This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
Alzheimer’s disease and related disorders affect a growing number of people worldwide. Quality of life is generally good in the early stages of these diseases. However, many individuals fear living through the advanced stages. Such fears are triggering requests for medical assistance in dying (MAiD) by patients with dementia. Legislation was recently passed in Canada and the province of Quebec allowing MAiD at the explicit request of a patient who meets a set of eligibility criteria, including competence. Some commentators have argued that MAiD should be accessible to incompetent patients as well, provided appropriate safeguards are in place. Governments of both Quebec and Canada are currently considering whether MAiD should be accessible through written requests made in advance of loss of capacity.
Aimed at informing the societal debate on this sensitive issue, this study will compare stakeholders’ attitudes towards expanding MAiD to incompetent patients with dementia, the beliefs underlying stakeholders’ attitudes on this issue, and the value they attach to proposed safeguards. This paper describes the study protocol.
Data will be collected via a questionnaire mailed to random samples of community-dwelling seniors, relatives of persons with dementia, physicians, and nurses, all residing in Quebec (targeted sample size of 385 per group). Participants will be recruited through the provincial health insurance database, Alzheimer Societies, and professional associations. Attitudes towards MAiD for incompetent patients with dementia will be elicited through clinical vignettes featuring a patient with Alzheimer’s disease for whom MAiD is considered towards the end of the disease trajectory. Vignettes specify the source of the request (from the patient through an advance request or from the patient’s substitute decision-maker), manifestations of suffering, and how close the patient is to death. Arguments for or against MAiD are used to elicit the beliefs underlying respondents’ attitudes.
The survey was launched in September 2016 and is still ongoing. At the time of submission, over 850 respondents have returned the questionnaire, mostly via mail.
This study will be the first in Canada to directly compare views on MAiD for incompetent patients with dementia across key stakeholder groups. Our findings will contribute valuable data upon which to base further debate about whether MAiD should be accessible to incompetent patients with dementia, and if so, under what conditions.
Medicine aims to relieve patient suffering and cure illness [
Outside of Canada, euthanasia and/or physician-assisted suicide have now been legalized in four countries (The Netherlands, Belgium, Luxemburg, and Colombia), four US states (Oregon, Washington, Vermont, and California) and the District of Columbia. In the scientific literature,
In most permissive jurisdictions, MAiD is not available to individuals with dementia. Patients are either still capable but not near enough to the end of life, or they are close enough to the end of life but no longer capable. However, in the Netherlands, where being at the end of life is not a condition of eligibility for MAiD, 109 euthanasia requests from competent patients in the early stages of dementia were granted in 2015 [
Until recently, MAiD was prohibited in Canada under several provisions of the
The Court’s ruling catalyzed the Government of Canada to engage in consultation and draft legislation specifying the eligibility criteria and procedural safeguards for access to MAiD. Bill C-14 came into force on June 17, 2016 [
The administering by a medical practitioner or nurse practitioner of a substance to a person, at their request, that causes their death, or the prescribing or providing by a medical practitioner or nurse practitioner of a substance to a person, at their request, so that they may self-administer the substance and in doing so cause their own death [
MAiD thus encompasses both euthanasia and physician-assisted suicide, as defined above. A person has a grievous and irremediable medical condition only if all of the following criteria are met:
(a) they have a serious and incurable illness, disease, or disability; (b) they are in an advanced state of irreversible decline in capability; (c) that illness, disease, or disability or that state of decline causes them enduring physical or psychological suffering that is intolerable to them and that cannot be relieved under conditions that they consider acceptable; and (d) their natural death has become reasonably foreseeable, taking into account all of their medical circumstances, without a prognosis necessarily having been made as to the specific length of time that they have remaining [
Contrary to recommendations made by a Provincial-Territorial Expert Advisory Group on Physician-Assisted Dying [
On June 10, 2014, eight months before the Supreme Court’s ruling in
Care consisting in the administration by a physician of medications or substances to an end-of-life patient, at the patient’s request, in order to relieve their suffering by hastening death [
This definition corresponds to euthanasia, as defined above. Nurse practitioners are not authorized to administer aid in dying under the Quebec legislation. The legislation became effective on December 10, 2015. Eligibility for MAiD is restricted to competent adults from Quebec who are at the end of their lives, have made persistent explicit requests for MAiD, and:
Suffer from a serious and incurable illness, are in an advanced state of irreversible decline in capacity, and experience constant and unbearable physical or psychological suffering which cannot be relieved in a matter the patient deems tolerable (article 26) [
For greater clarity, article 51 specifies that MAiD may not be requested by means of an advance medical directive.
Quebec Bill 52 has drawn both opposition and support, with some supporters recommending that access to MAiD be extended to incompetent patients, provided appropriate safeguards are in place [
In both Canada and the province of Quebec, eligibility criteria for MAiD currently exclude patients who have become incompetent due to Alzheimer’s disease or other forms of dementia. These are serious incurable conditions that progressively and irreversibly erode patients’ abilities to perform basic activities of daily living. Additionally, many affected individuals develop serious clinical complications (eg, eating problems, pneumonia) and distressing symptoms (eg, pain, dyspnea) that are difficult to manage [
Arguments in favor of MAiD generally include individual autonomy and freedom of choice, the inability to relieve suffering in some cases, the absence of a moral distinction between withholding/withdrawing potentially life-sustaining treatment and MAiD, and the claim that permitting MAiD allows the establishment of stronger safeguards and oversight for the entire spectrum of end-of-life medical care through carefully-designed regimes. Arguments against MAiD include: the sanctity of life; the need to protect socially vulnerable populations from abuse and social discrimination; concerns about the
Other arguments are raised against MAiD when referring specifically to patients rendered incompetent by advanced dementia: patients’ potential to adapt to their disease, which may change previously expressed wishes (the so-called “disability paradox”); the impossibility of health care providers and families engaging in meaningful conversations with the patient to confirm the wish to die; and practical difficulties in assessing suffering, balancing current preferences against earlier wishes laid down in a now-forgotten request, and choosing the right moment to carry out the request. Complying with an advance request for MAiD also raises the philosophical question of whether a request made by a previously competent person should have any authority over the life of a person who now has severe dementia [
Major groups of stakeholders likely to be impacted by extending MAiD to incompetent patients include older adults, relatives of patients with dementia, physicians, and nurses. Systematic reviews of quantitative studies from several countries, including Canada, show increasing support from these groups of stakeholders for MAiD in cases of competent terminally-ill patients experiencing severe pain who make an explicit request [
In conclusion, growing knowledge of possible clinical complications of advanced dementia, and current access to MAiD for competent adults, will likely trigger advance requests for MAiD from Canadians diagnosed with dementia [
Studies that have elicited attitudes towards assistance in dying for incompetent patients (mostly due to dementia).
First author (year of publication) | Country | Age of sample, |
Sex of sample, |
Response |
Sample |
Percent judging |
|
Koenig (1996) [ |
United States (Durham, NC) | 61% over 75 years | 23 | 86 | 168 | 14 | |
Van Holsteyn (1998) [ |
The Netherlands | 18 or older | unspecified | 46 | 911 | 45 | |
Ryynänen (2002) [ |
Finland | 18 to 70 | 42 | 59 | 587 | 48 | |
Rietjens (2005) [ |
The Netherlands | 20 to 93 | 39 | 78 | 1388 | 62 | |
Williams (2007) [ |
United Kingdom (London) | 43 (17) | 46 | 71 | 725 | 57 to 59 | |
Lindblad (2010) [ |
Sweden | 20 to 84 | 50 | 52 | 625 | 60 to 65 | |
Kouwenhoven (2013) [ |
The Netherlands | 53 (15) | 54 | 78 | 1960 | 77 | |
Rurup (2006) [ |
The Netherlands | 57 | 38 | 72 | 136 | 89 | |
Ryynänen (2002) [ |
Finland | 24 to 87 | 48 | 62 | 506 | 8 | |
Rietjens (2005) [ |
The Netherlands | 63% aged 40 to 55 | 76 | 81 | 391 | 6 | |
Rurup (2006) [ |
The Netherlands | 41 | 51 | 96 | 107 | 16 | |
Lindblad (2010) [ |
Sweden | 32 to 79 | 69 | 56 | 667 | 13 to 15 | |
Kouwenhoven (2013) [ |
The Netherlands | 51 (8) | 65 | 41 | 793 | 33 | |
Ryynänen (2002) [ |
Finland | 20 to 63 | 6 | 73 | 582 | 23 | |
Rurup (2006) [ |
The Netherlands | 34 | 17 | 94 | 148 | 57 | |
Gielen (2009) [ |
Belgium (Flanders) | 44 (9) | 12 | 70.5 | 415 | 52 to 56 | |
Inghelbrecht (2009) [ |
Belgium (Flanders) | 42% aged 36 to 45 | 12 | 62.5 | 3321 | 57 | |
Kouwenhoven (2013) [ |
The Netherlands | 44 (11) | 10 | unspecified | 1243 | 58 |
aRange reported when several items were used to measure acceptability
Restricted to Quebec with plans for extension to other Canadian provinces, this study will elicit and compare the attitudes of four groups of stakeholders (seniors, relatives of persons with dementia, physicians, and nurses) towards MAiD for incompetent patients with dementia, the beliefs underpinning stakeholders’ positions on this matter, and their opinions as to whether proposed safeguards can adequately protect incompetent patients. Based on findings in other countries [
An anonymous province-wide postal survey using clinical vignettes will be conducted on random samples of French-speaking Quebec residents belonging to one of four groups of stakeholders: (1) community-dwelling seniors aged 65 years or older, (2) relatives of persons with dementia, and (3) physicians and (4) nurses likely to be involved in end-of-life decision making. In Quebec, 94% of the population speaks French [
The random sample of community-dwelling seniors will come from the Quebec health insurance database. Relatives of patients with dementia will be reached through regional Alzheimer Societies. To protect their members’ right to privacy, participating Societies will randomly select a predetermined number of potential participants (proportional to the size of their memberships) and distribute the survey package directly to them per our instructions. Assistance in managing the survey will be provided by our research staff to any Society that expresses the need. Practicing physicians and nurses will be recruited through their respective professional bodies, excluding those in full-time administrative, teaching, or research positions due to their limited direct contact with patients. Specialties for physicians will be restricted to family medicine, geriatrics, internal medicine, neurology, psychiatry, and intensive care; for nurses, specialties will be restricted to geriatrics/gerontology and end-of-life care.
The survey and questionnaires were designed using strategies shown to maximize response rates and data quality [
Outline of the postal survey.
A letter of endorsement from the Federation of Quebec Alzheimer Societies is also included, with a postcard bearing the respondent’s name, to be returned separately from the questionnaire. Returned postcards make it possible to identify sampled individuals who have returned the questionnaire–either by mail or electronically–while preserving the anonymity of their answers. The postcard also serves the purpose of identifying sampled individuals who are no longer eligible (eg, seniors who are now institutionalized or too cognitively impaired to participate). The names of those who return the postcard are immediately removed from the mailing list to prevent further reminders. At the close of the postal survey, nonrespondents receive a 3-item questionnaire asking: (1) for their reasons for not participating (eg, felt questions were biased, lack of time, or doubt that anonymity can truly be preserved); (2) how comfortable they are with the current Quebec legislation that gives competent patients access to MAiD if certain conditions are met; and (3) whether they favor or oppose allowing physicians to administer MAiD to incompetent patients, with proper safeguards in place. The latter two questions will be used to assess nonresponse bias.
After stating the eligibility criteria for MAiD as defined in Quebec’s
Part 2 explores related issues, such as whether respondents have filled out an advance directive for themselves, personally know someone with dementia, or have ever accompanied a dying relative or friend. Respondents are also asked the likelihood that they would request MAiD in advance of loss of capacity, should they be diagnosed with Alzheimer’s disease, or ask a physician to comply with such a request drafted by a loved one under similar circumstances.
Part 3 collects sociodemographic data from all respondents (eg, age, gender, ethnicity, degree of religiosity) and contains additional group-specific questions. For seniors, these questions include educational attainment and self-rated health. Relatives are asked how long ago the person with dementia was diagnosed and their current level of cognitive functioning. Questions for physicians and nurses explore their experience in caring for dying and dementia patients, training in palliative care, and exposure to MAiD requests from patients or patients’ relatives. Physicians are also asked whether they would be willing to provide such assistance, were it legal. Few physicians currently administer MAiD in Quebec, and to preserve their anonymity, surveyed physicians are not asked whether they have in fact provided such assistance in the past.
Questionnaires were developed in English with input from renowned English-speaking content experts from countries where assisted dying is legal or not criminalized. The questionnaires were then translated into French and pretested through cognitive interviews performed by a research assistant with representatives of the four groups of stakeholders (n=20). Interviews were aimed at assessing the length of the questionnaires, clarity of the questions, uniformity in comprehension, and respondent comfort with the content [
The data will be analyzed in four consecutive steps. First, we will compare nonrespondents, respondents to the one-page questionnaire only, and late versus early respondents to the full questionnaire, to detect response bias and establish sample weights where needed. Response rates will be reported as recommended by the American Association for Public Opinion Research. In Step 2, we will study patterns of item nonresponse and determine whether imputing missing data would be appropriate [
The data from our four samples will first be summarized with proportions and associated confidence intervals. In the worst-case scenario of equal proportions for and against a given end-of-life practice, two-sided 95% confidence intervals for proportions require 385 respondents per sample when the semiinterval width is set at 5% (nQuery Advisor, version 7.0). The sample size required for reliably fitting multilevel models depends on several factors, including sample size at each level of the analysis, number and type of predictors included in the model, intraclass correlation, and model complexity. Recent Monte Carlo simulation studies on multilevel models for binary and continuous outcomes suggest that 100 to 200 level-2 units (ie, survey respondents) with 5 to 10 level-1 units (ie, questionnaire items) ensure model convergence and provide adequate power for testing fixed and random effects [
This study will investigate views on sensitive issues. While there are no physical risks to participants, psychological risks must be acknowledged. Questionnaires may trigger emotional distress in some participants or revive painful memories of the death and suffering of a loved one. In an effort to minimize these risks, the cover letter that accompanies the questionnaire includes contact details for a support resource, if needed. Participation in the survey is voluntary and answers are anonymous. Signed consent is not required; in anonymous surveys, implicit consent is inferred from respondents who return the questionnaire [
The survey was launched in September 2016 among relatives of patients with dementia, physicians, and nurses, and is still ongoing. The survey will be launched among older adults as soon as we receive a random list of names extracted from the Quebec health insurance database.
To the best of our knowledge, this study will be the first to uncover Quebec stakeholders’ attitudes towards MAiD for incompetent patients with dementia, which is a vulnerable and rapidly expanding population of patients. Dementia affects more than 37 million people worldwide, with a projected increase to over 115 million by 2050 [
Do stakeholders believe that MAiD should be made available to patients who have reached an advanced stage of dementia? Under what circumstances? Are other end-of-life practices viewed as more appropriate? Can proposed safeguards on MAiD adequately protect vulnerable individuals from abuse and coercion? Do views on these issues vary markedly within and between groups of stakeholders? By using a proven research methodology and identical questions across stakeholder groups, this study will provide answers to these questions, which have yet to be explored in Canada, and have only been partially investigated in other countries.
The current study has strengths and limitations. Strengths include: the timeliness of the survey, which will inform ongoing legislative activities; random selection of potential respondents from four highly relevant groups of stakeholders; the anonymity of answers, which decreases bias due to social desirability; the care taken in designing and testing the questionnaires with input from international content experts; and our decision to administer a common set of questions to all four groups of stakeholders, enabling direct comparison of their views on MAiD for incompetent patients with dementia. The presence of investigators on the research team who support, and others who oppose, extending MAiD to these patients is another strength, as it minimizes the risk of biased questions and increases uptake of research findings [
Although not without limitations, surveys contribute invaluable data to inform ethical debates, public policy development, and future research on sensitive issues such as whether MAiD should be extended to incompetent patients with dementia [
Low response rates threaten the external validity of findings from attitude surveys. To counter this problem, both the survey and the questionnaires were designed using strategies that comprehensive reviews have shown to be effective [
One limitation is the restriction of the survey to the province of Quebec. We plan to extend the survey to the rest of Canada in the near future, using the same questionnaires to enable provincial/territorial comparisons. Short case descriptions with a limited number of possible answers fall short of capturing the complexity of end-of-life decision making [
medical assistance in dying
We acknowledge the contribution of François Thibeault, PhD, who developed and pretested draft versions of the questionnaires in close collaboration with the coinvestigators. We also thank Simon Brière, Eng, for his assistance in creating and managing the Web survey. The study is funded by an unrestricted grant from the Alzheimer Society of Canada (Grant # 16 11). The Society approved the study protocol, but has no role in designing the questionnaires, data collection, statistical analysis, interpretation, or dissemination of survey results.
GB is the principal investigator and lead author. As coinvestigators, MA, JD, MFD, and SK contributed to the development of the study protocol and grant application to the Alzheimer Society of Canada. CR implemented the study and is managing the postal survey under the close supervision of the principal investigator. As part of his Master’s degree in health sciences, and under the cosupervision of GB, VT designed and is currently running the qualitative study exploring attitudes of persons with early dementia towards MAiD. All authors except VT were actively involved in designing the questionnaires, with special input from CMH, SP, and LVDB. All authors read and approved the final version of the paper.
None declared.