Article body

introduction

Medical assistance in dying (MAiD) has been legal in Canada since 2016. Since then, MAiD legislation and policy have evolved through a series of court challenges and amendments to the legislation. Initially, MAiD was only accessible to patients who had a grievous and irremediable medical condition for which natural death was reasonably foreseeable (1). After a court challenge (Truchon v Canada) and subsequent legislative changes, MAiD eligibility criteria expanded to include those whose deaths were not “reasonably foreseeable” (2). This expansion raised concerns about its implications for vulnerable populations. News media stories emerged that revealed individuals being offered or accessing MAiD due to a lack of other supports that may have otherwise ameliorated their suffering (3-4). Many of these cases have policymakers and ethicists re-assessing whether existing safeguards are sufficient to protect vulnerable Canadians.

Interestingly, there has been disagreement about who should be considered “vulnerable” — that is, susceptible to harm or exploitation. For instance, some scholars, patients, and disability rights advocates have argued that individuals with disabilities are particularly vulnerable under MAiD legislation (5-10), while other individuals with disabilities have argued that automatic assumptions of their vulnerability are themselves discriminatory (11-12). Part of the issue surrounding this tension may stem from a characterization of the concept of vulnerability that is limited and applied broadly to subpopulations. Feminist bioethicist Florencia Luna has recently challenged this traditional understanding of vulnerability and offered a more relational and contextual model that involves overlapping layers of vulnerability with cascading effects (13-14). This conception has practical implications for both public policy and societal attempts to address vulnerability. Elizabeth Victor and colleagues have taken Luna’s more nuanced view of vulnerability and applied it in three health and social care contexts (15). They show how it can provide a valuable lens to a) analyze cases of vulnerability within their broader contexts for a more nuanced understanding and b) guide practical actions in related healthcare and policymaking contexts.

In this paper, we build upon this work and apply Luna’s model specifically to emerging cases of MAiD in Canada. In what follows, we argue that Luna’s conception of vulnerability: 1) better explains some of the disagreements over MAiD (e.g., within the disability rights community); 2) accounts for the social determinants of vulnerability of MAiD patients, thereby providing opportunity for policy and practice changes that better address patient needs; and 3) can offer insight and guidance to governments so that they might better address vulnerability in society. When applied to MAiD, we contend that Luna’s approach to vulnerability will enable us to better identify patient unmet needs that may in fact be remediable and thus potentially avoid unnecessary MAiD deaths.

“Protecting the vulnerable” in MAiD policy and discourse

In the preamble to Bill C-14, the Federal Government stated that, “permitting access to medical assistance in dying for competent adults whose deaths are reasonably foreseeable strikes the most appropriate balance between the autonomy of persons who seek medical assistance in dying, on one hand, and the interests of vulnerable persons in need of protection and those of society, on the other” (1). However, in 2019, the requirement that one’s death had to be reasonably foreseeable in order to qualify for MAiD was challenged in a Québec court (16). Nicole Gladu, who lived with post-polio syndrome, was one of two Québec residents legally challenging Bill C-14 as being discriminatory and unconstitutional since it denied access to disabled individuals whose natural deaths were not reasonably foreseeable. Gladu rejected the idea that this requirement protected the vulnerable, vehemently denied that she was a vulnerable person in need of protection under the law, and asserted that “vulnerability is a concept used ad nauseum by paternalistic people in good health for standing in the way of MAiD” (11).

The Superior Court of Québec sided with Gladu and declared the “reasonable foreseeability of natural death” eligibility criterion in Bill C-14 unconstitutional (16). In 2021, the Federal Government responded to the ruling with the introduction of Bill C-7, which legally permitted MAiD for individuals whose natural death is not reasonably foreseeable (now referred to as Track 2 cases) and allowed for a waiver of final consent in certain circumstances (ss. 1 (7)). The passage of Bill C-7 called into question the balance between autonomy and vulnerability that was allegedly struck with Bill C-14. Some individuals have argued that autonomy has received disproportionate prioritization (7), and some MAiD providers now object to Track 2 cases because they feel the expanded eligibility goes too far (17). Additionally, since Bill C-7, there has been a notable increase in MAiD use. In 2024, MAiD was provided to 16,499 individuals, with track 2 cases more than doubling (18), and is now tied with cerebrovascular diseases as the fifth leading cause of death in Canada (19).

The enactment of Bill C-7 intensified debates about whether MAiD eligibility is empowering or endangering for vulnerable groups. These debates have included discussion about whether advanced directives ought to be permitted for individuals who may lose cognitive capacity to consent to MAiD (20-21), whether the law should extend to mature minors (22-23), and whether those whose sole underlying condition is a mental illness could ethically qualify for MAiD (24). There has also been notable public news media about MAiD as a potential threat to the lives of marginalized and vulnerable individuals, including for instance, Indigenous peoples (25), those with disabilities (5-10), prison inmates (26-27), those living in poverty (10,28-30), geographically isolated patients in rural regions (31), or those with mental illnesses (30,32-33). Bill C-7 was also met with particularly strong opposition from disability rights advocates (34-37) who feared that it would lead to further discrimination against people with disabilities and promote a narrative that the lives of disabled people are not of equal worth (6,38). Professor Emeritus and activist Catherine Frazee has argued that the legislation leads to a devaluation of disabled lives and makes them more vulnerable to social and healthcare discrimination (7).

Related discussions have not only addressed the vulnerability of particular groups or subpopulations, but also touched upon the broader conditions in which people are made vulnerable. For instance, some journalists and activists have noted the potentially coercive influence of poverty, homelessness, racialization, and limited social support services that may constrain people’s choices about end-of-life care (7-8,30). Similarly, some scholars have raised concern that individuals may be unduly vulnerable to pressures to access MAiD due to caregiver burnout, financial considerations, or despair, and that there is no oversight to prevent such situations (39). These ethical issues were mentioned in 2019 by Prime Minster Justin Trudeau:

[The Prime Minister] promised to ensure the proper supports are in place for palliative care and people with disabilities, so that no one chooses to end their life because they can’t get the help they need. “The essential element around society ensuring that everyone gets the supports, the treatment they need to live in dignity, and to make the choice of medical assistance in dying one that is made in a way that isn’t because you’re not getting the supports and cares that you actually need,” he said (40).

Since that time, a number of cases have been made public that reveal these remarks to be more aspirational than operational. For instance, news reports have illustrated stories of people being offered or pursuing MAiD when they primarily needed assistance to live. This includes a veteran who required a wheelchair ramp (41), individuals who were facing economic and food insecurity due to illness or a lack of adequate housing (42-43), and patients who needed additional home care services (44-45). Importantly, encouraging patients to pursue MAiD may be a violation of the Canadian Criminal Code. The provision of information about MAiD is allowable under s. 241(5.1) and helps patients to make informed decisions about their end-of-life care. However, under s. 241(1), it is illegal to “counsel” someone to end their life, i.e., to provide information with the aim of inducing, persuading, or convincing a person to die (47).

There have also been several MAiD cases reported in which the person requesting MAiD did not yet want to die but found MAiD to be more accessible than other basic supports or assistance that might have alleviated their suffering (48). These cases include persons living in poverty unable to access sufficient income supports (10), those with chronic and/or debilitating diseases struggling to access healthcare (49-50), and those with disabilities being denied appropriate care while in hospital (51). There have also been cases in which psychiatric care or assessment is underused by patients whose MAiD requests may be influenced by mental disorders. This has included a man in his late 40s who the Coroner’s Report indicated was suffering from post-COVID-19 vaccination syndrome (52), a 27-year-old woman with autism and other possible undiagnosed mental illnesses with no known eligible physical disability (46,48), and a woman suffering from post-concussion syndrome and related mental health issues, unable to access medical care during the COVID-19 pandemic (53). Notably, the British Columbia Civil Liberties Association, a group that initially led the charge for the 2015 decriminalization of physician-assisted death in Canada, has revised their prior position in light of such cases. The association recently warned about ethical abuses in the MAiD system, and called upon the Federal Government to act (54-55). These arguments showcase the rapidly growing concerns about providing MAiD for individuals who live in a state that provides insufficient social welfare to ensure the requisite conditions for a decent life (3,56-57). This view suggests that MAiD may not be a fully autonomous choice in the context of an inequitable society in which social, political, legal, medical, and economic structures make people vulnerable.

Despite these concerns, MAiD legislative expansion continues. During the passage of Bill C-7, the Federal Government initially stipulated a two-year end date on the prohibition of MAiD for those whose sole underlying condition is a mental disorder (s. 6). However, this has since been postponed until March 17, 2027. This delay was due to a lack of preparedness, as well as mounting ethical concerns identified by members of the mental health community. These included the added complexity of determining a mental disorder to be irremediable (58); difficulties in distinguishing MAiD requests from suicidality (59); questions of whether higher standards of capacity might be necessary for this population (60); determining the role of psychiatrists given their specialized training in understanding issues relevant to marginalized or vulnerable populations such as seniors, those suffering from depression, or people experiencing homelessness (58); and the need for improved access to mental health supports (61).

The underlying theme in these ethically dubious MAiD cases is the concept of vulnerability. These cases suggest that some MAiD recipients suffering from physical and mental illnesses also had intersecting suffering resulting from poverty and/or lack of adequate supports (56). Data from Health Canada also reveal that 47.1% of non-terminally ill Canadians who applied for MAID reported “isolation or loneliness” as one of the causes of their suffering (54). Social, economic, or untreated medical conditions that cause suffering may render individuals unduly vulnerable and affect the voluntariness of their requests for MAiD. Despite receiving considerable attention through recent news coverage and public debate, the concept of vulnerability itself has remained vague and ambiguous. It thus warrants further philosophical exploration if it is to have practical import in ethical policy-making decisions surrounding MAiD.

Vulnerability: From labelLed subpopulations to overlapping layers

Traditionally, in medicine, vulnerability has been viewed as a characteristic belonging to a group or subpopulation (e.g., children or pregnant women) that would lead to them needing special protections. This was notably seen in the iconic Belmont Report in 1979. One of the foundational texts examining the ethics of medical research, the report identified ethical principles that should inform the conduct and practice of research involving human subjects. It was also one of the first documents of its type to acknowledge vulnerability.

One special instance of injustice results from the involvement of vulnerable subjects in research. Certain groups, such as racial minorities, the economically disadvantaged, the very sick, and the institutionalized may be systematically sought as research subjects, owing to their ready availability in settings where research is conducted. Given their dependent status and their frequently compromised capacity for free consent, they should be protected against the danger of being involved in research solely for administrative convenience, or because they are easy to manipulate as a result of their illness or socioeconomic condition (62). This wording treats vulnerability in the traditional sense, as a characteristic of certain subpopulations. Such a view considers certain marginalized groups to be inherently and necessarily vulnerable (15). While the Belmont Report represented a paradigmatic shift towards accounting for and addressing the potential vulnerability of research subjects, this traditional approach possesses conceptual and practical limitations.

First, this account of vulnerability assumes that everyone in a group is necessarily vulnerable simply by virtue of being a part of that subpopulation. While the report does not suggest a quantitative degree of vulnerability, it’s use of the subpopulation framing of vulnerability provides the same protections to different individuals within a specific vulnerable population without distinction. This framing contributes to the stereotyping of whole categories of individuals and labels them all as “vulnerable” with no acknowledgement of potential differences between persons within that group (63) — it neglects the fact that individuals within a group may experience vulnerabilities differently (13). This lack of nuance can lead to the wrongful exclusion or wrongful inclusion of persons within the “vulnerable” category, i.e., because some individuals belonging to a vulnerable subpopulation might not actually be vulnerable. Conversely, others not in a vulnerable subpopulation might still require special protections. While the subpopulation model does not necessarily preclude others outside a vulnerable category from receiving special protections, it does not provide for them either, instead promoting the same protections for all individuals within a particular vulnerable group. If other legislation or guidelines similarly use the subpopulation approach, then indeed some persons falling outside of those predefined “vulnerable” groups may not receive the protections they need.

The rigidity of the traditional approach is also practically problematic as it ignores contextual factors and many of the potential intersecting sources that could make a person more or less vulnerable. For example, this could include socioeconomic factors, education, housing and/or income security, social circumstances, geographic location, among others. As other scholars have similarly noted, labeling people, instead of identifying conditions that trigger vulnerability, risks concealing conditions that may need to be addressed (15). If one of the goals of governments in acknowledging vulnerability is to be able to provide protections that address vulnerability, then an account that overlooks potential underlying sources would be a glaring oversight.

Florencia Luna attempts to solve the problem with the traditional approach by offering a model of vulnerability based on layers and not labels (64). In her account:

  1. No single standard or ideal exists and there are multiple factors or sources of vulnerability;

  2. they are deeply related to the context; and

  3. vulnerability is not an essential property of the research subjects or groups per se (13).

An approach that better identifies the conditions that trigger vulnerability will still have much overlap with the traditional subpopulation approach. Many of the same individuals will likely be identified as vulnerable by both accounts. However, there is a key conceptual departure from the traditional subpopulation approach. Some factors that lead to vulnerability may not clearly or neatly allow persons to fit into a traditional subpopulation group and will thus be better captured by this more nuanced approach. For instance, an individual might feel added pressure to consider MAiD as a result of their caregiver’s burnout. This factor may not in itself make a person part of a traditional vulnerability group, but does appear to be a layer of vulnerability with possibly profound effects. The practical upshot of this more nuanced model is that by identifying the particular conditions that trigger vulnerability, we are better able to meet the needs of vulnerable persons. Luna’s approach does not deem vulnerability to be inherent to the individual, but instead sees vulnerability as relational and contextual, as it acknowledges the environment in which an individual is situated. There are no fixed, predetermined, rigid categories or subpopulations. As she explains “we do not face a solid and unique vulnerability that exhausts the category” (13, p.89). Instead, there can be multiple layers of vulnerability (e.g., age, disability, socioeconomic conditions, housing, social circumstances, inability to consent), which can overlap and operate concurrently (14). Furthermore, such layers of vulnerability are not necessarily immutable, but may be acquired and/or removed. For instance:

The fact of being a woman does not in itself imply that one is vulnerable. A woman living in a country that does not recognize or is intolerant of reproductive rights acquires a layer of vulnerability (that a woman living in other countries that respect such rights does not necessarily have). In turn, an educated and resourceful woman in that same country can over-come some of the consequences of the intolerance of reproductive rights. Yet, a poor woman living in a country that is intolerant of reproductive rights acquires another layer of vulnerability. (She may not have access, for example, to emergency contraceptives and hence will be more susceptible to unwanted pregnancies) (13).

Additionally, Luna notes the potential for cascading effects of vulnerability. This refers to layers of vulnerability that may exacerbate already existing ones, or potentially generate new vulnerabilities (14). For instance, a lack of access to healthcare services can exacerbate an existing ailment or enable the proliferation of a new chronic disease that otherwise could have been detected and mitigated earlier. This worsening of health status may then threaten a person’s ability to work and undermine their source of income. Such a progression of events exemplifies the cascading potential of certain vulnerabilities.

Thus, this layered approach to vulnerability adds further context and nuance previously missing from the subpopulation categorizations. Further, this approach avoids viewing individuals only at a population level and instead focuses on the relevant factors and circumstances that may make a person vulnerable. The layered approach can thus “explain compounding vulnerability: how vulnerabilities interplay and render some persons more vulnerable than others similarly situated” (15, p.2). This better accounts for the differences and variations experienced by individuals who are members of the same group. Resisting the rigidity of the traditional approach, this layered model also has practical ethical consequences. Specifically, by acknowledging the diversity of potential underlying and intersecting sources that could make a person vulnerable, society is then in a better position to consider protections and safeguards that target and address those specific elements.

Luna’s relational and situational model contrasts with other scholars who have treated vulnerability as an intrinsic property of humanity. This essentialist or ontological view claims that human beings are fundamentally and essentially vulnerable (65,66), and that vulnerability is “universal and constant, inherent in the human condition” (67, p.1). However, there are conceptual and pragmatic issues with this essentialist approach to vulnerability. An overview of various accounts of vulnerability, and the challenges they may face, is beyond the scope of this paper. Suffice it to note that some argue that an ontological or essentialist view of vulnerability “loses its normative force if everyone is vulnerable” (14, p.103) and has limited ethical relevance in health policy. By that argument, if everyone is vulnerable, then no one is particularly vulnerable in a way that warrants special safeguards. Similar to the traditional approach, this model of vulnerability does not distinguish between individuals for the purposes of providing protection and addressing their vulnerability.

For the reasons outlined above, Luna’s layered approach appears best suited for navigating the nuances of vulnerability, particularly in a health policy context. While multiple bioethics scholars have embraced the layered model, some have challenged Luna’s account as lacking practical application and guidance (68). Practical application is vital if our theoretical understanding of vulnerability is to have pragmatic normative import in health and medicine. Luna has more recently attempted to satisfy this challenge and offers a method to better operationalize the layered account of vulnerability, which we discuss further below. Before that, however, we will show how Luna’s layers of vulnerability can better account for disagreements about MAiD within marginalized groups.

How layers can better account for disagreements about MAiD within marginalized groups

Some members of the disability rights community have been very concerned with expanded MAiD legislation. As noted above, Catherine Frazee expressed concern that Bill C-7 will lead to a devaluation of disabled lives, making them even more vulnerable to social and healthcare discrimination (7). Canadian legal scholar Trudo Lemmens has expressed concern that Bill C-7 prematurely enables the death of individuals and amounts to a “deadly form of discrimination” against those with disabilities or chronic illnesses (6). However, not all individuals within the disability rights community agree. As previously mentioned, Nicole Gladu, who suffered from post-polio syndrome, was one of the plaintiffs in the Québec case that led to Bill C-7. She argued vehemently that her disability did not make her vulnerable and that any alleged “safeguards” preventing her from accessing MAiD would be paternalistic and discriminatory (11,12). Gladu claimed that she never lacked anything needed for a full life. She went to university, became a journalist, worked in Paris and New York, including as a press attaché at the United Nations before retiring in Québec on the 14th floor of a condominium building, which as she explains, “offers not only all the necessities, but also a breathtaking view of the river that mirrors the beauty of the sunsets” (11). She described MAiD as an autonomous choice for herself, even envisioning the experience, “I want to be able to do [MAiD] effectively and without suffering, surrounded by my incredible friends, a flute of pink champagne in one hand and a canapé of foie gras in the other, as I watch the sun set over the river one last time from my living room window” (12).

Using the traditional subpopulation approach to vulnerability, it would be challenging to account for this intra-group division of opinion about disability. Disabled individuals are necessarily considered vulnerable if they meet the criteria to be a part of that group. However, with a relational and contextual approach to vulnerability, such as Luna’s layered model, this disagreement is better accounted for. It is the overlapping nature and interplay between layers of vulnerabilities that make some people more vulnerable than others. This model also acknowledges the role of socioeconomic status, social circumstances, education, and other factors that may create stark differences in vulnerability between individuals within the same group. This approach is congruent with intersectional feminist theory (69) that calls attention to the ways in which our identities are comprised of multiple, intersecting, social locations, such as race, class, gender, age, ability, or sexuality. In other words, none of us belong to a solitary subpopulation. This acknowledgement of diversity within marginalized groups reflects what some activists in the disability rights community have expressed. For instance, activist sb smith [sic] has argued:

Bill C-7 is currently being pushed into law in Canada, primarily due to the efforts of financially-privileged and/or newly disabled people, Dignity with Dying advocates, able-bodied lobbyists and government heads, and more. The progressives pushing for this bill to pass into law tout the arguments of the few financially-privileged and/or newly disabled people who agree that having the so-called choice to utilize no-holds-barred MAiD is, overall, a good thing for everyone. What these [MAiD advocates] don’t realize is that their class privilege clouds their perspectives on the way this legislation is and will continue to be used for harm with its money-saving potential to offload “burdens” (that is, disabled lives) from government-funded institutions like healthcare, housing, and so on… It is increasingly necessary for financially-privileged (and especially white, financially-privileged) disabled people to confront and be forthcoming about their wealth and social status. This is especially important given that the community is overwhelmingly comprised of us crips who (barely) survive on government benefits or are altogether ineligible for said benefits, have multiple side hustles, are under-paid, are completely unable to work, live in subsidized and/or group housing, are under-housed, are routinely rejected from medical insurance coverage, often go without medical treatments and prescriptions, are often denied medical care, and so on. Rich disabled people do not know what it means to be viewed as financial “burdens” on the institution of the Canadian nation state (70).

Such sentiments further exemplify how Luna’s layered model succeeds where the subpopulation approach to vulnerability failed. Equipped with this more nuanced conception of vulnerability, we can better understand the types of tensions raised by activist sb smith. This thinking can similarly be applied to other marginalized populations such as Indigenous peoples, prisoners, or individuals with mental illnesses, since members of these groups also experience diverse intersections of social locations and varying degrees of vulnerability.

Operationalizing A layered approach to vulnerability and applying it to MAiD

Luna proposes 2 steps towards operationalizing the layered model of vulnerability.

  1. Identify different layers or sources of vulnerability (e.g., those related to physical limitations, consent, dependency, exploitation, socioeconomic situations)

  2. Evaluate them both in terms of the extent of the harm (those with cascading effects may be more harmful) and the probability of its occurrence (13-14).

Luna states that layers of vulnerability are dispositional and “latent” until triggered by a stimulus condition. She uses the property of solubility to illustrate the concept. A cube of sugar has the dispositional property of solubility; it will dissolve in liquid. However, this property must be triggered and is only realized when the cube of sugar gets introduced into a liquid, i.e., the stimulus condition (14-15). Being able to identify and evaluate layers of vulnerability requires understanding their dispositional and latent nature, and anticipating triggers and stimulus conditions. This is crucial since it is the presence of the stimulus conditions that affects the likelihood that the disposition will manifest. “Researchers, healthcare professionals, ethics committees, and policy-makers should identify the stimulus conditions that can trigger the actualization of a layer of vulnerability” (15, p.3). Layers of vulnerability and their triggers or stimulus conditions may be different for individuals depending on their contexts and circumstances.

Consider the case of healthcare workers during the COVID-19 pandemic: this is not a subpopulation of individuals that would traditionally be considered a vulnerable group (14). Using the layered account though, we can see that as a result of their frontline work, they have a layer of vulnerability (step 1). They are vulnerable to falling ill themselves when tending to sick patients (e.g., during a pandemic). Inadequate testing, overstretched staff, and a lack of personal protective equipment would then be examples of stimulus conditions that could trigger the vulnerability. Assessing the presence of such triggers would then enable us to evaluate the vulnerability (step 2). Once layers or their stimulus conditions (i.e., triggers), are identified and evaluated, their inclusion in normative considerations would generate obligations. Specifically, mechanisms to avoid, minimize, or eradicate those vulnerabilities should be considered (14). Continuing our COVID-19 example, this might entail ensuring the procurement of sufficient protective equipment, making sick and personal time available, and/or training and hiring the appropriate amount of healthcare personnel.

Recently Elizabeth Victor and colleagues (15) applied this model of vulnerability to three health and social situations, generating meaningful insights. We see potential in applying this model to an analysis of MAiD. In many of the MAiD cases that are receiving ethical scrutiny, Luna’s model can assist in identifying layers of vulnerability among the patients requesting or receiving MAiD. This allows us to better identify and address unmet needs, potentially avoiding unnecessary deaths. Consider the following hypothetical case.

Richard is a 57-year-old Toronto man. He has a part-time job working as a mortgage broker and lives with his wife Ava in a recently updated bungalow in a nice suburban neighbourhood. He was diagnosed with multiple sclerosis (MS) in his 30s, but his symptoms have been effectively managed for years, and his quality of life is quite good. He is able-bodied, enjoys going for regular walks with his wife, and plays weekly pickleball with friends. He would not be considered vulnerable according to the subpopulation model of vulnerability since he does not fit into any predefined vulnerable group. However, Richard begins to experience a variety of odd and intermittent symptoms that include vision problems, feelings of weakness and fatigue, some issues with coordination, vertigo, and concentration problems. On some days Richard finds it difficult even to get out of bed. His family doctor retired several years ago and, having difficulty finding a new one, Richard has been relying on walk-in clinics when he becomes ill. He decides to visit the walk-in clinic near his home, and the doctor Richard sees initially attributes the new symptoms to aging. The doctor says that the symptoms are likely related to deteriorating eyesight and does not think there is anything more to worry about. Richard is then referred to an ophthalmologist; the earliest available appointment is two months away.

In the interim, Richard’s symptoms progress beyond his vison problems. His fatigue and weakness become chronic, he is no longer able to drive or go to work; he has to take a leave from his office job, which adds financial strain on his family because, like more than half of Canadians, they live paycheque to paycheque. Richard’s wife Ava has to take over all household chores as well as act as caregiver for Richard when his symptoms are particularly severe. Richard decides to visit the hospital, where he is then referred to a specialist who provides a requisition for an MRI, though due to healthcare backlog, it will take an additional month for Richard to receive the MRI. Five weeks later, Richard returns to the hospital for the MRI. A week after that, the results reveal that Richard has multiple new lesions in his brain and spinal cord. Further, some of his prior lesions appear to have grown. These are key indicators that Richard had been experiencing a MS relapse over the past year. Given that he has tried several popular MS treatments and still relapsed, Richard becomes concerned that he will continue to decline and never return to his “normal” life. Further exacerbating his concern is the fact that he and Ava are using up their savings, and that they will soon run out of money and struggle to purchase food or pay their rent. Fearing that he may become a physical and financial burden on his wife, Richard decides to inquire about MAiD.

Using Luna’s layered model, we can better analyze this case with respect to the multiple overlapping layers of vulnerability and their potential triggers.

The first layer of vulnerability is Richard’s difficulty in obtaining timely and appropriate medical assistance for his symptoms. Given the shortage of physicians and the overburdened medical system in Richard’s city, he does not have a family physician. Instead, Richard relies upon different walk-in-clinic doctors, which undermines continuity of care since these physicians do not know his detailed history with MS. Furthermore, Richard’s symptoms and experiences were originally dismissed, overlooked, or misunderstood by the clinic physician. MS symptoms are not always well understood or effectively treated within the healthcare system. Instead of receiving care and diagnostics, Richard’s symptoms were seen as “just part of aging.” There may be some level of healthcare ageism and ableism further triggering this vulnerability. Ableism is known to affect everyday living for persons with disabilities by leading to their experiences being dismissed and potentially preventing them from accessing needed assistance.

A second layer of vulnerability includes Richard’s precarious economic situation, leading to imminent housing and food insecurity. For Richard, MAiD appears to be largely a socio-economic consideration, and not something pursued out of a readiness to die. Yet, being unable to work leaves him without access to an income, and with his savings running out, he fears that he and Ava will be unable to survive. Further triggering this economic vulnerability is the inflationary economic situation in Richard’s city, leading to an increased cost of living. For example, the maximum monthly Ontario Disability Support Program (ODSP) payment for a single person is $1,408 (71) and the average monthly cost for an individual to live a healthy life in Toronto is over $5,000 (72). Without generational wealth to rely upon, Richard does not want his health to bankrupt his family or become a burden on the state. He begins to think that Ava would be better off without him.

A third level of vulnerability involves Richard’s reliance upon his spouse for care. Richard already relies upon his wife economically, domestically, and emotionally. He sees her burnt out and struggling to care for him. They have tried to obtain home care for Richard but find it challenging to schedule and it does not meet all their needs. He feels that he is a burden to her and that her health may also quickly deteriorate, a cascading effect that would leave him increasingly vulnerable. As a result, he has become embarrassed, distant, and socially isolated. Social isolation can be a significant contributor to the suffering that MAiD patients experience. As previously noted, almost half of non-terminally ill Canadians who applied for MAiD reported “isolation or loneliness” as one of the causes of their suffering (54).

Table 1

A layered analysis of Richard’s vulnerability

A layered analysis of Richard’s vulnerability

-> See the list of tables

When detailed in this manner, it is clear that although Richard was not initially a member of a subpopulation that is traditionally seen as vulnerable and in need of protection under the law, he has several layers of vulnerability that inform his decision to inquire about MAiD. It is thus important for MAiD assessors to 1) acknowledge the external and relational conditions that trigger Richard’s economic, healthcare, and emotional vulnerabilities, and 2) consider the ways in which these vulnerabilities might affect the voluntariness of his request for MAiD. A nuanced analysis of vulnerability suggests that while Richard’s illness may indeed be serious, incurable, and in decline, his degree of suffering may be triggered or mitigated by situational factors. It is also unhelpful to simply label Richard as “vulnerable” because of his MS without paying attention to relevant contextual details. By identifying the layers of Richard’s vulnerability, what triggers them, their probability, and the level of potential harm, healthcare professionals and others can more accurately and meaningfully assess and address his needs. It would be worthwhile to explore remedies and trigger-mitigation strategies for Richard, where possible, to see if they might alter his request for MAiD. While this case is hypothetical, it reflects real issues experienced by Canadians exploring MAiD.

Acknowledging social determinants of vulnerability and opportunities for practice and policy changes

As demonstrated above, the traditional subpopulation approach to vulnerability can result in the mistaken inclusion or exclusion of persons within the category “vulnerable.” By looking at groups as a whole, instead of individual patients and their unique circumstances, the subpopulation approach has, at times, led the health and public policy sectors to neglect factors that could make a person vulnerable. By contrast, Luna’s model of vulnerability enables better patient-centred care and analyses as it allows for deeper consideration of the social determinants of vulnerability and health (e.g., socio-economic factors, poverty, housing insecurity, food insecurity, substance addictions, past traumas, coercive or abusive family environment, isolation, psychiatric disorders, and more). This nuance is paramount for MAiD, as recent cases demonstrate that people have applied for the service after having been made vulnerable due to unmet needs. Trudo Lemmens has raised concern over such a situation, arguing that in many cases MAiD is quicker and more easily accessible than certain medical or financial supports, including long-term care, specialized pain clinics, and the Canada Pension Plan Disability Benefits (6). As one palliative care physician noted, MAiD can be arranged in a more timely and organized process than can access to mental health services (30). Others similarly fear that MAiD could be chosen as “the antidote to a system that fails in compassion or equitable palliative care access” (73, para.17). It is ethically worrisome that a patient can more readily access MAiD than support through the social safety net. For instance, someone could access MAiD after the 90-day waiting period but have to wait over 120 days to access specialized long-term care, specialized pain clinics, or disability benefits (6). Such a scenario makes those patients unduly vulnerable to MAiD since it would become their only option for relief from otherwise preventable suffering.

This presents an opportunity to reconsider practice and policy in light of concerns about vulnerability. Policy and practice require sufficient safeguards that acknowledge and address the multiple ways in which patients might be vulnerable. Here we consider three such proposals: 1) An amendment to the Federal Government’s language about “protecting the vulnerable” in MAiD legislation; 2) Enhanced MAiD assessments that better screen for social determinants of vulnerability and use a multidisciplinary approach to address them; and 3) establishing an oversight body with members from multiple stakeholder groups with special expertise or experience relating to some of the most relevant sources of vulnerability. We detail each proposal in turn.

First, we recommend an amendment or update to the Federal Government’s language about “protecting the vulnerable” in MAiD legislation. Given the ongoing expansion of MAiD eligibility, it is time to formally revisit the balance between promoting autonomy and protecting the vulnerable, and the government’s role in doing both. When those in power speak about protecting “the vulnerable,” to whom are they referring and how is such vulnerability determined? Providing such information and identifying how current safeguards protect “the vulnerable” on Government of Canada websites and in official communications about MAiD could help to improve transparency and accountability.

Second, we acknowledge that some guidance to healthcare professionals has stipulated that they should seek to understand a patient’s circumstances, perspective, and reason for considering MAiD (74), and identify if patients are seeking MAiD due to unmet needs (75). However, policies have failed to offer more concrete guidance on what healthcare providers are to do in such circumstances (76), and how such unmet needs would be addressed within clinical contexts. Nor have there been any changes in practice that would ameliorate the concern that such unmet needs are making patients unduly vulnerable to MAiD. If MAiD is being requested due to such reasons, it is unclear how such requests could be seen as meeting the condition of grievous and irremediable suffering.

The assessment process for MAiD ought to include a more robust analysis of the many triggers to vulnerability and unmet needs that led to the patient’s request. Policymakers and regulatory bodies should provide more guidance for assessors regarding social determinants. This would include how to explicitly look for potential layers of vulnerability that might otherwise be overlooked, and to subsequently refer patients to multi-disciplinary specialists as needed (e.g., social workers that can assist with housing, disability benefits, food programs, and more).

In particular, the assessment process must better screen for underlying mental health conditions that may be leading to the MAiD request. Current legislation requires the MAiD recipient to have grievous and irremediable suffering caused by illness, disease, or disability that is not solely caused by mental illness. However, as many MAiD cases reveal, patients often have a combination of physical and psychological conditions. Despite the presence of physical disease or illness, the additional accompaniment of psychological conditions may signal a potential vulnerability and/or unmet need that prompted the MAiD request. It is vital to ascertain whether pre-existing mental health conditions, or unmet needs triggering mental health conditions, may be significantly contributing to the suffering leading to the request for MAiD. This may be of particular import in situations where the patient has a psychiatric history of suicidal ideation. In line with the Ontario Coroner’s MAiD Death Review Committee, we therefore recommend MAiD practitioners consider using “psychiatric assessment when a person is requesting MAiD with a complex medical condition and concurrent mental illness… [and] referrals to other mental health professionals (including psychologists, social workers, mental health nurses and psychotherapists)” (52, p.18). This practice would better identify whether other psychological, social, or economic factors are creating undue vulnerability for the patient and influencing their desire for MAiD. Attention to the overlapping layers of vulnerability would then enable the provision of mitigating factors for their triggers. Such practice changes would better ensure that such patients could access support for their psychological, social, or socioeconomic issues that may be causing their underlying suffering.

Applying Luna’s approach in this way would involve greater effort to hear from those requesting MAiD. However, it has the potential to better capture what patients are telling us, and specifically their first-person narrative accounts of what it is like to live with certain mental disorders or in certain socio-economic situations. The gathering of such information will result in a more robust evidence base on why people are requesting MAiD, the factors precipitating their request, and services they may require to address any unmet needs. We also support further empirical research on how best to implement Luna’s approach and gather this type of evidence.

Third, some scholars, such as Lyon, Lemmens, and Kim (48), have argued for better multidisciplinary oversight through the establishment of a more transparent and safeguarded compliance regime founded on a variety of stakeholder perspectives.

The federal government might establish an independent and transparent public body more representative of most clinical specialties and other stakeholder groups reflected in MAID requests (e.g., disability, poverty, Indigenous, and palliative, cancer, and psychiatric care specialists), as well as a public meta-regulator to provide oversight and standardization for MAID (48, p.19).

This advisory and oversight body could review particularly ethically concerning MAiD cases, as well as the ongoing MAiD landscape with a particular eye to vulnerability, in order to advise future legislation and policy. This is in line with recommendation 16 from the Final Report of the Expert Panel on MAiD and Mental Illness. They suggest that “the federal government should play an active role in supporting the development of a model of prospective oversight for all or some Track 2 cases that could be adapted by provinces and territories” (77, p.16). Such oversight should also include enhanced data collection and transparency around aspects of MAiD practice. These data would include the rationales for MAiD requests, attempts made to address unmet needs when identified, physician rationales when refusing patient requests for MAiD or other services, and the number of MAiD deaths per provider. Our deeper understanding of vulnerability supports each of these proposals. The bolstering of the assessment process including the use of various health professionals, and the establishment of a diverse oversight body of stakeholders would create a MAiD regime better equipped to speak to and/or address the variety of existing and evolving triggers for vulnerability.

Conclusion

Equipped with a more nuanced approach to vulnerability, we are better able to acknowledge and identify various sources and triggers of vulnerability in our society. However, this identification generates normative obligations. We acknowledge that this will likely involve grappling with macro and micro resource-allocation issues that are beyond the scope of this paper. Similarly addressing the sources of vulnerability and people’s unmet needs will likely surpass the responsibilities of healthcare workers and medical institutions. Government bodies must therefore consider their priorities when citizens are being offered a dignified death but not the means to live. As Frazee notes, “it’s cold comfort, I think, to be offered the choice to die when you are not offered the choice to live a dignified life” (7, para.13). While it is understandable that in desperate circumstances people may opt for MAiD, several current cases suggest that this can occur as a result of structural coercion that makes persons vulnerable, thereby undermining meaningful autonomy (56). Importantly, we do not have robust or sufficiently reliable public data regarding the prevalence of requests that are influenced by unmet needs. By highlighting related cases that have received public attention, we do not intend to overstate the frequency of such requests (nor suggest that all requests will be approved and/or lead to provisions). Rather, we aim to draw attention to the fact that these cases exist, that the ethical stakes are high, and that with the ongoing expansion of MAiD eligibility, it is critical that we learn from feminist bioethicists and disability rights scholars who have, for years, identified the ways in which the law can fail to protect vulnerable individuals.

MAiD was initially created in response to the Carter v. Canada case in order to facilitate access for patients with reasonably foreseeable deaths or who were experiencing grievous and irremediable suffering. The expansion of MAiD, and our nuanced understanding of vulnerability, call us to reflect upon the myriad sources of human suffering in our current social, political, and economic context. Parallel to efforts to improve the MAiD system in Canada, we argue that the Government of Canada has a responsibility to also improve the social safety net. The suffering that results from poverty and inadequate housing is not irremediable and political inaction is not ethically neutral (78). It is incumbent upon the state that wishes to offer death as an option in healthcare, that they first ensure the necessary steps are taken to mitigate needless suffering. This would involve addressing the social, economic, or medical unmet needs that make citizens vulnerable, through policy initiatives such as publicly funded mental healthcare, increases to disability support benefits, and bolder affordable housing initiatives.

We acknowledge that our analysis is not without limitations. While the overlapping layered model of vulnerabilities has more explanatory power than the traditional approach, and has the potential to assist in establishing more ethical MAiD policy, we acknowledge that the exact standards for determining undue vulnerability are unclear (i.e., what would be the threshold for someone to be declared unduly vulnerable to MAiD?). As Diana Myers (79, p.204) has noted, the precise range of life-sustaining resources sufficient for persons to make autonomous choices in a health care system represents a threshold that is notoriously difficult to determine. Although in many cases excessive vulnerability will be obvious, there will also be borderline cases. This is especially true since we all have some degree of contextual factors that make us more or less vulnerable in certain circumstances. How many triggers or factors, and to what degree, would be required for someone to be deemed unduly vulnerable? For full disclosure, the authors of this article are not fully aligned on the answer. The lead author doubts that sufficient safeguards could exist that would prevent MAiD from posing harm to vulnerable persons, and that some level of prohibition may thus be required. The other two authors contend that outright prohibition would dismiss the agency of ill and/or disabled individuals to determine their own reasonable threshold for suffering, and so risks preventing ethical MAiD provisions. Collectively, we believe that Luna’s richer and more nuanced understanding of vulnerability has both inherent and instrumental value, and the potential to positively affect MAiD policy, practice, and both public and bioethics discourses.