• The number of Canadians dying prematurely by “medical assistance in dying” (MAiD) has risen thirteenfold since legalization. In 2016, the number of people dying in this way was 1,018. In 2022, the last year for which data are available, the number was 13,241.
• MAiD in Canada is the world’s fastest-growing assisted-dying program.
• MAiD is now tied with cerebrovascular diseases as the fifth leading cause of death in Canada. Only deaths from cancer, heart disease, COVID-19, and accidents exceed the number of deaths from MAiD.
• Assisted dying was not meant to become a routine way of dying. Court rulings stressed that it be a “stringently limited, carefully monitored system of exceptions.” Then Minister of Justice and Attorney General Jody Wilson-Raybould agreed: “We do not wish to promote premature death as a solution to all medical suffering.” The Canadian Medical Association likewise stated that MAiD was intended for rare situations.
• MAiD assessors and providers do not treat it as a last resort. The percentage of MAiD requests that are denied continues to decline (currently it is 3.5 percent). MAiD requests can be assessed and provided in a single day.
• Government departments and agencies continue to state that Canada’s MAiD experience is similar to that of other jurisdictions, that the rate of increase is expected, and that the growth is gradual. The data contradict these statements.
• Health Canada has dramatically underestimated what a “steady state” of MAiD deaths would look like and how quickly Canada would reach the 4 percent threshold of total deaths. This threshold was reached in 2022, eleven years ahead of what Health Canada predicted only months earlier, and double its prediction just four years earlier.
• Despite the importance of accurate vital statistics, some provinces’ death records do not record MAiD as a cause of death, instead recording the underlying condition that led to the MAiD request and subsequent death. Further, Health Canada reports on the number of MAiD deaths, but Statistics Canada does not consider MAiD a cause of death. These inconsistencies in reporting have an impact on research about MAiD and about causes of death more generally.
• The systematic underestimation of MAiD in government statements and reporting is a serious impediment to understanding the scale of MAiD’s normalization in Canada and its abnormality with regard to other countries where some form of assisted dying is permitted.
• For policymakers and the public to properly understand the Canadian reality, it is essential that government agencies collect consistent data and issue correct statements.

Contrary to repeated official statements by the Government of Canada, Canadian “medical assistance in dying” (MAiD) is the world’s fastest-growing assisted-dying program. In 2016, the first full year in which assisted dying was legal in Canada, 1,018 Canadians died by MAiD. The number has increased thirteenfold since then, and in 2022, the last year for which data are available, more than thirty-six MAiD deaths occurred per day, on average. 1 1 Health Canada, “Fourth Annual Report on Medical Assistance in Dying in Canada 2022,” Government of Canada, October 2023, https://www.canada.ca/en/health-canada/services/publications/health-system-services/annual-report-medical-assistance-dying-2022.html. MAiD is effectively tied with cerebrovascular diseases as the fifth most frequent cause of death in 2022. 2 2 Although Statistics Canada does not include MAiD as a “cause of death,” as discussed in more detail later in this report, the total number of MAiD deaths stated in Health Canada’s reporting can be compared with Statistics Canada’s data on causes of death. Health Canada, “Fourth Annual Report”; Statistics Canada, “Table 13-10-0394-01. Leading Causes of Death, Total Population, By Age Group,” November 27, 2023, https://doi.org/10.25318/1310039401-eng. Only deaths from cancer, heart disease, COVID-19, and accidents surpass the number of deaths from MAiD.

This increase dramatically exceeds the projections of Health Canada. In May of 2022, for instance, Health Canada predicted that by 2033, MAiD deaths would stabilize at 4 percent of total deaths, 3 3 Government of Canada, “Regulations Amending the Regulations for the Monitoring of Medical Assistance in Dying,” Canada Gazette, Part I 156, no. 21 (May 21, 2022): 2567, https://canadagazette.gc.ca/rp-pr/p1/2022/2022-05-21/pdf/g1-15621.pdf. a figure that was double its first public prediction of a “steady state” in MAiD barely four years prior. 4 4 Government of Canada, “Regulations for the Monitoring of Medical Assistance in Dying,” Canada Gazette, Part II 152, no. 16 (August 8, 2018): 3089, https://canadagazette.gc.ca/rp-pr/p2/2018/2018-08-08/pdf/g2-15216.pdf. Yet, Health Canada’s estimate was proven to be inaccurate even before the end of 2022. Health Canada’s fourth annual report on MAiD, released in October 2023, revealed that Canada surpassed the 4 percent mark in 2022—in effect, eleven years ahead of schedule.

The growing number of MAiD deaths, and the continued expansion of eligibility criteria, is far beyond the expectations set in Carter v. Canada, the court case that decriminalized assisted dying in 2015. Consequently, MAiD is now far more than exceptional: it is routine. Almost no MAiD requests are denied by clinicians, and the median time between written request and death from MAiD in 2022 was merely eleven days. Despite judges’ and policymakers’ claims or expectations, MAiD is no longer an option of “last resort.”

This report examines the data that are available about the number of Canadians who have died by MAiD each year since its legalization, the rapid increase in this number from year to year, and how the Canadian numbers compare to those of other jurisdictions where some form of assisted dying is legal. The report calls into question the accuracy of official statements, which continue to underestimate MAiD’s rate of growth in Canada, and it describes deficiencies within the current MAiD reporting system.

In jurisdictions around the world in which “assisted dying” is legal, the term may refer to euthanasia (in which a doctor or other medical professional ends a patient’s life with their consent, as for example by lethal injection), or to assisted suicide (in which a doctor or other medical professional assists a patient in ending their own life, as for example by providing the patient with a lethal substance to consume), or to both of these. In Canada, while both euthanasia and assisted suicide are legal, almost all cases of MAiD are euthanasia, provided by a doctor or nurse practitioner. 5 5 Health Canada, “Fourth Annual Report.” This report uses the terms “assisted dying” and “MAiD” interchangeably.

Unlike in most jurisdictions that have legalized assisted dying, the criminal prohibition in Canada on assisted death was overturned not through the political system but through the courts. In 2015, the Supreme Court of Canada in Carter v. Canada unanimously upheld a trial judge’s determination that the criminal prohibition on euthanasia and assisted suicide was unconstitutional. Yet the Carter decision, and the subsequent federal legislation in Bill C-14 6 6 Parliament of Canada, Bill C-14, An Act to Amend the Criminal Code and to Make Related Amendments to Other Acts (Medical Assistance in Dying), Statutes of Canada 2016, https://www.parl.ca/DocumentViewer/en/42-1/bill/C-14/royal-assent. that responded to the court ruling the following year, was not a carte blanche for assisted dying for all serious illnesses and disabilities. For both the courts and policymakers, the societal “consensus” that mandated the decriminalization of some form of assisted dying was also the same consensus for assisted dying to be limited—and used only for rare cases.

Assisted dying was clearly not meant by the courts to become a normal way of dying. The British Columbia trial judge in Carter (2012) stressed the need for a “stringently limited, carefully monitored system of exceptions”—a set of circumstances outside the norms of routine healthcare. 7 7 The Supreme Court at Carter v. Canada (Attorney General), 2015 SCC 5, [2015] 1 S.C.R. 331 at para 29 cited Smith’s reasoning in Carter v. Canada (Attorney General), 2012 BCSC 886 at para 1243: “Permission for physician-assisted death for grievously ill and irremediably suffering people who are competent, fully informed, non-ambivalent, and free from coercion or duress, with stringent and well-enforced safeguards, could achieve that objective in a real and substantial way.” The danger posed by an unfettered MAiD regime was recognized: it was because of “the consensus that unlimited physician assisted death would pose serious risks” that an assisted-dying system required strict limits—short of a complete ban—for it to be constitutional. 8 8 Carter v. Canada (Attorney General), 2012 BCSC 886 at para 342. Moreover, the Supreme Court further affirmed the trial judge’s determination that the same “strong consensus” in Canada was that “if it is ever ethical in an individual case for a physician to assist in death, it would be only in limited and exceptional circumstances, where it is clearly consistent with the patient’s wishes and best interests, and in order to relieve suffering.” 9 9 Carter v. Canada (Attorney General), 2012 BCSC 886, para 342; Carter v. Canada (Attorney General), 2015 SCC 5, [2015] 1 S.C.R. 33, para 24. It is important to recognize that a patient’s “wishes” can, and often do, clash with their “best interests,” even when a patient is not suicidal.

Assisted dying was also never intended by policymakers to be a panacea for all illnesses. In introducing Bill C-14, Minister of Justice and Attorney General Jody Wilson-Raybould agreed: “We do not wish to promote premature death as a solution to all medical suffering.” 10 10 Canada, House of Commons Debates, April 22, 2016, 42nd Parliament, 1st Session, vol. 148 (045), https://www.ourcommons.ca/DocumentViewer/en/42-1/house/sitting-45/hansard. The role of medical professionals is to be more than simply providers of a service. That a patient may legally qualify for assisted dying does not mean that they should receive it. Furthermore, “evidence, presented over the past year, confirms that medical assistance in dying may pose risks to the vulnerable,” Minister Wilson-Raybould said in Parliament, “even in circumstances where there is a general consensus that the person should be eligible for the procedure.” 11 11 Canada, House of Commons Debates, May 31, 2016, 42nd Parliament, 1st Session, vol. 148 (062), https://www.ourcommons.ca/DocumentViewer/en/42-1/house/sitting-62/hansard.

It was not just the courts and government that envisioned a limited assisted-dying program. Following the Supreme Court’s decision in Carter (2015), the Canadian Medical Association released a statement that “there are rare occasions where patients have such a degree of suffering, even with access to palliative and end-of-life care, that they request medical aid in dying. In such a case, and within legal constraints, medical aid in dying may be appropriate.” 12 12 Canadian Medical Association, “CMA Examining Supreme Court Ruling Striking Down Ban on Doctor-Assisted Death,” Longwoods, February 6, 2015, https://www.longwoods.com/newsdetail/4968, (emphasis added). This wording was similar to then Justice Minister Wilson-Raybould’s: even when MAiD may be legal, it is not necessarily clinically appropriate. The lead lawyer for the plaintiff in Carter (2015) likewise stressed that they were arguing for physician-assisted dying specifically, which would help ensure that legalized MAiD would be performed rarely, noting that

all doctors believe it is their professional and ethical duty to do no harm. Which means, in almost every case, that they will want to help their patients live, not die. It is for the very reason that we advocate only physician assisted dying and not any kind of assisted dying because we know physicians will be reluctant gatekeepers, and only agree to it as a last resort. 13 13 Supreme Court of Canada, “Webcast of the Hearing on 2014-10-15,” https://www.scc-csc.ca/case-dossier/info/webcastview-webdiffusionvue-eng.aspx?cas=35591&id=2014/2014-10-15--35591&date=2014-10-15&fp=n&audio=n.

Unlike the earlier expectations that MAiD requests would be granted only in rare circumstances, Canada is now far beyond the limitations expected or recommended. Just four years after the Supreme Court’s Carter decision, a Superior Court of Quebec judge in Truchon v. Canada ruled that Canada’s MAiD program was unconstitutional because it did not allow for the assisted dying of persons not at end-of-life. 14 14 Truchon c. Procureur général du Canada, 2019 QCCS 3792. Although dozens of disability-rights groups urged then Minister of Justice and Attorney General David Lametti to appeal the ruling, Minister Lametti chose not to do so. 15 15 Inclusion Canada, “Advocates Call for Disability-Rights Based Appeal of the Quebec Superior Court’s Decision in Truchon & Gladu,” October 4, 2019, https://inclusioncanada.ca/2019/10/04/advocates-call-for-disability-rights-based-appeal-of-the-quebec-superior-courts-decision-in-truchon-gladu/. Instead, to comply with the lower-court judge, the government introduced Bill C-7, 16 16 Parliament of Canada, Bill C-7, An Act to Amend the Criminal Code (Medical Assistance in Dying), Statutes of Canada 2021, https://www.parl.ca/DocumentViewer/en/43-2/bill/C-7/royal-assent. which came into effect in 2021.

Bill C-7 moved far beyond the judicially imposed ruling in Truchon, let alone the Supreme Court’s ruling in Carter. Existing safeguards were weakened. The minimum ten-day assessment period for MAiD was removed entirely, allowing for the same-day assessment and provision of assisted dying for almost all MAiD requests. The requirement that the patient give final consent before administration of MAiD was no longer mandatory. Moreover, MAiD became available to persons whose death is no longer “reasonably foreseeable” but who consider that their physical suffering, from a disability for example, is intolerable to them. In a surprise last-minute amendment to C-7 in the Senate, MAiD was expanded further, without any additional legal safeguards, to include mental illness as a qualifying condition. This clause was scheduled to take effect in March 2023 but has since been deferred to 2027. 17 17 Parliament of Canada, Bill C-39, An Act to Amend the Criminal Code (Medical Assistance in Dying), Statutes of Canada 2023, https://www.parl.ca/DocumentViewer/en/44-1/bill/C-39/royal-assent; I. Grant, “Legislated Ableism: Bill C-7 and the Rapid Expansion of MAiD in Canada” (SSRN Scholarly Paper, Rochester, NY, 2023), https://doi.org/10.2139/ssrn.4544454.

Despite these expansions, it was still the government’s stated intent that MAiD would remain rare. At a Dying with Dignity Canada presentation in 2021, Minister Lametti emphasized that MAiD is “a very serious decision that is only taken under the most tragic of circumstances and the most personal circumstances.” 18 18 Dying with Dignity Canada, “Bill C-7 Update with the Honourable David Lametti,” April 21, 2021, YouTube video, at 23:40, https://www.youtube.com/watch?v=eD71TvVKGZE&ab_channel=DyingWithDignityCanada. Minister Lametti stressed to the Toronto Star in 2022 that since MAiD was in effect a “species of suicide,” it applied only to “a group within the population who, for physical reasons and possibly mental reasons, can’t make that choice [to commit suicide] themselves to do it themselves. And ultimately, [MAiD] provides a more humane way for them to make a decision they otherwise could have made if they were able in some other way.” 19 19 A. Raj, “Is Canada Expanding Medical Assistance in Dying Too Quickly?” Toronto Star, November 18, 2022, in It’s Political, podcast, https://www.thestar.com/podcasts/its-political/is-canada-expanding-medical-assistance-in-dying-too-quickly/article_caea1226-64ea-56a6-912b-08c5b7d10aa7.html.

Since Bill C-7 does not contain any additional safeguards for cases in which mental illness is the sole underlying condition, it is significant that policymakers continue to envision MAiD as a last resort. Minister Lametti claimed in Parliament that MAiD for mental illness would be granted only “where everything has been tried, where the person is an adult capable of making up their own mind and there is no remedy.” 20 20 Canada, “Evidence,” Standing Committee on Justice and Human Rights, February 14, 2023, 44th Parliament, 1st Session, no. 050, https://www.noscommunes.ca/documentviewer/en/44-1/JUST/meeting-50/evidence. In the implementation of one of the delays of MAiD for mental illness alone, then parliamentary secretary to the Minister of Families, Children and Social Development, Ya’ara Saks, claimed that this expansion will apply only to those “who have been presented treatment after treatment” and “have tried everything imaginable to address their suffering.” 21 21 Canada, House of Commons Debates, February 15, 2023, 44th Parliament, 1st Session, vol. 151, no. 161, https://www.ourcommons.ca/documentviewer/en/44-1/house/sitting-161/hansard#Int-12065422.

Nevertheless, in contrast to the repeated assertions of policymakers and judges since Carter, the clinicians who administer MAiD do not treat it as an option of last resort. This can be seen from three indicators: the relative frequency of MAiD, the diminishing percentage of denied MAiD requests, and the average time used by clinicians to assess and provide a MAiD death.

To be clear, warning lights have long been flashing that MAiD assessors and providers are not treating MAiD as a last resort. Even before Bill C-7, which removed the provision that death had to be “reasonably foreseeable” to qualify for MAiD, the Attorney General of Canada did not contest expert evidence that “some clinicians gained comfort with extending prognostic timeframes out to many years.” 22 22 J. Arvay, Letter re: Lamb and BCCLA v. AGC to the Supreme Court of British Columbia, B.C. Civil Liberties Association, File no. 20001, September 6, 2019, https://bccla.org/wp-content/uploads/2019/09/2019-09-06B-LT-Court_adjournment-of-triaL.pdf.

Also not contested by the government was that patients could qualify for MAiD on the basis of “intent to refuse treatment” as opposed to actual refusal of treatment. 23 23 J. Arvay, Letter re: Lamb and BCCLA v. AGC. This also implies, according to the Health Law Institute at Dalhousie University, that the mere intent to stop eating or drinking could qualify a patient for assisted dying, 24 24 Health Law Institute, Dalhousie University, “VSED and VSPeC,” End-of-Life Law and Policy in Canada (blog), accessed February 23, 2024, http://eol.law.dal.ca/?page_id=2475. compromising the efficacy of safeguards.

This expansive way of evaluating eligibility for MAiD has been matched by continued legislative expansion. The following year, after Bill C-7 included virtually every disability as a potentially qualifying condition, “vision/hearing loss” was recorded by MAiD assessors so frequently as a qualifying form of intolerable suffering that it was listed as a distinct reporting category in the annual Health Canada report on MAiD. 25 25 Health Canada, “Fourth Annual Report,” 27.

Federal data provide further evidence that medical professionals are not viewing MAiD as an option of last resort only. Statistics Canada does not recognize MAiD as a cause of death (which will be explained in a later section of this report), but the number of people dying by MAiD, as reported in Health Canada’s annual MAID reports, can be compared with the number of people who die from causes that Statistics Canada recognizes. When the two sets of data are brought together, it can be seen that MAiD is now at least the fifth leading cause of death in Canada, on par with the number of deaths from cerebrovascular conditions (figure 1).

Statistics Canada reported 13,915 deaths caused by cerebrovascular diseases in 2022. 26 26 Statistics Canada, Table 13-10-0394-01. Although Health Canada’s 2022 MAiD report does not list cerebrovascular diseases as a specific category for the underlying medical condition in a MAiD request, it does specify that cardiovascular conditions, which would include cerebrovascular diseases, were the second most common underlying condition for MAiD deaths, at 18.8 percent of all MAiD deaths in 2022, or roughly 2,490 deaths out of 13,241 total MAiD deaths. 27 27 Health Canada, “Fourth Annual Report,” 5. It is unknown how many of these MAiD deaths from cardiovascular conditions recorded by Health Canada were recorded as cerebrovascular deaths by Statistics Canada, but it is likely that some were. Therefore, the adjusted number of cerebrovascular deaths (less the MAiD deaths with cerebrovascular as the underlying condition) is likely on par with the number of MAiD deaths in 2022.

The original intention of MAiD safeguards was to ensure that MAiD requests would be carefully scrutinized rather than merely rubberstamped. The trial judge in Carter twice cited that only one in ten explicit assisted-suicide requests in Oregon ended in assisted death as evidence of the meticulous deliberations that physicians would take if assisted dying were to be legalized in Canada. 28 28 Carter v. Canada (Attorney General), 2012 BCSC 886 at paras 418, 436. This was also a key argument in the Supreme Court’s decision. 29 29 Carter v. Canada (Attorney General), 2015 SCC 5 at paras 117–21. Yet in 2022, over 81 percent of all explicit MAiD requests ended with death from MAiD. A notable reason for this unprecedented high mortality rate from assisted dying is that the percentage of MAiD requests that is denied in Canada is low and has continued to decrease each year. According to federal data, 8 percent of total written requests for MAiD in 2019 were deemed ineligible. In 2020, the percentage was 6.1 percent, in 2021 it was 4.1 percent, and in 2022 it was 3.5 percent. 30 30 Health Canada, “Fourth Annual Report,” 7. The actual percentage of requestors who are denied is likely even lower, given that when a person is denied, they may continue to pursue their request with other medical professionals and may eventually be approved.

Further cause for concern is found in the length of time that clinicians use to assess requests, explore alternatives to MAiD, provide additional disability or palliative services, and support patients in their healthcare decisions. All of these necessary services require time for consultations, connecting with available resources, and offering alternative treatments. Yet, for patients whose MAiD request was granted, the median length of time between the request and the provision was only eleven days in 2022, and nine days in 2021 (figure 2). 31 31 Whenever a figure is updated in a more recent government report, this Cardus report uses the more recent figure. For example, it uses the median of twenty-nine days for 2022, which was recorded in the Oregon Year 26 (2023) Annual Report, rather than the median of thirty days, which was recorded in the Year 25 (2022) Annual Report.

In contrast, as depicted in figure 2, Oregon, which requires patients to have a terminal illness with a prognosis of under six months, has nearly triple the median time for the assessment and provision. It is also noteworthy that, in 2020, after Oregon provided an exemption to the fifteen-day waiting period between a first and second request, the median time for assisted death dropped from forty-three to thirty-four days. 32 32 Oregon Health Authority, “Death with Dignity Act Annual Reports,” Year 23 (2020) Annual Report, https://www.oregon.gov/oha/ph/providerpartnerresources/evaluationresearch/deathwithdignityact/pages/ar-index.aspx; Oregon Health Authority, Year 26 (2023) Annual Report. In Canada, when the ten-day minimum assessment period was entirely removed in 2021, there was no significant change in the median length of time to assess and provide MAiD.

Canada’s adoption of MAiD has dramatically exceeded expectations, with a much higher number of deaths than expected. This number has continued to increase year after year. Even critics could barely imagine the consequences of the legalization of MAiD. Bioethicist Margaret Somerville testified at the Special Joint Committee on Physician-Assisted Dying on February 4, 2016, that she estimated there could be “between 11,000 and 12,000 Canadians being killed by lethal injection” every year if Canada followed the Belgium and Netherlands model—although “I could almost not believe it when I worked out those figures.” 33 33 Canada, “Evidence,” Special Joint Committee on Physician-Assisted Dying, February 4, 2016, 42nd Parliament, 1st Session, no. 012, https://www.parl.ca/documentviewer/en/42-1/pdam/meeting-12/evidence. Medical practitioners writing in the New England Journal of Medicine, one of the most highly regarded medical journals in the world, stated in 2020 that “approximately 2000 euthanasia cases are expected in Canada each year.” 34 34 I. Ball et al., “Organ Donation after Medical Assistance in Dying—Canada’s First Cases,” New England Journal of Medicine 382, no. 6 (2020): 576–77, https://doi.org/10.1056/NEJMc1915485. This proved to be a dramatic underestimate just one year later. In 2020, there were 7,611 MAiD deaths. In 2021 there were 10,092, and in 2022 the number was 13,241, already surpassing Somerville’s estimates.

Health Canada has also made inaccurate predictions concerning MAiD death numbers, predicting in 2018 that “Canada would reach a steady state of 2.05 percent of total deaths attributed to medical assistance in dying.” 35 35 Government of Canada, “Regulations for the Monitoring of Medical Assistance in Dying,” (2018), 3089. In November of 2022, it stated that “Canada would reach a steady state growth in total deaths attributed to MAiD from 2023 to 2033, reaching 4% of total deaths in 2033.” 36 36 Government of Canada, “Regulations Amending the Regulations,” (2022), 2567. In reality, the 4 percent rate was exceeded in 2022, just months after this statement was made.

Thus it cannot be said that MAiD is being provided only on rare occasions. The head of Quebec’s commission on end-of-life care, which monitors MAiD in that province, sparked headlines when he claimed that in Quebec “we’re now no longer dealing with an exceptional treatment, but a treatment that is very frequent.” 37 37 J. Serebrin, “Quebecers No Longer Seeing Doctor-Assisted Deaths as Exceptional, Says Oversight Body,” CBC News, August 15, 2023, https://www.cbc.ca/news/canada/montreal/quebecers-maid-no-longer-last-resort-oversight-body-1.6936530. While the rate of MAiD in Quebec is higher than in most other provinces, his statement about MAiD being unexceptional is true of Canada as a whole.

Legislators and judges may be wrong, of course, in estimating the outcome of their decisions. But MAiD in Canada is not only proceeding far beyond the expectations that continue to be set for it. It is also outpacing every other jurisdiction in which assisted dying exists, as demonstrated in the next section of this report.

More than sixteen jurisdictions globally have decriminalized or legalized assisted dying in some form. Of these, Belgium, the Netherlands, Switzerland, and the US states of California, Oregon, and Washington report the highest rates of assisted deaths. When Canada is graphed alongside them, the rapid rise in Canada’s MAiD program is evident (figure 3).

Either in absolute numbers or when weighed as a percentage of deaths, Canada’s MAiD program is by far the world’s fastest-growing assisted-dying program. No other assisted-dying program has seen such a dramatic rate of growth. This nation, with its nascent MAiD program, has already outpaced the jurisdictions that have had assisted-dying programs for much longer. In 2016, the first year of legalization, MAiD deaths as a percentage of all deaths in Canada already matched Oregon’s percentage (where assisted dying was legalized in 1997) and Washington’s (legalized in 2009). In 2019, Canada’s percentage surpassed Switzerland’s (legalized in some forms in 1941), and in 2020, surpassed Belgium’s (conditionally decriminalized in 2002).

It is important to note that MAiD deaths as a percentage of all deaths is an imperfect method of comparison across jurisdictions, since each jurisdiction has its own legislation setting out the circumstances in which assisted dying is permitted. As such, a precise apples-to-apples comparison is not possible.

For instance, in Oregon, Washington, and California, only assisted suicide is an option (euthanasia is excluded), and only for those with a terminal disease that has a prognosis of under six months. 41 41 J. Nicol, “Medical Assistance in Dying: The Law in Selected Jurisdictions Outside Canada,” Library of Parliament Hill Studies, Publication No. 2015-116-E, 2015 (rev. 2021), https://lop.parl.ca/sites/PublicWebsite/default/en_CA/ResearchPublications/2015116E. In Belgium and the Netherlands, by contrast, both euthanasia and assisted suicide are available, for terminal and non-terminal physical and mental illnesses or conditions. 42 42 Nicol, “Medical Assistance in Dying.” In Switzerland, only assisted suicide (not euthanasia) is ethically permitted for any consenting persons who are not “healthy.” 43 43 Switzerland’s Criminal Code does not lay out detailed regulations for assisted suicide, but Swiss medical bodies have established their own ethical guidelines for physicians: “Not ethically justifiable in accordance with these guidelines is the performance of assisted suicide in persons who are healthy”; Swiss Academies of Arts and Sciences, “Dying and Death,” https://www.samw.ch/en/Ethics/Topics-A-to-Z/Dying-and-death.html.

Canada’s model is similar to that of Belgium and the Netherlands in that it now allows assisted death for persons with non-terminal illnesses, which expands eligibility to those living with disabilities (whose death is not reasonably foreseeable). While MAiD on the sole basis of mental illness has been deferred until 2027, Canada effectively allows MAiD on the basis of a “completed life,” a criterion rejected in the Netherlands, according to Madeline Li, the scientific lead of the Canadian Association of MAiD Assessors and Providers’ nationwide training curriculum. 44 44 Canada, “Evidence,” Special Joint Committee on Medical Assistance in Dying, October 18, 2022, 44th Parliament, 1st Session, no. 020: 12, https://www.parl.ca/Content/Committee/441/AMAD/Evidence/EV11991759/AMADEV20-E.PDF.

Since the implementation of Bill C-7 in 2021, therefore, Canada has permitted clinicians to perform an assisted death on patients who otherwise would not have died but possibly have lived for years or decades. As a result, Canada’s situation is different from that of the jurisdictions that restrict assisted dying to patients whose death is reasonably foreseeable. In those jurisdictions (Oregon, for instance), there is a direct zero-sum relationship between MAiD and natural death: one MAiD application denied, or never requested in the first place, is equal to one natural death gained. In Canada, by contrast, where the death does not need to be reasonably foreseeable, a denied or not-requested MAiD application may not mean a subsequent, imminent natural death.

Despite the fact that MAiD deaths as a percentage of all deaths is an imperfect method of comparison across jurisdictions, this Cardus report uses this data point because the Canadian government reports MAiD deaths as a percentage of all deaths and uses this to characterize Canada’s growth as consistent and expected. 45 45 It is interesting that federal agencies continue to use “MAiD as a percentage of total deaths” as a metric, since the federal government does not actually consider MAiD to be a “cause of death” for formal reporting purposes. Instead, Statistics Canada records only the underlying illness or disability as the cause of death. The only federal MAiD data come from the annual report that Health Canada produces based on data submitted by the provinces and territories. This Cardus report points out that the percentage in Canada is in fact not consistent with the percentage in other jurisdictions.

While the rate of assisted dying generally increases in every jurisdiction that legalizes it, the rate of increase in Canada has been unusual. California, with a population similar to Canada’s, legalized assisted dying the same year that Canada did. As of 2022, California still has not reached the rate of assisted deaths that occurred in Canada in year one of legalization.

Some of the international jurisdictions discussed above, like California, allow only assisted suicide (self-administered) rather than euthanasia (clinician administered), whereas Canada allows both. The question arises of why Canada adopted an expansive approach to assisted dying similar to the approach in Belgium and the Netherlands, which is associated with higher numbers, instead of permitting assisted suicide only, as is the case in several US states, an approach that is associated with lower and more stable growth.

The Netherlands is the only country with an assisted-dying rate that is comparable to Canada’s. It was the first country to effectively decriminalize euthanasia in 1981 and to formally legalize it in 2002. 46 46 C.F. Gomez, Regulating Death: Euthanasia and the Case of the Netherlands (New York: Free Press, 1991), 32; W. Rooney, U. Schuklenk, and S. van de Vathorst, “Are Concerns About Irremediableness, Vulnerability, or Competence Sufficient to Justify Excluding All Psychiatric Patients from Medical Aid in Dying?,” Health Care Analysis 26, no. 4 (2018): 326–43. https://doi.org/10.1007/s10728-017-0344-8. In 2013, thirty-two years after the country’s effective decriminalization, the percentage of total deaths crossed the 3 percent bar. 47 47 Author’s calculations using data sources cited in figure 3. Canada crossed the 3 percent threshold just six years after legalization. No jurisdiction with an assisted-dying program other than Canada and the Netherlands has crossed the 3 percent threshold. In the seventh year of its legalization, Canada crossed the 4 percent line.

Every Canadian province has seen a surge in MAiD deaths since legalization (figure 4). While mortality rates are usually similar across Canada, this is not the case with assisted dying. 48 48 Statistics Canada, “Table 13-10-0800-02, Age-Standardized Mortality Rate by Leading Cause of Death and Sex,” https://doi.org/10.25318/1310080001-eng. No research examining the striking disparities in the number of MAiD deaths between provinces has yet been published.

The disparities may be attributed at least partly to differing safeguards that each province has implemented. For instance, in Newfoundland and Labrador, where nurses follow a practice guideline that they may not initiate conversations with patients about MAiD, 49 49 College of Licensed Practical Nurses of Newfoundland and Labrador, “Medical Assistance in Dying Practice Guideline,” June 2021,https://www.clpnnl.ca/storage/Medical%20Assistance%20in%20Dying%20MAiD%20Practice%20Guideline%202021%20in%20template.pdf. The government did not release Newfoundland and Labrador’s data in 2016 and 2017 because of potential confidentiality issues. MAiD represented almost no deaths in 2016–17 and reached 1.52 percent of deaths in 2022. While this rate is low compared to that in other Canadian provinces, it is more than double the rate seen between 2016 and 2022 in the US jurisdictions that have the largest number of assisted deaths. Alberta, with its centralized MAiD intake protocol that may reduce the incidence of “doctor shopping,” had a rate in 2022 that was slightly above that of Belgium (2.5 percent versus 2.7 percent). Ontario’s 2022 rate of 3.3 percent is significantly above Belgium’s, while the rates in British Columbia (5.5 percent) and Quebec (6.6 percent) both surpass and continue to increase much faster than the Dutch rate (5.1 percent) (figure 4).

This rapid increase that is stated above is not reflected in Canadian official statements, however. Here are three examples: 50 50 Sources used for these calculations are provided in figure 3.

• The Department of Justice issued a press release in early 2020 stating that “MAiD deaths as a percentage of all deaths in Canada remains consistent with other international assisted-dying regimes.” 51 51 Government of Canada, “Government of Canada Proposes Changes to Medical Assistance in Dying Legislation,” Department of Justice news release, February 24, 2020, https://www.canada.ca/en/department-justice/news/2020/02/government-of-canada-proposes-changes-to-medical-assistance-in-dying-legislation.html. In fact, in 2019, MAiD deaths as a percentage of all deaths in Canada was 2.00 percent, higher than that of the world’s oldest assisted-suicide program, that of Switzerland, which caused 1.78 percent of that country’s total deaths in 2019.
• In internal briefings and talking points prepared for cabinet ministers by Health Canada and the Ministry of Justice, ministers were advised that “as expected, there has been a consistent and gradual increase in MAiD deaths over the last three years.” 52 52 Government of Canada, “Government of Canada Proposes Changes to Medical Assistance in Dying Legislation.” In fact, the number of MAiD deaths more than doubled during those years: from 4,493 deaths in 2018 to 10,092 deaths in 2021. 53 53 Health Canada, “Fourth Annual Report.” Nor was the increase expected. Health Canada’s briefing did not inform the cabinet ministers that the federal minister of health, as part of a public notice published in the Canada Gazette, had predicted in 2018 that Canada’s MAiD program was meant to already have “reach[ed] a steady state of 2.05 percent of total deaths attributed to medical assistance in dying.” 54 54 Government of Canada, “Regulations for the Monitoring of Medical Assistance in Dying,” 3089.
• Statistics Canada stated in 2022 that “the observed increases in medically assisted deaths are consistent with those seen internationally in jurisdictions where some form of assisted dying is legal, including the American states of Oregon and Washington, and the ” 55 55 Statistics Canada, “Medical Assistance in Dying, 2019 and 2020,” The Daily, January 10, 2022, https://www150.statcan.gc.ca/n1/daily-quotidien/220110/dq220110d-eng.htm. In fact, MAiD deaths as a percentage of total deaths in Canada in 2022 (4.18 percent) was over five times that of Oregon and Washington in the same year (0.63 percent and 0.65 percent, respectively). Canada moreover surpassed the rate of MAiD deaths that the Netherlands, the jurisdiction with the largest number of assisted deaths, reached just in 2020, almost two decades after formal legalization in that country.

Policymakers and researchers need accurate vital-statistics data so that they can understand the frequency and trends in natural deaths (from aging or disease) and unnatural deaths (from homicide, suicide, and accidents). For example, in Ontario, if a person with a terminal illness dies in a car accident, the accident is recorded as the immediate cause of death, although the illness could have been a contributing factor. 56 56 Ontario Ministry of Government and Consumer Services, Office of the Registrar General, “Handbook on Medical Certification of Death & Stillbirth,” 2019 ed., https://www.publications.gov.on.ca/store/20170501121/Free_Download_Files/300146.pdf. Consistent and public reporting is crucial for improvements to public health, not just within the province but across the country. This is acknowledged by Health Ontario, which quotes from the UN “Handbook of Vital Statistics Methods” that “it may truthfully be said that virtually every largescale problem in preventive medicine has been brought to light—in part at least—by statistics of death, and further that the adequacy of remedial or curative action is, in the last analysis, reflected in these same statistics.” 57 57 UN Statistical Office, “Handbook of Vital Statistics Methods, 1955,” quoted in Ontario Ministry of Government and Consumer Services, “Handbook on Medical Certification of Death & Stillbirth,” 2.

At present, there is a variety of ways in which Canadian federal and provincial government agencies record MAiD deaths for vital-statistics purposes. Some record MAiD deaths as “natural deaths,” by assigning the underlying condition as the immediate cause of death and MAiD as the secondary cause. Other government agencies record MAiD deaths as “unnatural deaths,” by assigning MAiD as the immediate cause of death. And some government agencies do not include MAiD in the death record at all, because they record only the underlying condition that led to MAiD.

A majority of provinces and territories (British Columbia, Alberta, Saskatchewan, Nova Scotia, New Brunswick, Prince Edward Island, Yukon, and Northwest Territories) do record MAiD as a cause of death, either as immediate cause or secondary cause. 58 58 J. Brown, L. Thorpe, and D. Goodridge, “Completion of Medical Certificates of Death after an Assisted Death: An Environmental Scan of Practices,” Healthcare Policy 14, no. 2 (2018): table 1, https://www.longwoods.com/articles/images/HPL_Vol14_No2-Brown-Table-1.pdf. Health Canada recommends that death certificates state the immediate cause of death as “the toxicity of the drugs administered for the purposes of a medically-assisted death.” 59 59 Health Canada, “Guidelines for Death Certificates,” April 26, 2017, https://www.canada.ca/en/health-canada/services/publications/health-system-services/guidelines-death-certificates.html.

In 2016, the policy of the Office of the Chief Coroner of Ontario was that MAiD be recorded as the immediate cause of death, with the qualifying illness recorded as contributor: “In the current legislative framework, the immediate cause of death will generally be provided as Combined Drug Toxicity, with the underlying condition that led to the MAiD request being provided as the Contributing Factor.” 60 60 Ontario Hospital Association, “Spotlight on Medical Assistance in Dying: The Role of the Coroner in Ontario,” September 2016, https://www.oha.com/Legislative%20and%20Legal%20Issues%20Documents1/Spotlight%20on%20MAID%20-%20Interview%20with%20the%20Coroner%20-%20final.pdf. But by November 2018 at the latest, and despite no change in legislation, the official recommendation changed:

When completing the death certificate, please note that the illness, disease or disability leading to the request for medical assistance in dying is to be recorded as the cause of death. . . . No reference to MAiD or the drugs administered for the purposes of MAiD should be included on the Medical Certificate of Death. 61 61 Health Ontario, “Process Overview and Checklist: Reporting a MAID Death to the Office of the Chief Coroner of Ontario.” Copy in possession of the author.

Other provinces followed suit. As in Ontario, death certificates in Quebec and Newfoundland and Labrador contain no mention of assisted dying. By 2018, the death certificates issued in seven provinces did indicate MAiD but classified the manner of death as “natural.” 62 62 Brown, Thorpe, and Goodridge, “Completion of Medical Certificates,” table 1. Note that since 2018, the only change has been in Saskatchewan. See S. Taylor and A. Bridges, “Sask. Ends Practice of Recording Medically Assisted Deaths as Suicides,” CBC News, October 5, 2018, https://www.cbc.ca/news/canada/saskatchewan/maid-medically-assisted-deaths-suicide-saskatchewan-1.4851451. In Alberta, a MAiD death is recorded as caused by “administration of drugs X, Y, Z (due to) underlying illness, disease or disability” in an “unclassified” manner of death. 63 63 Brown, Thorpe, and Goodridge, “Completion of Medical Certificates,” table 1.

Statistics Canada, the agency that ultimately reports the national death statistics, takes another approach. According to Statistics Canada, no person in Canada dies from MAiD, even if the underlying illness or disability was not terminal. 64 64 Statistics Canada, “In the database, the underlying cause of death is defined as the disease or injury that initiated the train of morbid events leading directly to death. As such, MAID deaths are coded to the underlying condition for which MAID was requested,” Twitter, November 28, 2023, 10:34 a.m, https://twitter.com/StatCan_eng/status/1729539259076334034. MAiD is not “an official cause of death, and most cases involve people with serious underlying health conditions.” 65 65 APF Canada, “Claims Mislead on Canada’s Medical Assistance in Dying Law,” AFP Fact Check, August 31, 2022, https://factcheck.afp.com/doc.afp.com.32HB2WX. Statistics Canada records the underlying illness or condition as the only cause of death.

This practice results in a discrepancy between two federal departments, because Health Canada does report the number of MAiD deaths, through a separate data-collection process. Therefore, in official reporting, Health Canada’s annual reports on MAiD show an explosion in the number of MAiD deaths, whereas this explosion is completely hidden when Statistics Canada’s dataset of causes of death is consulted. As noted earlier, when MAiD is graphed alongside Statistics Canada data on causes of death, MAiD is virtually tied as the fifth leading cause of death as of 2022 (figure 1).

Compounding the problem, there is no method to independently review the accuracy of Health Canada’s MAiD reporting in Canada. The only monitoring agency for MAiD deaths across Canada is the federal minister of health, as tasked by Bill C-14. The same medical professionals who directly assess requests and administer MAiD are the ones who submit the data to Health Canada. The submission includes information about whether the recipient received palliative care, the age at death, and the percentage of rejected MAiD applications. Health Canada then publishes an annual report on the number of MAiD deaths.

Yet Statistics Canada, which policymakers and researchers rely on for economic and social data, is the agency responsible for official data on causes of death, and Statistics Canada does not consider MAiD to be a cause of death.

These criticisms about MAiD reporting and oversight are not new. Others, including the Canadian Society of Palliative Care Physicians and academic researchers, have criticized federal MAiD reporting as inaccurate or misleading. 66 66 Canadian Society of Palliative Care Physicians, “CSPCP Statement in response to the Special Joint Committee on MAiD report: MAiD in Canada: Choices for Canadians,” March 2023, https://www.pallmed.ca/wp-content/uploads/2023/03/CSPCP-response-to-AMAD-report_final-Mar-2023.pdf; J. Kotalik, “Federal Annual Reports on MAiD: Informative but Incomplete Picture,” in Medical Assistance in Dying (MAiD) in Canada: Key Multidisciplinary Perspectives, ed. J. Kotalik and D.W. Shannon (Cham, Switzerland: Springer, 2023), 127–42, https://doi.org/10.1007/978-3-031-30002-8_8. For instance, the widely repeated statistic that a high percentage of MAiD applicants received palliative care 67 67 J. Downar, S. MacDonald, and S. Buchman. “What Drives Requests for MAiD?” Canadian Medical Association Journal 195, no. 40 (October 16, 2023): E1385–87. https://doi.org/10.1503/cmaj.230259. was, until 2022 (the latest year with a publicly available report), based on a single question on the federal MAiD reporting form, with “yes,” “no,” and “do not know” as the response options (figure 5). The form did not ask MAiD assessors and providers about the type or quality of palliative care the patient received; it could have been anything from full hospice care to a simple prescription for painkillers. 68 68 Health Canada as of 2023 will address this concern by asking new questions about the type and duration of disability supports and palliative care. It is unclear as to why this information was not collected until 2023, however.

Additionally, the data-reporting mechanisms do not include any input from MAiD recipients prior to death, nor from medical professionals not otherwise involved in the assessment or provision. Instead, all reporting is self-reporting, recorded and submitted by the same persons who assess and provide MAiD. The same persons who could face criminal repercussions for not strictly following the criminal exemptions for assisted suicide and homicide under Canada’s federal laws regulating MAiD are the only ones reporting whether they adhered to the regulations.

Other countries have methods for verifying how accurate assisted-death reporting is. The Netherlands has Regional Euthanasia Review Committees that publish the number of assisted deaths, and these data can be compared against data given in death certificates. 69 69 A. van der Heide et al., “End-of-Life Practices in the Netherlands under the Euthanasia Act,” New England Journal of Medicine 356, no. 19 (2007): 1957–65, https://doi.org/10.1056/NEJMsa071143. This method of verification or comparison cannot be used in Canada, because an increasing number of provinces have removed any mention of MAiD from death certificates, as previously discussed.

Inconsistency in recording data by provincial and federal jurisdictions has consequences for the federal reporting about MAiD and all research or estimates based on it. For one, it limits researchers’ ability to verify MAiD numbers through federal and provincial comparisons, which is one tool to ensure reliability in reporting. For another, it affects research done using death certificates, which are an important resource for medical research. 70 70 J. Cohen et al., “Using Death Certificate Data to Study Place of Death in 9 European Countries: Opportunities and Weaknesses,” BMC Public Health 7, no. 283 (2007), https://doi.org/10.1186/1471-2458-7-283. These data issues may undermine policymakers’ ability to make informed decisions.

The systematic underestimation of MAiD in government statements and reporting is a serious impediment to understanding the scale of MAiD’s normalization in Canada and its abnormality with regard to other countries where some form of assisted dying is permitted. The record of deliberations in the Carter decision and in Parliament shows that when MAiD was decriminalized and legalized, it was not intended to become routine. There were frequent assertions that assisted dying was intended as a “last resort.” MAiD in Canada was intended to reflect assisted-dying regimes in other jurisdictions, and the public message to Canadians continues to be that Canada’s experience is similar to that of other jurisdictions, that the rate of increase is expected, and that the growth is gradual.

Yet, as demonstrated by the data discussed in this paper, MAiD in Canada is no longer unusual or rare. Federal predictions about the expected frequency of MAiD have significantly underestimated the numbers of Canadians who are dying by this means. More troubling, instead of physicians acting as “reluctant gatekeepers” for assisted dying, as the lawyers for the plaintiff in Carter envisioned, they appear highly favourable to MAiD requests, as shown by the available data on length of time from assessment to provision, the percentage of MAiD requests that are denied, and the sheer prevalence of occurrences.

Internationally, Canada is increasingly an outlier, as the world’s fastest-growing assisted-dying program and heading toward further liberalization of the eligibility criteria. It is difficult to understand how the Government of Canada continues to claim that the rate of MAiD in Canada is similar to rates in other jurisdictions. Understanding this growth is essential for evaluating the use and possible abuse of MAiD, particularly in light of future expansion to eligibility criteria.

For policymakers and the public to properly understand and respond to the Canadian reality, it is essential that government agencies collect consistent data and issue correct statements.