After an excellent introduction to Medical Assistance in Dying, and its potential impact upon the Healthcare System, Dr. Francis Bakewell takes a further look at potential implications and ramifications within the Emergency Department (ED).
Last week, after much review and attempts at amendment by the Senate, the federal government’s legislation on Medical Assistance in Dying (MAID) received royal assent and became law. Bill C14 allows for aid in dying, either through prescription of lethal medication or through administration of medication, to be provided to persons over the age of 18 who meet all of the following criteria1:
- Have a serious and incurable illness, disease, or disability
- Are in an advanced state of irreversible decline in capability
- Have enduring physical or psychological suffering as a result of their illness and decline that is intolerable to them and cannot be relieved under conditions they find acceptable
- Have a reasonably foreseeable death (though a prognosis as to specific length of time remaining does not necessarily need to have been made)
While the debate around medical assistance in dying (MAID) has dominated the national discourse over the past two years, it has received comparatively little attention in the world of emergency medicine – and understandably so. Most emergency physicians are unlikely to ever provide MAID themselves, unless they specifically seek involvement in clinical assessments outside of their day-to-day work in the ED.
However, as with any innovation, especially one that represents such a fundamental change in both the scope and philosophy of medicine, the effects of MAID will almost certainly be felt both upstream and downstream of the service itself. Canadian emergency physicians deal with chronic suffering, death and suicide on a daily basis – it would be foolish to think that we won’t face MAID in some form as well.
There are three main areas where MAID is likely to affect the practice of emergency medicine:
- In communication about end-of-life issues with patients and their families
- In dealing with potential complications of MAID undertaken outside of hospital settings
- In the implications it might have for treating other suicide attempts and reconciling the traditional ED approach to suicide and incapacity with an increasing social and medical acceptance of ‘capable suicide’.
Conversations about end-of-life care take place regularly in the ED and patients may now bring up MAID as part of that discussion. Patients with a terminal illness express a desire for a hastened death not infrequently – 17% of terminally ill cancer patients in one study were found to have a high desire2. Other studies report that over half of patients with AIDS and ALS would consider MAID, while nearly all of the latter group at least want the option available3,4. Smaller numbers of patients, however, will actually discuss or request MAID. In Oregon, 1 in 6 terminally ill patients will discuss MAID with their families, 1 in 50 with their physician and only 1 in 500 will directly access it5.
It’s important to note that patients who request MAID do so for a variety of reasons – uncontrolled pain is rarely the only or even the primary reason underlying a request. More often, patients cite their primary motivation as being a loss of control, a loss of dignity and a desire to not be dependent on others for personal care6.
While there is no data on how frequently patients discuss this wish in the ED, it isn’t hard to imagine patients bringing it up during end-of-life conversations – in part, perhaps, because our patients are often confused by the end-of-life terminology we use. In one study, ED patients were asked to define 10 common end-of-life terms and were then scored for correctness. While 56% were able to correctly define DNR, only 9% and 7% were able to correctly define physician-assisted suicide and euthanasia – the two most poorly defined terms7.
One can also imagine ED patients with serious pre-existing illness bringing up MAID in the context of worsening symptoms, a decline in function, a perceived ‘failure’ of their palliative measures, upon finding out that their illness has progressed or been complicated by some new diagnosis.
When our patients inevitably do broach the topic of assisted dying, we might be hesitant to discuss it with them at all. We may fear the strong emotions involved, the potential for psychological harm to the patient, and quite frankly, that it will just take too much time to address8. We may think that discussing such psychosocial factors is simply not our responsibility9, or (and perhaps quite legitimately) we might suspect that a patient’s desire to die in the ED has been triggered by whatever setback has landed them there in the first place and like many wishes for a hastened death, it is only temporary10.
As understandable as these innate responses may be, we ought to try and avoid them. Simply blocking or dismissing a patient’s request in the ED, even in the form of blanket reassurance and expressions of hope, must be considered an unacceptable abandonment of the patient at a time when they may be at their most vulnerable. We must find a genuine way to address our patient’s concerns.
What should we do when faced with a patient’s request for MAID during an ED visit?
There exist a variety of suggested approaches for responding to a patient’s desire to die11,12. Though most have been developed for use in the clinic and palliative care setting, the basic elements are translatable to even brief discussions in the ED:
- Be aware of your own initial response to a patient’s request: be open, listen actively and acknowledge the patient’s feelings and concerns, even if you have strong opinions on the issue (one way or the other).
- Attempt to understand exactly what the patient means when they express a wish for a hastened death: is it a passive wish that they will simply die sooner rather than later, or is it an explicit wish to take active measures to end their life?
- What does the patient understand about their illness thus far and their prognosis? What do they see as their options at the end-of-life? Are they in a position to make a capable, informed choice?
- Are there factors contributing to this desire that ought to be otherwise addressed (e.g. inadequate symptom management, poor social supports, depression)? Are there steps you can initiate in the ED today to help address some of these factors?
- As is so often the case, a crucial step in the ED will be to ask why today? Is the patient initiating a discussion in the ED because they are distraught by a sudden change in their condition? Or have they been consistently expressing the desire for hastened death for some time? Do they fear that they cannot talk about it with their primary provider, or have they already tried but been dismissed? As with so many seemingly ‘non-emergent’ concerns, attempting to get to the root of why an issue has come up today will help determine the appropriate course of action – does the patient need more information about their condition and prognosis, more aggressive symptom management or counselling, or referral to a provider who can fully assess their desire and eligibility for MAID?
It needs to be stressed that it will often be impossible and likely inappropriate, to attempt to achieve a full understanding of a patient’s desire to die in the ED setting. Not only is there not enough time, but the lack of a therapeutic relationship and the fact that we will not actually be the physician proposing a treatment plan means that our goals of discussion need to be adjusted. We should not expect to fully assess a patient’s capacity, nor to obtain their consent for medically assisted death in the ED. Rather, our focus must be on assessing their capacity and consent to be referred for an appropriate assessment. Undergoing an assessment for MAID obviously does not mean that the patient will be approved for MAID, and thus a referral requires a much less intensive level of discussion than obtaining consent for the act itself. Ensuring that the patient understands their illness, prognosis, options and attempting to first address any other remediable factors is likely a sufficient approach in the ED prior to a referral. Using this approach will likely mean sending some patients for assessment who will then be deemed ineligible, or whom will change their mind. However, as is often the case in the ED, we ought to prioritize sensitivity over specificity – as gatekeepers to the medical system, we should not turn away those who may have nowhere else to turn.
In settings that do not yet have an established local or institutional referral process, there may be other ways patients or clinicians can access more information. For example, in Ontario the Ministry of Health and Long-Term Care has established a toll-free referral support line to help physicians arrange for consultations for patients seeking MAID (1 844 243 5880).
As discussed in Part 1 of this series, the actual process of MAID can take several forms, including patients self-administering a lethal dose of medication orally.
The Special Joint Parliamentary Committee that advised the federal government on MAID advocated that patients should have the ability to receive assistance outside of a hospital setting13 and Bill C-14 explicitly acknowledges that self-administration of medication should be a viable option1.
Self-administration may not become part of the assisted dying landscape in Canada for a while, however, as the most commonly used oral medications for assisted death in regions like Oregon and Washington (secobarbital and pentobarbital) are currently not available in Canada – though that may change14.
And while those drugs are the most common options, they’re not a patient’s only option. According to Death with Dignity, the increased cost of secobarbital and pentobarbital in the US has led some physicians in Washington State to prescribe their own mixture of phenobarbital, chloral hydrate and morphine – all of which are available in Canada .
If patients do end up self-administering medications in Canada, especially if they do so in the absence of medical supervision, the available evidence suggests that some complications will occur.
Data from Oregon’s experience with assisted death thus far (1998-2015) suggests that of 991 total patients who self-administered medication, about half (486, or 53.9%) did not have a medical provider present at the time of their death15. There were 27 complications (mostly regurgitation) in the 556 cases where the presence or absence of complications was recorded, for an overall complication rate of 4.9%. However this number is unreliable, as there is no data available on complications for the other 44% of cases. There have been 6 reported cases in the past 17 years where patients regained consciousness after ingesting medication. Complications aside, of the 536 cases where the data is available, there is a reported time range from ingestion to death of 1 minute to 104 hours, with a median time of 25 minutes.
Data from the Netherlands in the 1990s suggests similar issues16. Of 114 cases of assisted death using self-administered oral medications, 10% had some technical problem (most commonly difficulty was ingesting the drug), 7% had complications such as myoclonus, cyanosis, vomiting, or extreme gasping, and 15% had a longer than expected time to death or never actually became comatose. The median time to death was 30 minutes, with a range of 1 minute to 14 days. In 19% of cases, the time to death was longer than the physician had expected. In 18% of cases a physician was present and decided to administer a lethal medication intravenously for any of the above reasons.
The occurrence of complications and delayed deaths such as these is not necessarily a reason to restrict self-administration, but patients and providers need to be aware of the risks and have a plan in place to deal with them should they arise.
While it’s unknown just how many of the above cases sought or received further emergency medical care, it’s reasonable to assume it may happen. Patients with a terminal illness and a documented wish to not be resuscitated are already often brought by their family members to the ED in extremis and understandably so – it is difficult to watch a loved one apparently suffering and not call 911. We can expect this may also happen in some of these cases of regurgitation or prolonged time to death. What should be the appropriate response of the ED physician if they receive a patch call from paramedics, or if they see a patient in their ED with an incomplete assisted death?
As others have pointed out17, there are essentially three options available to the ED physician:
- Fully resuscitate the patient
- This is the option that will likely be most familiar to ED physicians, being the treatment course pursued in most cases of patients who have intentionally ingested a medication to end their lives. Obviously, however, it is unlikely to be the ethical choice in most cases of MAID. Patients who come in after an incomplete assisted suicide will have already undergone an assessment of their capacity and consented to a hastened death. So long as they have evidence of this and are seeking care simply due to a complication or a prolonged death, then resuscitating them would represent a clear violation of their autonomy and would arguably consist of medical assault. Furthermore, given that many of these patients may already have an underlying terminal illness and significant medical conditions, the likelihood of a successful resuscitation (and therefore the balance between beneficence and maleficence) would be seriously in question.
- The more difficult case may be where it is unclear whether or not a patient has changed their mind. This is particularly conceivable in the form of a loved one or substitute decision maker (SDM) bringing in an unconscious or minimally conscious patient and asking for them to be resuscitated, insisting that they changed their mind after ingesting a medication at home. The ED physician may be torn between wondering if it is in fact the SDM who is having second thoughts, while also not wanting to withhold treatment from a patient who may genuinely have changed their mind. Organizational policies on resolving end-of-life care disputes, while imperfect, may offer some guidance. For example, the College of Physicians and Surgeons of Ontario’s policy on Planning for and Providing Quality End-of-Life Care, recently amended in May 2016, suggests that in cases where a physician disagrees with an SDM about providing full resuscitative care to a patient, and pending further attempts at conflict resolution, they should provide resuscitation – “unless the patient’s condition will prevent the intended physiologic goals of CPR (i.e., providing oxygenated blood flow to the heart and brain) from being achieved18.”In some cases, a massive barbiturate overdose in a patient with a pre-existing terminal illness may well meet the criteria for withholding resuscitation that this amendment sets out. When an ED attempt to resolve a dispute fails, the ED physician may be expected by professional policies to attempt resuscitation, although this certainly raises concerns about the ability of SDMs to potentially override a patient’s final wishes.
- Palliate the patient
- A second, and likely more often appropriate option available to ED physicians when faced with a patient with an incomplete assisted death would be to pursue palliative measures, such as symptom management and possibly further sedation. This would seem to be in keeping with a MAID patient’s presumed wishes (i.e. the avoidance of pain and suffering, without any further prolongation in life) and is analogous to the care already provided in EDs to many other patients who come in with an expected death but who are seeking symptom control.
- While it may seem an ideal option, one could argue, however, that this approach doesn’t go far enough in following a patient’s wishes. Patients who receive MAID will presumably have already discussed palliative options at end of life and rejected them; they may have sought MAID precisely because a prolonged dying process or coma was unacceptable to them. Furthermore, if the patient is already in a coma with a slowly deteriorating cardiovascular state then the palliative approach may not involve doing very much at all, other than offering reassurance to loved ones.
- Still, except in cases where there is a significant doubt or dispute about a patient’s wishes (see above), this approach will usually be a reasonable option, as it at least goes partway to respecting a patient’s wishes and represents a type of care most emergency physicians should already be familiar with providing.
- Actively hasten the patient’s death
- The final option involves the ED physician administering a lethal medication parenterally, in effect converting an assisted suicide to euthanasia.
- This option may seem like that which most closely honours an incomplete MAID patient’s autonomy and wishes, by hastening a quick and painless death. However, there are several important reasons why that may not be the case and why this course of action is unlikely to ever be the appropriate one.
- First, a patient may have elected to self-administer a lethal medication at home precisely because they did not want to die in a hospital with an intravenous line in their arm. For those patients who view MAID as a means to re-assert control at the end of their life, this option wrests that control back into the medical realm once again. Second, while the patient may have consented to an assisted death, they likely did not consent to have that assistance provided by an ED physician – while it might be in keeping with their general wishes, to hasten their death in the ED would be to undertake an irreversible and arguably non-emergent medical act without having gone through any sort of consent process – something that both physicians and patients should understandably balk at.
- It is conceivable, and perhaps advisable, that as part of the initial consent process for self-administered MAID a patient should also be counselled on and provide consent for the actions to be undertaken in the event of a complication or delayed time to death. However, it is unclear if a patient’s request for parenteral intervention in this setting would constitute advanced consent for euthanasia, which is not allowed under current legislation. And even if such an advance request was allowed, ED physicians might still be uncomfortable, having never been part of the consent process in the first place. In such cases, consulting the initial attending physician to come and administer a lethal medication in the ED would likely be the more appropriate action.
Regardless of which of these options is the right one for any given case, one thing should be stressed upstream of the ED: physicians who are providing MAID to patients in the form of a prescription for self-administration must discuss with them the potential, albeit unlikely, risk of these complications and delays. They must establish (and document) a plan with patients and their loved ones for what to do if a complication occurs, and how to avoid a trip to the ED.
The final potential impact of MAID may be on a much more common occurrence in the ED: patients who have not sought a medically assisted death, but rather present to the ED after having simply attempted suicide by themselves. While ED physicians have a great deal of experience in treating these patients, the growing acceptance of MAID may raise questions about how we treat this population as well and whether we are being ethically consistent. What’s the difference between a terminally ill patient who has sought MAID and a terminally ill patient who has tried to end their life on their own?
Generally speaking, the ED approach to patients who have tried to harm themselves is to fully resuscitate them, under the assumption that their actions were influenced by mental illness and that they likely lack the capacity to make autonomous choices19.
This approach is so well established that instances of withholding resuscitation from patients with an incomplete suicide can become well-known case reports and the source of much debate. One such case is that of Kerrie Wooltorton, a 27-year-old Englishwoman with a history of mental illness who presented to the ED in 2007 after a toxic alcohol ingestion, carrying a set of advance directives requesting only palliative care – and who received it20.
The majority of the time, the assumption that mental illness underlies an attempt at self-harm is likely correct – post-mortem analyses find that 90% of suicides occur in the setting of concomitant depression, substance abuse, or psychosis21.
However, it would seem that not all suicides are the result of mental illness, or at least mental illness alone. For example, age and medical illness are known to also be significant risk factors for suicide. Men over the age of 85 in particular have one of the highest rates of suicide in Canada22, and studies have shown that, while depression is still the strongest risk factor for suicide, COPD, cancer, spinal disorders and strokes are all independent risk factors as well23. When surveyed prior to the introduction of the Death with Dignity Act, 58% of Oregon emergency physicians reported having seen incomplete suicides in terminally ill patients in the ED, with 19% of physicians having seen at least 5 such cases24.
The increasing social acceptance and support for MAID clearly suggests that, in the eyes of the public at least, suicide can sometimes be a capable, rational choice – to escape intolerable suffering, reassert control and retain a sense of dignity and independence in death. As with other medical decisions, its validity as an autonomous choice comes from it being a free and informed decision made by a patient who can understand the relevant information and appreciate the consequences of their action – not from it simply being condoned by a physician. With MAID, assessment by multiple physicians is a procedural safeguard to ensure a patient is capable of deciding to end their life with assistance, but the assessment does not in and of itself make the decision valid. A capable patient might conceivably make the same decision on their own and try to end their lives by their own hand.
The issue for the ED physician lies in assessing the capacity involved of a patient who presents with an incomplete suicide and deciding when to treat a patient against their wishes, and when to respect a patient’s request to withhold treatment. Obviously, with the time sensitive nature of resuscitation and the often information-poor context of the ED, this presents some difficulties. While absolute guidelines are impossible, some general principles might be considered:
- In emergency situations where a patient’s capacity cannot be fully assessed, the right course of action is to err on the side of prolonging life. Even those who advocate for a greater understanding and acceptance of rational suicide agree that this principle is paramount25. This approach of course means that some fully capable patients who made an autonomous choice to die will be resuscitated against their will, but this is generally regarded as a lesser evil than potentially not resuscitating a patient who tried to kill themselves impulsively or as a result of mental illness.
- However, it must be stressed that attempts at self-harm do not necessarily entail incapacity. This is becoming more apparent as MAID gains wider acceptance. While perhaps uncommon, a patient presenting to the ED with an incomplete suicide may have been capable of making that decision and may be still capable of making the decision to refuse further care. Every patient who presents with self-harm should still be assessed for the capacity to make decisions about their care.
- Likewise, attempts at self-harm do not by themselves demand full resuscitative measures, simply because they are cases of self-harm. This final point is perhaps more controversial. While some have argued that attempts at self-harm should nullify any pre-existing advanced directives26, their rationales (that directives should only apply for deteriorating ‘natural’ illness, that patients likely did not intend them to apply in cases of self-injury, or that the AD may itself reflect suicidal thinking) are unconvincing. Patients who refuse resuscitation are likely more interested in the outcomes of resuscitation than in why they need it in the first place, and while an advanced directive signed yesterday may indeed reflect suicidal thinking, one from a year ago more likely reflects a pre-existing firm and consistent desire to avoid the various traumas of aggressive resuscitation. The principles of beneficence (likelihood of benefit) and non-maleficence (likelihood of harm) that guide decisions around resuscitation in patients with significant co-morbidities should remain the same, regardless of whether an acute injury or illness is precipitated by the individual or not.If you already wouldn’t recommend intubation or chest compressions for a patient with metastatic cancer who presents with signs of multi-organ failure from an acute deterioration, then you shouldn’t reflexively do so simply because the deterioration is the result of self-injury (all other things being equal).
While its exact impact remains to be seen, the legalization of MAID in Canada is likely to have effects throughout medicine, including in the emergency department. We may see it come up in end-of-life discussions, in the care of patients with potential complications, or simply in deciding how to treat a patient who comes in after an attempted suicide. Direct experience with MAID in the ED may turn out to be uncommon, but it is precisely in preparing for the uncommon but complicated that the practice of emergency medicine distinguishes itself. We’d be well served to plan for it now.