Article Text

Download PDFPDF
Medical assistance in dying in hospice: A qualitative study
  1. James Mellett1,2 and
  2. Mary Ellen Macdonald3
  1. 1John Dossetor Health Ethics Centre, University of Alberta Faculty of Medicine & Dentistry, Edmonton, Alberta, Canada
  2. 2Biomedical Ethics Unit, McGill University Faculty of Medicine and Health Sciences, Montreal, Quebec, Canada
  3. 3Division of Oral Health and Society, McGill University, Montreal, Quebec, Canada
  1. Correspondence to James Mellett, John Dossetor Health Ethics Centre, University of Alberta Faculty of Medicine & Dentistry, Edmonton, AB T6G 2T4, Canada; jmellett{at}


Objectives The modern hospice movement has historically opposed assisted dying. The 2016 legalisation of medical assistance in dying (MAID) in Canada has created a new reality for Canadian hospices. There have been few studies examining how the legalisation of MAID has affected Canadian hospices. Our objective was to identify the challenges and opportunities hospice workers think MAID brings to a hospice.

Methods This qualitative descriptive study included four focus groups and four semistructured interviews with Canadian hospice workers at two hospices, one which allowed MAID on site, and one which did not. Thematic analysis was used to understand and report these challenges and opportunities.

Results We constructed five themes. These themes detailed participants’ beliefs in the abilities of hospice care, and how they felt MAID challenged these abilities. Further, participants felt that MAID itself created challenging situations for patients and families, and that local policies and practices led to additional institutional challenges. Some participants also felt that allowing MAID in hospice provided opportunities for more extensive end-of-life options.

Conclusions The legalisation of MAID in Canada has created both challenges and opportunities for Canadian hospices. A balancing of these challenges and opportunities may provide a path for Canadian hospices to navigate their new reality. Increasing demand for MAID means that hospices are likely to continue to encounter requests for MAID, and should enact supports to ensure staff are able to manage these challenges and make best use of the opportunities.

  • ethics
  • hospice care

Data availability statement

Data are available on reasonable request. Reasonable requests for study data should be directed to the corresponding author, or to the Institutional Review Board at McGill University’s Faculty of Medicine (514-398-8302).

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key messages

What was already known?

  • The hospice movement has opposed assisted dying.

  • Assisted dying has had diverse impacts on hospice care worldwide.

What are the new findings?

  • Medical assistance in dying (MAID) has challenged hospice identity and aspects of practical care.

  • MAID has created opportunities for hospice.

What is their significance?

  • Clinical

    • Canadian hospices will need to balance these challenges and opportunities.

  • Research

    • Furthering knowledge of assisted dying and hospice.


Recent changes in the Canadian political and legal arenas now allow some patients to seek assistance from physicians and nurse practitioners in ending their lives, a procedure known nationally as medical assistance in dying (MAID) and l’aide medicale á mourir (AMM) in the French-speaking province of Quebec.1 Canadian hospices now face the reality that assisted dying is a healthcare option for many patients at end of life. While both hospice and MAID share the common aim of providing patients with control over the dying process, and are rooted in the values of patient dignity, and the relief of suffering, the hospice principle of not hastening death conflicts fundamentally with MAID.2–4

Research from international jurisdictions has shown that the legalisation of assisted dying has affected hospice communities in a variety of ways. In Belgium, assisted dying and the country’s broader palliative care community have developed a symbiotic relationship. For example, in some regions, palliative care staff are involved in assisted dying consultations, and even in the assisted dying procedure. Some palliative care facilities allow assisted dying on site.5 One regional palliative care organisation has endorsed assisted dying as an option within palliative care.5 In the USA, the hospice programmes in states that allow assisted dying permit varying levels of participation in assisted dying.6 7 Some allow for hospice staff to discuss with patients their interest in assisted dying, provide patients with information about the law, and direct patients to organisations that can facilitate assisted dying.7 These hospices allow staff to be present at the time of assisted dying, and permit staff to assist patients should complications arise.7 Other hospices have taken a separate approach, forbidding staff from discussing assisted dying with patients and from attending when the patient consumes the life-ending substance.7 Still others have taken more moderate paths, falling somewhere between the latter two approaches.7

Research on the experiences of Canadian hospice palliative care workers since legalisation has shown challenges have emerged, including emotional and ethical difficulties for staff.8–10 However, there have been few studies exclusively examining the experiences of Canadian hospices since legalisation. We specifically wanted to examine how the shared and divergent values between the good deaths provided by MAID and hospice communities could create challenges and opportunities for a hospice. The study described in this article addresses this knowledge gap.


Methodological framework

We used qualitative description as a methodological framework. Qualitative description aims to describe a phenomenon through exploration of participants perspectives of, and experiences with, the phenomenon.11 12 Qualitative description does not require researchers to interpret the meanings of the phenomenon within the participants’ lived experience (eg, as required by phenomenology).13 Rather, qualitative description requires researchers to stay close to the words of participants and is therefore useful when there is limited prior research on a subject, providing a foundation for future, more interpretive, research.11 12

Sampling and recruitment

This project employed two interrelated sampling processes, one for the selection of hospice sites, the other for the selection of hospice workers. Two hospice sites were invited to participate in the project, one that offered MAID (Site 1) and one that transferred patients off-site to receive MAID (site 2). Site 1 is located in Alberta, Canada, and site 2 is in Quebec, Canada. Both hospices are in-patient facilities that are physically separate from a hospital or other acute care centre and are operated by third part organisations rather than by a government health authority. Both sites are staffed by interdisciplinary teams that include physicians, registered nurses, licensed practical nurses, care aids and administrative support staff. With the help of key informants in each site, a convenience sample was used to recruit individual participants from a range of professions. Our recruitment goal was to maximally include all hospice workers interested and available during the study timeline. The key informants advised on effective recruitment strategies and distributed advertisement posters through the organisations’ internal email as well as in staff areas. The first author also spent time at one of the research sites to discuss the study with staff and ask if they were interested in participating in the study.

Data generation and analysis

Data generation occurred between August 2019 and February 2020. We used focus groups as the primary method of data generation, as they provide a space for participants to share and reflect on their perspectives and experiences with each other.14 As an exception, individual interviews were offered to any staff who were unable to participate in the focus groups, or who were uncomfortable sharing their views in a group setting. Individual interviews and focus groups were conducted concurrently. Before the commencement of a focus group or interview, participants were made aware of the study’s goals, and its role as a requirement for JM’s MSc degree. While we did not explicitly use informational redundancy as a formal end point, we did notice repetition in the data at each individual site, and then once sites were combined. Field notes and audio recordings were taken during the sessions.

We drafted the interview guides with open-ended questions to allow participants the freedom to describe experiences and perspectives they believed relevant to the study. As is common in qualitative research, these guides were modified throughout the study to explore novel perspectives and experiences raised in prior focus groups and interviews (see online supplemental file 1). The interview guide used for both individual interviews and focus groups was largely the same, save for variation that occurred from interview to interview, or focus group to focus group, as the guide was modified to explore new perspectives. All focus groups and interviews were conducted by JM as part of his master’s thesis project in Biomedical Ethics. An undergraduate research assistant assisted in note taking for two focus groups and one individual interview, while another master’s student assisted in note taking for another focus group. No one else was present for the interviews or focus groups.

We used thematic analysis guided by Braun and Clarke to make sense of the data.15 To develop themes, JM first assigned a code using QDA miner lite to each segment of the data that was relevant to the research question.15 Each code was named so as to briefly describe what was said by participants in the data segment. JM then wrote a narrative summary of each focus group or individual interview, in which similar codes were grouped together, their respective data segments combined and summarised, and a descriptive category name assigned that described the summary.

With the supervision of MEM, JM then wrote a synthesis of all narrative summaries that combined similar descriptive categories and their respective data summaries. Data from focus groups and individual interviews were integrated by connecting responses from individual interviews to collective discussions that had occurred in the focus groups.14 16 Together with MEM, he interpreted and combined descriptive summaries through the lens of our theoretical framework to develop themes that described the challenges and opportunities participants believed MAID brought to their hospice.

Throughout the study, we engaged in a process of reflexivity to explore how our participation in the study would impact the data and themes generated, and how these themes were interpreted and discussed. JM is a junior qualitative researcher at the master’s level, and through the study learnt to plan and execute a qualitative project. He was supervised by MEM, an anthropologist and senior qualitative researcher with PhD and postdoctoral training working at a major Canadian university. MEM’s supervision ensured that the project was conducted in accordance with established qualitative principles and standards, and that thematic analysis was performed in accordance to the principles of Braun and Clarke.15 JM had previously met some participants at one hospice where he had volunteered. While this previous association may have made participants feel less comfortable participating, he felt that the impact was the opposite: participants were actually more comfortable given their previous relationship with him. JM’s previous volunteer experience in hospice care was the reason for his interest in developing the study. During data analysis and interpretation, JM continuously reflected on how his past experiences at that hospice, and relationships with staff, impacted this process and strived to minimise this impact and report data with maximal integrity.

Theoretical framework

We developed the theoretical framework of Hospice as a Place to use as a sensitising construct for this study, as we were unable to locate other frameworks relevant for Canadian hospices. This framework combines Yi-Fu Tuan’s Humanist Perspective on Place with Timothy Kirk’s Hospice Care as a Moral Practice.17 18 When combined, Hospice as a Place provides a framework to understand hospices through the interrelation between the hospice space, and the people and activities contained within. This construct was used as a baseline to understand the impact of MAID on Canadian hospices.


We conducted four semistructured focus groups and four semistructured individual interviews, with two focus groups and two individual interviews occurring at each site. In total, 24 workers participated in the study, including physicians, registered nurses, licensed practical nurses, resident care aides, managers, outreach workers, social workers, spiritual care workers, unit clerks and administrative support workers. Thirteen participants were recruited from site 1 and 11 from site 2. The samples at both sites were similar in terms of occupation, however, site 2 had more administrative and outreach workers. The majority of participants identified as female. No participants later asked for their contributions to be removed from the study. All participant quotes have been rendered to English for the benefit of the reader and to maintain confidentiality. All sessions save for one interview were conducted at the hospice sites, with one interview being conducted over the phone.

Challenges to the identity of hospice

Participant responses highlighted how the availability of MAID on-site in a hospice challenged what they felt to be the identity of their workplace. This theme is rooted in three concerns. First, participants felt that the primary purpose of a hospice was to provide ‘hospice care’ and the good death this care facilitates; allowing MAID on-site was seen to have fundamentally challenged this purpose. For example, a site 1 nurse said:

To me, MAID is totally opposite from what we're trying to do here, because MAID just puts a finish to everything, and we're actually here to not necessarily give people a longer life, but to that quality of life right, and, and their comfort, so…

Included in this challenge was a worry that patients would choose MAID before considering hospice care, foregoing the benefits of this care and a good death.

Second, participants worried that allowing MAID on-site could change the role of a hospice to a ‘drop-in’ centre for MAID. Site 2 participants were concerned that if MAID was allowed on-site, hospitals could simply pass off patients who had requested MAID to hospice staff. Participants at site 1 felt that their hospice was already beginning to be used by hospitals for such a purpose. Third, participants at both sites were also concerned that allowing MAID on-site could negatively impact community perceptions of the hospice and its staff. They were concerned that the lay community already had misconceptions about hospices expediting the death of patients; they worried that allowing MAID on-site could exacerbate these concerns. A site 2 participant worried that allowing MAID on-site:

…could send mixed messages that 'oh as soon as you step inside the hospice, the palliative care, well, in a week from now they will give you the shot and you’ll be gone,’ you know. So, this is not the image obviously we want to, to be associated with the palliative care. It’s not the place, it’s not the espace, you know, where you euthanize a dog that is sick or something.

Struggling to reconciling a patient’s interest in MAID with staff beliefs in hospice care

Participants believed that hospice care was able to provide a comfortable and relatively pain-free end-of-life experience for most patients, and that the good death facilitated by hospice care was meaningful for patients and families. Amid these background beliefs, some site 1 participants struggled to reconcile a patient’s interest in MAID with their own professional beliefs. This struggle was a challenge for participants, with some experiencing emotional difficulties and frustration as a result. Participants held varying understandings about patients’ reasons for wanting MAID, including a concern that the patient was not satisfied with the hospice care they had received; thus, there had been a ‘failure of care.’ Over time, however, participants noted a transition in staff views, seeing MAID requests as less about a failure to provide satisfactory care and more as an expression of the patient’s end-of-life preferences.

MAID creates challenging clinical situations involving patients and families

Participants at both sites described how the availability of MAID for hospice patients, either on or off-site, had resulted in challenging clinical situations involving hospice patients and their families. These situations could be emotionally difficult for staff. Contrary to the previous themes, many of these situations were not inherent to the MAID treatment; rather, they were the result of challenges and irregularities in implementing the MAID laws and local MAID policies.

Participants described cases in which staff worried that patients who had already received MAID were actually unsure that they had really wanted it, or did not actually have the capacity to consent to the procedure. In other situations, patients were seen to refuse pain medications to satisfy eligibility criteria for MAID. Finally, one participant described hearing of clinical complications during the MAID procedure. Participants worried about and described cases in which patients’ decisions to receive MAID could cause conflict within families, as well as stress and anxiety for individual family members. However, participants also noted scenarios in which a MAID death brought ‘peace’ to a patient’s family as it was what the patient wanted.

Different hospice responses to MAID lead to unique challenges

Participants’ descriptions demonstrated how the different approaches to MAID taken by each hospice led to unique challenges. At site 1, MAID was allowed on-site, although it was performed by an external team from the provincial health authority. Participants described that this situation resulted in disruptive care. For example, while the MAID team performed the procedure, they left the patient’s body for hospice staff. Further, the separation of professional teams and tasks led participants to experience the MAID provision as a subversive and secretive activity in their hospice. Further, despite the separation between the MAID and hospice teams, hospice staff were requested to help the MAID team with MAID assessments; yet, the separation prevented them from attending the MAID procedure.

At site 2, patients who requested MAID were transferred off-site to receive the procedure. Participants believed that this transfer system negatively impacted continuity of patient care and felt that transfers were emotionally difficult on staff who had developed a bond with the patient. Further, they felt that transfers prevented patients from receiving MAID in a ‘nice’ hospice setting. One participant said:

The only reason that I would like to see [MAID] done here, is because I think, I don't like the fact that we transfer them to another facility, who knows how they're being taken care of, and if they stay here, we can still give them really good care, physical care, psychological support, before.

MAID in hospice provides opportunities to provide more extensive end-of-life options to patients and families

Despite the challenges many participants felt the legalisation of MAID created for their hospice, some participants viewed MAID as an opportunity to provide patients with more extensive end-of-life options within a hospice. They felt that offering MAID on-site would allow patients to choose their preferred death and dying experience, would provide another option for symptom control, and would allow patients to receive MAID in a ‘nice’ hospice setting as opposed to a hospital. A site 2 nurse noted:

It would probably be a relief for some patients to know that that’s an option, and that they wouldn't have to leave. ‘Cause right now they have to leave to go to the hospital, they don't get to stay with us, which is sad. You know to have to go somewhere foreign like that, and I mean, the hospital is not really where you'd want to go to die, but it would be nice, it would be really nice if they could stay with us. And like I said I'm sure it would be a big relief, or just like, just to know that it’s an option …

That said, participants still did not see MAID as a part of hospice care. Rather, they saw it as an alternative treatment that could co-exist with hospice care in a hospice facility. This separation helped participants avoid the conflict they felt MAID had with the commitments of hospice care and the death and dying experience this care facilitated.


Our results provide insight into the challenges and opportunities MAID brings to Canadian hospices. Like Freeman et al, our findings highlight that the introduction of MAID has left many hospice staff struggling to reconcile patient requests for MAID with their own beliefs regarding hospice facilities and hospice care.10 Since its inception in 1967, the modern hospice movement has been rooted in providing a certain dying experience, in which the patient’s end-of-life needs are addressed through holistic comfort care, enabling the patient to find meaning as they face their upcoming death alongside their family and friends.18–20 To many in the hospice community, assisted dying offers a contrasting dying experience, with participants in our study commenting on how the death offered by MAID is fundamentally incompatible with that offered by hospice, as they felt it does not allow staff to focus on providing the patient with comfort and quality of life and death.7 21 22 Like Freeman et al, our participants also found assisted dying to be uncomfortably quick in contrast to the longer approach to death common in hospice.10 Similarly, our participants struggled to accept a patient’s request for MAID, sometimes feeling that the care they offered had failed the patient.10

While our participants were concerned that MAID was not congruent with the ideology of hospice care, previous national and international studies have shown that many patients who received assisted dying had previous contact with palliative care services.6 23 These findings suggest that while palliative care has much to offer, like all fields of healthcare it has limitations. Palliative care is not able to completely meet the end-of-life needs and preferences of all patients. Thus, for some patients at least, hospice care and assisted dying are not mutually exclusive; rather they can be sequential. Similarly, some of our participants clearly felt that there was benefit for patients to receive hospice care before receiving MAID. The framework of ‘reflective awareness’ may help hospice staff reflect on and balance their own beliefs regarding hospice care and assisted dying, allowing them to respond to MAID requests as patients decide to access a legal healthcare option in alignment with their preferences.24

Our results also extend previous findings that a hospice’s response to legalisation can create challenges for a hospice and its staff.8 10 Specifically, our results aligned with and elaborated those of Freeman et al that demonstrated how the separation of MAID and hospice care can lead to communication issues among staff, and a feeling among hospice staff of being disconnected from the patient’s care surrounding MAID.10 These feelings were most pronounced when a patient was transferred off-site to receive MAID, with staff in our study reporting distress with this occurrence.10 Participant descriptions of potential harms to patients created through the transfer system raises ethical questions regarding the duties of hospices to their patients when a MAID request is made. Discussions in the hospice literature on obligations to those who choose an assisted death focus on the principle of non-abandonment and may provide a starting point for scholars and hospice administrators examining these questions.7 25–27

While our study highlighted challenges that legalisation has brought to Canadian hospices and their staff, our results also identified opportunities that hospice workers feel MAID could bring to their hospices, making this the first Canadian study to do so. Participants felt that allowing MAID on site could allow hospices to offer patients a death that better aligned with their preferences, gave patients another option for symptom control, and allowed patients to receive MAID in a comfortable hospice setting. These findings resonate with Hurst and Mauron’s position that palliative care and assisted dying share common goals and values, and instead of being antagonistic towards one another, could instead act complementarily to give patients maximal end-of-life choices.2

As the Canadian hospice community continues to reconcile with the legalisation of MAID, a balancing of the challenges and opportunities MAID brings to a hospice will need to be considered. The prioritisation of these challenges and opportunities may determine whether Canadian hospice and assisted dying share the integrated approach seen in Belgium or the heterogeneous approach seen in the United States.5–7 Regardless of the approach, the growth of MAID demand in Canada likely means that hospices and their workers will continue to encounter patient requests for MAID.28 Hospice administrators should take steps to ensure supports are available to staff to manage challenges relating to communication between MAID and hospice teams, patient families and staff distress (should ethical questions surrounding a MAID provision arise.)8


As with all qualitative research, caution should be taken in transferring the results of this study to other contexts.29 The two hospice sites included in this study each exist in their own unique social contexts, and as such readers should evaluate similarities between their contexts and the contexts of the study sites—residential hospices in Alberta and Quebec Canada—when looking to transfer the study’s findings.29 All qualitative research will only provide a ‘limited perspective’ given the nature of this form of inquiry. In keeping, we did not aim to be representative of the entire county. Our design, rather, was meant to bring forward some of the issues that hospices are experiencing to foster dialogue on a topic that has not been given much research attention to date. We feel that our choice of site sampling provided one way to start this discussion.

Given that focus groups were the primary method of data generation in this study, it is possible that participants with minority or controversial beliefs were not comfortable sharing their perspectives. This limitation was partially addressed with the option of an individual interview. Indeed, three of four participants who opted for an individual interviews did cite concerns over social desirability of their views. Fortunately for us, the participants in focus groups and individual interviews were enthusiastic about discussing MAID. We feel that despite the issue of social desirability, we were able to capture a range of perspectives of the issue: the data generated did not uniformly match one perspective, opinion or experience.


The legalisation of MAID in Canada has created new challenges and opportunities for the Canadian hospice community. A balancing of these challenges and opportunities may provide a path for Canadian hospices to navigate their new reality. However, increased demand for MAID means that hospices are likely to continue to encounter requests for MAID and should enact supports to ensure staff are able to manage these challenges.

Data availability statement

Data are available on reasonable request. Reasonable requests for study data should be directed to the corresponding author, or to the Institutional Review Board at McGill University’s Faculty of Medicine (514-398-8302).

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by the Institutional Review Board at McGill University’s Faculty of Medicine. (study number A06-B36-19B).


We would like to thank the participants for interest, enthusiasm and participation in the study. A thank you to the Macdonald Lab for their feedback on this manuscript.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors This manuscript was developed from the MSc (Bioethics) thesis of JM, supervised by MEM. JM designed the design, generated the data, led the data analysis and completed the first draft of the manuscript. MEM supervised the study design and process, contributed both substantively and editorially to the manuscript. JM is the guarantor of this submission.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests JEM declares a competing interest as he has volunteered at one of the hospice sites.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.