Article body

introduction

Medical assistance in dying (MAiD) is now in its ninth year as a legal option in Canada. During its short life, it has gone through two stages: MAiD 1.0, as defined by Bill C-14, passed in June 2016, and then MAiD 2.0, as amended by Bill C-7, passed in March 2021. Since it is unlikely that further changes will happen anytime soon, it seems appropriate to take stock of how far we have come, and where we might go next.

MAiD 3.0 could incorporate any of three additional amendments: 1) extending eligibility to patients who present solely with a mental disorder, 2) extending eligibility to mature minors, and 3) allowing advance requests for patients with degenerative disorders like Alzheimer’s disease who fear losing capacity before reaching the point at which they would wish to request MAiD. All three of these measures were recommended first by the 2015 Provincial-Territorial Expert Advisory Group and then repeatedly thereafter by the Special Joint Parliamentary Committee assigned to this file (1-3). For the record, I support all three. However, they are all contentious and I’m not going to discuss any of them.

Instead, I want to examine a view that has been advanced by some commentators: not merely that MAiD in Canada should not go farther, but that it has already gone too far. The critics of this persuasion are wont to say such things as: the current regime is much too permissive, we are currently sliding down a slippery slope, or MAiD in Canada is out of control (4-8). I’m interested in why they think this way.

As a preliminary, I’m going to draw a very simple distinction between two different kinds of objection to MAiD, which I will call principled v pragmatic. Principled objections usually take the form of contending that MAiD is just inherently wrong, or at least ethically problematic, by virtue of intentionally causing death or violating the sanctity of life (9-12). The Catholic Church provides an extreme version of this sort of opposition. Pragmatic objections, on the other hand, raise no ethical issue with MAiD itself, but are critical of the legal regime established to administer it: its eligibility criteria are too broadly drawn, or its procedures are too lax, or its safeguards are insufficiently stringent. It is objections of this latter sort with which I am here concerned. For the sake of the argument, I will assume that those who have voiced pragmatic objections to the current MAiD regime in Canada have no principled issue with the procedure itself[1], though I will later have reason to revisit this assumption.

Critics of Canada’s current MAiD regime have raised a wide variety of objections. I am particularly interested in two lines of argument which, whatever their differences, share a common assumption, i.e., that among medical services in general, and end-of-life medical services in particular, there is something special about MAiD. I want to ask what that is.

The first line of argument starts with numbers — more particularly, the number of MAiD provisions in Canada. So let’s take a look at the numbers that have so exercised the critics. In 2022, there were 13,241 MAiD provisions across the country, constituting 4.1% of all deaths in that year (14). According to the critics, both that total and that percentage were too high (6). The 2022 total was 31.2% higher than the previous year, in line with a 31% average annual growth rate for MAiD provisions in the three-year period from 2019 (14). According to these critics, that growth rate was also too high (6).

When these numbers are claimed to be too high, the obvious question is: too high relative to what? As far as I am aware, there is no Platonic ideal for the correct number of annual MAiD provisions for a given population size. MAiD is not like, say, immigration, where the government can set an annual intake target and then work toward hitting it. Instead, the annual number of MAiD provisions will be determined by the intersection of demand (the number of patients qualifying for the service and requesting it) and supply (the number of practitioners providing it). Since these are the factors driving the annual numbers in Canada, we should expect those numbers to continue to rise. The current legal regime for delivering MAiD is enormously popular among Canadians. In a 2023 Ipsos poll, 84% of respondents said that they supported the Supreme Court’s 2015 decision in Carter v. Canada and 78% supported the Bill C-7 removal of the “reasonably foreseeable natural death” requirement (16). When a sample of Canadians was asked whether they would request MAiD if they were facing a long and painful death, just under 50% indicated that they either probably would or definitely would (17). That’s a lot of potential demand, which we should expect to increase as the population ages. (The cohort of Canadians 65 years of age and older grows by about 3.5% annually.) In fact, the demand is already increasing: the annual growth rate in MAiD provisions is driven almost entirely by the annual growth rate in MAiD requests (14). On the supply side, the number of MAiD providers has also grown, with an average annual increase of 18% since 2019.

How high might the numbers go? One very speculative calculation projected that MAiD deaths might top out eventually at just over 10% of all annual deaths in Canada (17). While that projection cannot be dismissed out of hand, it does seem a little on the high side. In 2023, the annual growth rate declined by half to 15.8%, suggesting that the curve may be starting to flatten. In that year, MAiD provisions accounted for 4.7% of all deaths (15).

One way to demonstrate that the numbers are already too high would be to compare them with jurisdictions abroad that have their own assisted dying regimes. Currently, the comparator of choice for critics is California (6). On the surface at least, the comparison is apt: two jurisdictions with roughly equal populations, similar access to high-quality health care, and no substantial differences in the leading causes of death or overall death rates. And the disparity in the numbers is undeniably stark: in 2022 the 853 assisted deaths in California constituted just 0.27% of all deaths. We might therefore conclude, with the critics, that the Canadian numbers are too high. Or we might conclude that the California numbers are too low. At least at this point, both conclusions are equally in play.

So what explains the disparity? One obvious possibility is the different eligibility criteria in the two regimes: at least since the passage of C-7, Canada’s criteria are more expansive than California’s. In California, as in all US states that have legalized assisted dying, a diagnosis of a terminal condition is required, with death expected within six months. However, the authors of a recent side-by-side study of the two jurisdictions found that this difference does little to account for the numbers gap (17). Instead, they identified three principal factors: 1) Canadians are much more aware of MAiD and its availability than are their Californian counterparts; 2) Canada has many more MAiD providers than California; and 3) Canadian health care authorities and institutions provide robust support to members of the public seeking to access MAiD, support that is lacking in California. More awareness, more providers, more support — none of these sounds like a bad thing. Since just as many Californians as Canadians indicate that they would (probably or definitely) seek to access MAiD, were they facing a long and painful death (17), the appropriate conclusion would seem to be that they are being underserved by their current assisted dying regime. It’s not that Canada is doing worse, but that California could be doing better.

However that may be, these differences between Canada and California make comparisons between the two regimes of questionable relevance. We get more reliable results when we look to jurisdictions with regimes more like ours in Canada. The nearest analogue is the Netherlands, where the 8,720 MAiD deaths in 2022 represented a 13.7% increase over the previous year and constituted 5.1% of all deaths during that year. Quite clearly, in this context Canada is not an outlier.

There is in any case something about this whole controversy over the numbers that I find puzzling. Suppose that we were talking, not about MAiD, but about some other medical service for which there is continuing, and growing, demand — such as joint replacements. In 2020-21, 110,585 hip and knee replacements were done in Canada (18). That total represented a decrease over the previous year, due to COVID-19, but, leaving that factor aside, the number of replacement procedures grew by an annual average of 10% during the previous two decades. Do we have any reason for thinking that either that total or that growth rate is too high? Should we not instead think it a good thing that increasing numbers of patients who need, and will benefit from, a joint replacement are getting one?

Or consider abortions. In 2022, 97,211 surgical abortions were performed across Canada, about 10,000 more than the previous year (19). Is either that total or the year-over-year increase too high? Abortion is equally legal in every part of the country, but access to it is notoriously uneven, with facilities much more restricted in some provinces than in others, and with poor and rural women disadvantaged by comparison with their more affluent urban sisters. If the number of abortions increases annually, either because barriers to access are removed or because new providers begin offering the service, shouldn’t we think that a good thing, just because more women who want and need an abortion are getting one?

If we regard an increasing number of joint replacements or abortions as a success, with supply having risen to meet demand, why should we think that an increasing number of MAiD provisions is a failure, or somehow a problem? If more awareness, more providers, and more support are good things for these other services, why are they a bad thing for MAiD? Why should we think differently about MAiD than we do about other medical procedures? What’s so special about MAiD?

I’ll come back to that question, but first I want to consider another line of argument that begins in a different place but leads us to the same destination. The focus of this one is not on numbers but on disability. MAiD 2.0, initiated in 2021 by Bill C-7, is defined principally by the two-track system it introduced, in which Track 1 is the pre-existing MAiD 1.0 (with minor modifications), while Track 2 imposes additional procedural safeguards for patients whose natural death is determined not to be reasonably foreseeable. To be clear, it is not the addition of Track 2 that is driving the growth in the MAiD numbers; in 2022, provisions in this category made up just 3.5% of the total (14). This change was widely reported as making MAiD available for the first time to persons with disabilities. But this was a misperception. Under the terms of the original Bill C-14, to qualify for MAiD a person had to have “a serious and incurable illness, disease or disability”. Persons with disabilities have therefore been eligible for MAiD ever since the inception of the program. The innovation of C-7 was to open up eligibility for those with a longer trajectory to their natural death.

As such, the bill was greeted with alarm by many disability rights advocates (20). Their concerns about the introduction of Track 2 were many and varied, but one theme tended to stand out. The critics make the point that Canadians with disabilities live in a society pervaded by ableism — the mindset that having a disability makes one less than fully human (8). As a result, they may seek an assisted death having internalized the ableist valuation that their lives are not worth living. As Anita Ho has put it: “...ableist ideology that treats a life with impairment as categorically worse off than one without impairment can impede some people’s capacity to form the self-trust and self-confidence that are essential to possessing and exercising autonomy...” (21, p.342). Under these oppressive circumstances, some have argued, persons with disabilities are unable to make a genuinely autonomous choice for MAiD, in which case it is wrong — because discriminatory and unjust — to offer them that option (8,22). It is worth noting that Ho does not herself endorse this conclusion (21). Nor is it so clear that autonomous decisions are impossible under conditions of oppression (23).

For various reasons, I find this line of argument odd. For one thing, from the outset the campaign to legalize assisted dying in Canada has been consistently led by persons with disabilities. Sue Rodriguez had amyotrophic lateral sclerosis (ALS). Austin Bastable, who advocated for legalization during the 1990s, was almost completely paralyzed by multiple sclerosis (MS). Steven Fletcher, the Conservative MP who in 2014 introduced two private Member’s bills to establish a legal regime for MAiD, had become paraplegic at the age of twenty-three as the result of an automobile accident. Gloria Taylor, one of the plaintiffs in Carter v. Canada, also suffered from ALS. Of the two plaintiffs whose challenge of the “reasonably foreseeable natural death” provision in Bill C-14 led to Bill C-7, Jean Truchon had had cerebral spastic paralysis from birth and later also developed spinal stenosis and necrosis of the spinal cord, while Nicole Gladu suffered from post-polio syndrome. Presumably none of these determined advocates shared the opinion that they were unable to make an autonomous choice for an assisted death.

That opinion also does not seem to be shared by most Canadians with disabilities. In a 2023 Ipsos poll, 83% of respondents who self-identified as having a disability supported the Carter decision. More to the point, 78% supported C-7’s removal of the “reasonably foreseeable natural death” requirement, exactly the same level of support found for respondents with no disability (15). So we seem to have somewhat of a disconnect between those disability activists who oppose MAiD, on the one hand, and the broader community of people with disabilities for whom they purport to speak, on the other.

However, what I find most perplexing about the ableism argument is its evident inconsistency. Anyone with a serious disability will frequently be called upon to make decisions about the course of their medical treatment, including very consequential ones where both benefit and risk are sizeable and uncertain. Sometimes those decisions will concern whether to continue treatment for their condition at all or, instead, opt for hospice or palliative care. Those who make the latter choice will have determined that their expected quality of life is no longer sufficient to justify continuing to prolong their life by every means possible. In the starkest cases they will be making a decision to die.

The woman known to the world only as Nancy B. had Guillain-Barré syndrome, an irreversible neurological disorder that left her completely immobilized and dependent on a ventilator. Her intellectual capacity was unaffected by her condition. In 1992, at the age of twenty-five, having been on respiratory support for two-and-a-half years, she requested its removal in order to escape the suffering caused by her immobility. With the ventilator, she could potentially have lived for a long time; without it, she would die quickly. While her decision was not opposed by her health care providers, or by the hospital in which she was resident, as a matter of law it was unclear whether disconnecting her from the ventilator would constitute either criminal negligence causing death (a culpable homicide) or assisting a suicide. Justice Jacques Dufour, who heard the case in the Quebec Superior Court, affirmed Nancy B.’s legal right to refuse continued use of the ventilator, even though such refusal would certainly precipitate her death (24). Five weeks after the judgment, her attending physician induced Nancy B. into a coma and removed the ventilator; she died comfortably in her sleep.

No one suggests that persons with disabilities lack the capacity to make autonomous decisions concerning the management of their medical condition, up to and including the decision to forego further life-sustaining treatment. Any such suggestion would be rightly dismissed as offensively paternalistic. Only in the case of MAiD is that capacity put in question. So I return to my earlier question: What’s so special about MAiD?

The answer to this question may seem obvious: alone among medical interventions, MAiD both causes death and is intended to do so. That makes it special enough to require a statutory exemption from the general legal prohibition of consensual homicide and assisting a suicide.

I believe, however, that this answer is seriously misleading; MAiD is not as special as it is commonly made out to be. To see why, we need to situate it within the broader context of end-of-life care. The question for each one of us is not whether we are going to die: that much is a certainty. The question, instead, is when and how. Many people are content to let nature determine the timing and manner of their death. But some, especially those with a serious and incurable medical condition, wish to have some say in the matter, especially if there is a substantial likelihood that their natural end will be marked by serious suffering. Even before MAiD was legalized, those who were so disposed had available to them a number of ways of managing their exit from the world. If their condition required life-sustaining treatment — whether this took the form of technological support or surgical intervention or continuing medication — they could refuse further treatment and so hasten their death. If it did not require such treatment, they still had the option of seeking death by refusing food and water. Plus, of course, the time-honoured method of death by overdose of pharmaceuticals.

All of these methods were perfectly legal. The advent of legal MAiD provided a new way of hastening death, with considerable advantages over the previously available methods. Unlike refusal of treatment, it does not require dependence on life-sustaining treatment. Unlike refusal of food and water, it is quick and easy and can be arranged for a specified time so as to be able to share one’s last moments with family and friends. Unlike overdosing, it is done under medical supervision and therefore guaranteed not to fail and leave the would-be suicide severely brain damaged. The fact that MAiD enables this kind of effective control over the timing and manner of one’s death at least partly accounts for its enormous popularity.

Seen from this perspective, MAiD is not so special (except in a good way). It is not even unique in both causing death and being intended to do so: refusing food and water and overdosing on medication share both of these features. We could argue whether refusal of life-sustaining treatment causes death (I think it does), but it is clear that it will usually have the effect of hastening it. Whether it is intended to have this effect will vary from case to case; some patients may be very happy to find themselves living on without the treatment in question. But there are clear cases in which treatment is refused precisely in order to hasten death and avoid further suffering. Nancy B.’s case was one of them.

I have contended on a number of occasions that, all relevant factors being equal, there is no significant ethical difference between MAiD and other end-of-life treatment options that also have the effect of hastening death, such as cessation of life-sustaining treatment (25). One of those occasions was my affidavit in 2011 on behalf of the plaintiffs in Carter v. Canada. Similar testimony was submitted at that time by other expert witnesses. In her judgment, Madame Justice Lynn Smith reviewed the ethical evidence submitted by both sides and concluded as follows: “The preponderance of the evidence from ethicists is that there is no ethical distinction between physician-assisted death and other end-of-life practices whose outcome is highly likely to be death. I find the arguments put forward by those ethicists ... to be persuasive.” (26, para 335)

Not everyone will agree with me, or with Justice Smith, on this contentious ethical issue. What is important is to note the ground on which we are now contending. In terms of the distinction I drew at the outset, we are here dealing with a principled issue about MAiD, not a pragmatic one. The point of contention is whether there is anything inherently problematic with it, or whether, ethically speaking, it is on a par with other end-of-life practices that can also hasten death. I earlier reported that critics of MAiD on pragmatic grounds seem not to, and sometimes explicitly claim not to, have any principled objection to it. I am now coming to doubt this.

Of all end-of-life treatment options, MAiD is by far the most closely scrutinized and most stringently regulated. I earlier mentioned the figure of 13,241 MAiD deaths for 2022. We have that figure because MAiD provisions are required to be reported to provincial and territorial authorities and the resulting data are published annually by Health Canada. Responsibility for monitoring MAiD cases, so to ensure compliance with the statutory requirements, also falls to the provinces and territories, each of which has set up a review process just for this purpose. There is no similar system of reporting and review of patient deaths that have been hastened by cessation, at the patient’s request, of further life-sustaining treatment, despite the fact that the annual number of such deaths is vastly higher (many more patients with grievous and irremediable medical conditions choose hospice/palliative care rather than MAiD).

I cannot think of a single concern about MAiD and patient autonomy that does not apply equally to treatment refusal. But I also do not know of a single MAiD critic who raises similar concerns in this latter context. They all treat MAiD as special and in need of a higher level of scrutiny. I can think of no reason for this exclusive attention to MAiD except that they find the procedure itself to be peculiarly problematic and therefore in need of some kind of special justification. The principled critics of Canada’s MAiD policy are open about this: they think that MAiD is inherently wrong, thus that the ideal number of annual MAiD provisions is zero. The pragmatic critics do not raise this issue, but I find it impossible to explain their concerns without attributing something like this view (maybe a weaker form of it) to them as well.

In this respect, the debate over abortion once again provides an illuminating parallel. Critics who argue that abortions are too easily accessed, or that either the annual total of abortions or the year-over-year rate of increase in these numbers is too high, or that in a society pervaded by sexism women are not capable of making autonomous decisions about their own reproduction, are really objecting to abortion itself on principled grounds. They believe that it is inherently wrong, and thus that the annual number of abortions should be zero, or close to it.

So, until a better explanation of what the critics find to be so special about MAiD comes along, my hypothesis will continue to be that their objection to it is, at bottom, principled rather than pragmatic. If I am right about this, then the ultimate issue for them is not the annual number of MAiD provisions or the annual growth rate of these numbers, and not the conditions of autonomous choice for persons with disabilities, but the ethical status of MAiD itself. If that is indeed the field on which the battle is to be fought, then I’m pretty confident I know who will be the winner.