Breslin_cropJonathan Breslin, PhD, is an Ethicist at the Southlake Regional Health Centre and Mackenzie Health in Ontario


Following the recent passage of Canadian legislation governing assistance in dying many might ask whether the law should be amended in the future to allow for request for medical assistance in dying in advance of becoming mentally incapable of doing so. This question is most relevant to people diagnosed with conditions like Alzheimer dementia (AD). The federal government made a commitment to study advance requests prior to passing Bill C-14. There are a host of problems with advance requests for medical assistance in dying that warrant deep and thorough reflection.

At first glance the idea of allowing advance requests for medical assistance in dying seems intuitive. Most Provinces in Canada have legislation that supports patients expressing wishes about medical treatment in advance, including when and under what conditions they would want life-sustaining treatment. So it seems natural to extend the practice to include medical assistance in dying. If people can refuse chemotherapy, feeding tubes, ventilators, or surgery in advance of becoming incapable, knowing this would result in their deaths, they should also be able to request medical assistance in dying in advance (so the argument goes). It would enable the patient to retain some level of autonomy, despite mental incapacity. People diagnosed with AD typically won’t find themselves dependent on life-sustaining interventions that they can refuse in advance, with the exception of feeding tubes at the very end stage of the disease. This means that the existing process of expressing wishes about medical treatment in advance are of limited use in the case of AD because there are few life-sustaining interventions to refuse through the later stages.  Allowing advance requests for medical assistance in dying will enable those diagnosed with AD to avoid the prolonged deterioration associated with the disease.

When people utilize what are often referred to as ‘advance directives’ to express wishes about medical treatment in advance, this is not an informed consent process. People who complete advance directive documents are not required to actually inform themselves about any of the illnesses or treatments about which they express wishes, nor are they required to discuss their wishes with health care professionals.  In fact, they don’t even have to be diagnosed with any illness or condition; an advance directive can be entirely hypothetical.  This is one of the main reasons why, in Ontario for example, those wishes do not actually function as directives to health care professionals but rather are wishes that must be interpreted and incorporated into the decision making process by substitute decision makers.  Allowing advance requests for assistance in dying crosses the line from expressing wishes to providing advance consent, which has implications for how that advance request is treated later on.  An advance request for assistance in dying would be viewed as consent that must be respected, not an expression of wishes that must be interpreted under the circumstances.

Moreover, serious problems surface when one begins to consider implementation, particularly in the case of AD.

Firstly, the patient’s substitute decision makers and care providers will face a difficult task when determining whether the patient’s situation meets the conditions expressed in the advance request. While this is a problem with advance directives in general – statements always require interpretation – it’s particularly so in the context of advance requests for assistance in dying in AD. For example, suppose someone states that she would like assistance in dying once she is no longer able to recognize her family. How would care providers and substitute decision makers determine when that condition is met? When she forgets her family members’ names? Or when her behavior indicates that she no longer recognizes them?  If there are any moments of recognition of any family member, would that render the condition unmet? What if the person doesn’t recognize family members as family members, but is still appears to enjoy interacting with them?  In order to avoid such problems, the advance requests would have to be extremely carefully and specifically worded, with extensive clarifications and qualifications – a task that might be impossible to achieve in practice.

Secondly, personality change is common in the course of AD and patients may find pleasure in things they never enjoyed before.  So, if the conditions in the advance request are met, but the person finds pleasure in certain new experiences and has some modicum of perceived quality of life, should the advance request be carried out regardless?  While patients with advanced AD may not be technically mentally capable of requesting assistance in dying, they do still have subjective experiences and can express wishes and preferences.  Should their wish to die be acted upon out of respect for their former capable wishes, or should their current best interests take priority?  This is an ethical dilemma discussed at length in the literature on advance directives in dementia.

Thirdly, patients who retain mental capacity can not only verbally express their suffering but also articulate and explain the meaning of their suffering; this is not so for patients with AD.  Yet research demonstrates that requests for assistance in dying typically are not driven by physical symptoms but rather by forms of existential suffering: loss of dignity, loss of independence, fear of being a burden on others, and so on.  These forms of suffering can only be reliably assessed via interaction with a patient who is able to express and explain it to the care provider.  This conversation cannot take place with someone in the advanced stages of AD, which will make it extremely difficult for the clinician to determine if the person is indeed suffering the way they anticipated they would when they completed the advance request.

The only jurisdiction in the world that allows advance requests for assistance in dying is The Netherlands, and there is something valuable to learn from their experience.  Despite legal permissibility of advance requests for assistance in dying, published evidence reveals that such requests have never been acted upon in the case of people with AD who are no longer mentally capable.  In a 2010 study (5), the 434 Dutch physicians surveyed admitted to a total of five cases in which assistance in dying was performed on patients with AD, but in all five cases the patients were still mentally capable and able to express their wishes.  There has not been one single reported case of assistance in dying being performed on a patient with AD even in the presence of a written advance request.  When asked to explain why they have not complied with advance requests for assistance in dying from AD patients, the physicians confirmed some of the problems described above: 76% reported that it was impossible to determine the moment at which the patient’s circumstances met the criteria in the advance request, and 54% that it is impossible to determine whether a patient with AD experiences his/her condition as unbearable or hopeless suffering (specific criteria under Dutch legislation).  While this is limited evidence, and doesn’t necessarily establish that it never happens, it is quite instructive that clinicians working in the most liberal assistance in dying jurisdiction in the world have a serious problem with acting upon advance requests for assistance in dying in AD patients.

I’m not arguing for the categorical prohibition on advance requests for assistance in dying but I do think we should approach this issue very carefully to ensure appropriate protection for some of our most vulnerable Canadians.  One safeguard that could perhaps address some of the problems described above would be to implement additional eligibility criteria to govern the validity of advance requests.  One example of such a criterion would be to say that advance requests for assistance in dying can only be made for the very end stages of diseases like AD, when a patient becomes non-communicative, bed-bound, unaware of their external environment, and unable to consume enough food and fluids.  In the end-stages of the illness relying on objective assessments of suffering is “safer” and less open to abuse.  Safeguards such as this may be the only way to move forward with advance requests for assistance in dying while balancing the values at the heart of this issue: respecting autonomy, preventing suffering, and protecting the vulnerable.