legalisation, the North American experience with legalised MAiD was confined to ‘assisted suicide’, which was legalised in Oregon in 1997.43 Almost 30 years of research document various aspects of the Oregon model, which was later emulated by other US jurisdictions.44 Beyond provider administration of the lethal medication, the Oregon model does not require continued involvement of the provider. The Oregon providers’ obligations end when the prescription is given to the patient while the Canadian provider continues to have a role central to the entire process. The Oregon patient must pick up, store the medication and then choose when and how to ingest it, and many who complete the qualification procedure choose not to proceed to pick up or ingest the medications. Research from Oregon reveals that only about half of the people who asked for MAiD and completed the qualification process actually ingested the lethal substance.44 In the Oregon model, the link between provider role and death is diffused by time, location, and the patient’s initiative in taking the lethal substance.

Reports of provider experience in Oregon are mixed. Studies indicate that doctors sometimes describe participating in MAiD as difficult, troubling, extracting a significant emotional toll and ‘an excruciating thing to do’.43 On the other hand, a 1999 survey of Oregon doctors reported that only four of the 144 providers who returned the survey expressed ‘ambivalence’, and only one reported deciding not to provide it again.45 Seminal research by Emanuel et al reported data about physicians who provided MAiD in the USA.36 46 These authors identified an inverse correlation between the doctors’ perception of the severity of physical pain expressed by the patient and the oncologists’ acceptance of the acceptability of MAiD.36 A majority (54%) of interviewed provider oncologists expressed comfort in knowing they had helped a patient end their life the way they wished. One-quarter of the providers reported regret after providing MAiD for specific or non-specific reasons. Specific reasons for regret included reflecting on a patient who ‘might have benefited from living to the end’ and feeling empathy for family members who did not get to say goodbye. Non-specific regrets included worries about ‘playing God’ and feeling conflicted or ‘at odds’ with themselves or their role. One in six providers reported experiencing emotional distress severe enough to affect their medical practice.46

Experiences in other jurisdictions: A multinational 2006 article reviewed publicly available provider statements (mainly from The Netherlands and Oregon) and reported ‘emotional and psychological effects’ of MAiD. This study concluded that

Physician participation in assisted suicide or euthanasia can have a profound harmful effect on the involved physicians. Doctors must take responsibility for causing the patient’s death. There is a huge burden on conscience, tangled emotions and a large psychological toll on the participating physicians (Stevens,47 213).

The Dutch have longitudinal experience of doctors directly administering the lethal agent because doctor-administered MAiD has been legal since 1984. Research indicates that doctors sometimes find MAiD difficult or troubling.48–50 However, it has been argued that the experiences of Dutch providers cannot be directly compared with those of North American providers, given the vast differences in the cultural and medical milieu (Norwood,51 pages 94, 153).

In summary, cross-cultural literature reveals reports of retrospective regrets or burdens on some physician providers of MAiD, but these seem to occur in a minority of those interviewed in qualitative studies and have variable effects on the providers’ well-being. Based on these studies, we wondered if a small but not inconsequential subgroup of doctors involved in the provision of MAiD in Canada would experience negative emotions or regret.36

Our study presents participants’ perceptions of the emotional rewards and challenges of providing MAiD and their moral reactions and reasoning. We wondered if self-reports of provider experience in Canada might vary from other jurisdictions because of the specific statutory conditions influencing the subjective experiences of those working within these conditions. Relevant elements for provider experience included (1) requiring the provider to be responsible for steps in between qualification and provision, (2) requiring the provider to be present from administration until death and (3) allowing the provider to deliver a rapid-acting lethal agent directly and intravenously (instead of self-administration). If these constitute a burden on the practitioners, then we would expect that to appear in accounts of their subjective experience. During the conceptualisation phase of our study, the impact of these factors had not yet been studied in Canada. This study explores the self-reported provider experience when the providers were required to be face to face with patients and causally responsible for administering a rapid-acting lethal substance.

Table 1

Participant self-reported demographics and professional attributes

Inductive and ChatGPT-enhanced thematic analysis provided congruent themes on provider emotional and moral impact. Factors that enhanced or detracted from participant experience of MAID provision were extracted. The excerpts from our participants include at least one excerpt from each study participant and no participant’s voice is repeated within the same category.

Our research results provide valuable information that can be used for the normative arguments made when legislating and regulating MAiD. Doctors are rarely thought of as a vulnerable group despite the fact that they take on many moral burdens in their work and are known to have high rates of substance abuse and suicide.63 Doctors’ occupations are stressful in many ways, and difficult aspects include witnessing crises, suffering and death and making morally difficult high-stakes decisions in conditions of uncertainty.

Our results support the idea that both patient and provider autonomy enable providers to feel connected to patients and morally comfortable with their work. Providers determine whom they approve as qualifying for MAiD and for whom they would provide. Our participants spoke about the importance of moral judgements and the difficulties of medical uncertainties that affect their moral reasoning. A strong sense of connection to the patient, combined with provider reliance on the patient to know what was ‘right for them’, gave providers a sense of moral security and purpose to continue providing MAiD.

We identified variability in the impact and moral reasoning from provider to provider. We propose that future legislators of MAiD should expect variability between providers in interpreting the law and justifying whom they approve. This variation must be anticipated within the tolerances of what society intends for the breadth of assisted dying. We recommend further study about our impression that factors that enhance provider sustainability for providing MAiD (such as their connection to the patient, patient gratitude and a sense of purpose) may also influence some providers to have difficulty saying ‘no’ to patients not fully meeting eligibility criteria and thus provide MAiD at the margin or even outside the intended legal scope.

A significant limitation of this study was methodological choices that narrowed and probably skewed the pool of participants. To honour principles of research participant safety, we chose a recruitment strategy that required the participants to reach out to us via email. We also required a time commitment from participants as we chose to do extended (120 min) face-to-face interviews. Thus, our participants were restricted to providers who would take initiative and commit significant time. It is possible that MAiD providers with more negative experiences or experienced regrets would be less likely to volunteer for a lengthy interview.

Another limitation is the timing of our use of AI-enhanced analysis. We added the ChatGPT analysis in 2023 when ChatGPT was in its first months of use (introduced 30 November 2022). There was no literature to guide us on the reliability of artificial intelligence in thematic analysis. We considered our use of AI-enhanced analysis to be low risk because we used it to complement the analysis that had been completed almost 3 years earlier. We decided to adopt the novel strategy on our previously analysed data to see if it could assist in identifying human bias unknowingly embedded in our previously completed analysis. We did not find any significant differences. In summary, Chat GPT was useful in providing an excellent thesaurus of synonyms for ideas we already identified. Since that time, there are helpful publications about the strength and weaknesses of AI-enhanced analysis64 and further researchers using these will have published experience to guide use and interpretation.

Not applicable.