Physician-Assisted Suicide and Euthanasia in the Intensive Care Unit: A Dialogue on Core Ethical Issues

Abstract

Objective

Many patients are admitted to the intensive care unit at or near the end of their lives. Consequently, the increasingly common debate regarding physician-assisted suicide and euthanasia (PAS/E) holds implications for the practice of critical care medicine. The objective of this manuscript is to explore core ethical issues related to PAS/E from the perspective of healthcare professionals and ethicists on both sides of the debate.

Synthesis

We identified four issues highlighting the key areas of ethical tension central to evaluating PAS/E in medical practice: (1) the benefit or harm of death itself, (2) the relationship between PAS/E and withholding or withdrawing life support, (3) the morality of a physician deliberately causing death, and (4) the management of conscientious objection related to PAS/E in the critical care setting. We present areas of common ground as well as important unresolved differences.

Conclusions

We reached differing positions on the first three core ethical questions and achieved significant agreement on how critical care clinicians should manage conscientious objections related to PAS/E. The alternative positions presented in this paper may serve to promote open and informed dialogue within the critical care community.

INTRODUCTION

In the wake of recent legal and social developments in the United States and Canada, physician-assisted suicide and euthanasia (PAS/E) have become the subject of intense discussion (1, 2). Following symposia at recent critical care meetings, several authors of this paper engaged in a vigorous debate on the ethics of PAS/E and its implications for the practice of critical care medicine. Seeking to maintain collegial co-operation in the face of a potentially divisive issue, we have chosen to write together to present our perspectives on ethical issues in a concise format to generate crosstalk by those with opposing views on PAS/E with the goal of increasing our collective understanding on this very difficult consideration within medicine. Joined by academic ethicists, palliative care clinicians and a jurist, we have identified four core ethical questions central to the debate over PAS/E.

By convention, physician-assisted suicide (PAS) refers to prescription of lethal medication to be voluntarily self-administered by the patient. Euthanasia refers to deliberate, direct causation of death by a physician (3). It is important to distinguish between voluntary euthanasia (VE, which refers to the patient’s specific, consistent and thoroughly considered request), involuntary euthanasia (IVE, in which the patient is killed against his/her will), and non-voluntary euthanasia (NVE, in which the patient neither consents nor objects to euthanasia because of decisional incapacity). NVE has been proposed and subsequently opposed in the intensive care unit (ICU) (4, 5); and it is illegal even in countries such as Belgium and Netherlands that have legalized euthanasia (6). We unanimously oppose NVE and IVE and all subsequent references to euthanasia in this paper signify VE. Physician-assisted suicide and euthanasia (PAS/E) are sometimes referred to as “physician-assisted death” (PAD) (7). By agreement of the authors, we have chosen to use the term PAS/E without wishing to assign any connotation (positive or negative) by this choice.

Although requests for PAS/E may be infrequent in the ICU context, this issue warrants intensivists’ consideration for several reasons. End-of-life care is a core aspect of critical care practice. Some hold that PAS/E is equivalent to widely accepted practices of withholding or withdrawing of life sustaining therapy (WWLST). Euthanasia has been employed to expedite death during WWLST in some jurisdictions (5). Furthermore, euthanasia has been considered to enhance number and quality of organs for donation (8), and intensivists may be involved in such discussions. We unanimously agree that high quality palliative care must be the first priority in the care of all suffering patients near the end of their life. Given that patients most commonly request PAS/E because of concerns over loss of autonomy and fear of dying in pain (9), assurance of excellent palliative care and lack of abandonment during suffering will help to minimize requests for PAS/E. However, such requests may persist even after such assurances and after achieving excellent symptom control.

In this article, we describe major alternative perspectives on four questions central to the ethical analysis of PAS/E in the form of a dialogue between those who favor the legalization of PAS/E and those who oppose it (Table 1). We do not discuss legal considerations or social policy implications, such as whether PAS/E poses a risk to certain vulnerable populations. The questions under discussion highlight key areas of ethical tension. It is our hope that this distillation of our debate into co-written reflections upon these questions may help change thinking for the reader and shape future discussions on PAS/E globally.

Table 1.

Core ethical issues under discussion About Physician-Assisted Suicide/Euthanasia in Critical Care.

Core Ethical Issue	Position 1 (perspectives supporting PAS/E*)	Position 2 (perspectives opposing PAS/E)
Are there patients for whom death is beneficial?	Some patients may benefit from death Quantity of life can be sacrificed in the interest of quality of life Some suffering will only end with the death of the patient	The benefit of death is unknown and unknowable to medicine For death to be beneficial, the condition of being dead must be superior to the condition of remaining alive The medical profession (and the patient) has no empirical knowledge of the condition of being dead (i.e. whether there is life after death or what it is like to be dead) Ending life early precludes potential opportunities for relational and spiritual healing at the end of life
Is physician-assisted suicide/euthanasia morally equivalent to withholding/withdrawing life support?	There is no ethically meaningful difference between PAS/E and WWLST The key considerations in PAS/E and WWLST are the same: respecting patient values and sacrificing quantity of life in the interest of quality of life The overall intent in both PAS/E and WWLST is the same: provide comfort The physician who performs WWLST that results in death has the same moral agency in that death as the physician who performs PAS/E	PAS/E and WWLST are ethically distinct because of differences in intent and mechanism of effect WWLST is not carried out in order to deliberately end the patient’s life, although death is often a foreseen consequence of WWLST The goals of WWLST (comfort, removal of burdensome therapies) are achieved irrespective of whether the patient dies following WWLST, whereas the goals of PAS/E are achieved only through the death of the patient
Is it morally acceptable for physicians to cause death intentionally?	Intentionally causing death of patients may sometimes be morally acceptable PAS/E is distinguished from murder by the presence of consent and a compassionate motivation There is no reason to allow passive treatment plans that shorten life (i.e. WWLST) but not active ones If death is certain and the possibility of suffering is significant, PAS/E can be defended on grounds of harm reduction	Intentionally causing death of patients is not morally acceptable Human persons have an intrinsic and incalculable value that transcends circumstance and preference Respect for persons necessarily entails that we cannot make a ‘somebody’ into a ‘nobody’ The duty to respect preferences stems from the intrinsic value of the person; honoring a preference for death necessarily takes aim at the very basis for respecting the patient’s preferences
What is a reasonable accommodation between the right of patient access to PAS/E and the right of conscientious objection?	Consensus was reached on this issue Conscientious objections should be accommodated without unduly obstructing patient’s access to medical interventions permitted by law Physicians must discuss all legitimate options for treating suffering, and recognize the distinction between restricting their own actions and obstructing the patient’s right of access Special constraints upon conscientious objection apply in the ICU because hospitalized patients often have little or no ability to choose their attending physician In the ICU context, transferring care to an alternate attending physician upon the patient’s request because of conscientious objection does not constitute a referral for PAS/E and does not imply moral culpability if the patient subsequently undergoes PAS/E

^*PAS/E = physician-assisted suicide and euthanasia

1. Are there patients for whom death is beneficial?

Position 1: Some patients may benefit from an intervention to cause death

The value of life is great but not infinite, in our opinion. We accept that we should not prolong life at all costs, particularly if those costs include intolerable pain and suffering according to the patient. We can choose WWLST, even if that would result in patients dying sooner than they otherwise might. We are obligated to treat physical suffering, even if there is a chance that intervention to relieve pain might shorten life. If we accept that life does not need to be prolonged at all costs, and that life can even be shortened (however unintentionally or passively) in the interest of comfort, then we implicitly accept that the value of life is not infinite.

We know that for many dying patients, the motivation for requesting PAS/E may not be physical suffering but ‘loss of control,’ ‘pointless suffering,’ ‘deterioration or loss of dignity,’ and ‘weakness or tiredness’ (9, 10), for which there are no effective medical treatments (though still we advocate attempts at palliating these types of suffering). Death may not be a pleasant experience, but it may be the only way to end these forms of suffering. Accordingly, we find it sometimes justifiable to accelerate a patient’s death deliberately as a means of ending suffering. We accept that positive aspects of being alive are sometimes outweighed by burden of being alive and suffering, and by the value of honoring a patient’s considered wish for death.

Position 2: Benefit from an intervention to cause death is unknown and unknowable to medicine

In the ICU, we sometimes witness severe and protracted physical and psychological suffering. Death may seem a kind of relief, as the patient no longer has physical sensation. On this basis some have argued that death can sometimes be beneficial.

We agree that the value of life per se is not infinite and should not be prolonged at all costs. However, we find it hard to argue confidently that death is itself a benefit (12). Death is beneficial to patients if and only if they are better off dead. However, in determining whether patients are better off dead, we are critically limited by unknowable variables.

First, doctors and patients have no idea what it is like to be dead. Some assume (apart from any evidence) that death is merely the negation of existence or that it is necessarily a better condition than a life of suffering. All of us hold beliefs about the possibility (or impossibility) of life after death and its nature and quality, beliefs to which we cling with varying degrees of tenacity and confidence (11). However, none of us form such beliefs on the basis of empirical evidence—the critical evidence standard in medicine. The afterlife, if it exists, is inaccessible to science. This inaccessibility does not favor those who believe that there is no life after death—absence of evidence is not evidence of absence (12). This lack of evidence precludes any genuine knowledge of the relative benefit of death over life, even a life that seems at the time not worth living.

Second, while the dying process will certainly have unwanted negatives in terms of pain and suffering (for which we advocate aggressive and much improved approaches to palliative care), the dying process can also be a time of existential and spiritual healing through growth in personal and relational wholeness as well as individual learning for patients, their loved ones, and those caring for them (13–15). Ending the patient’s life before the natural dying process runs its course potentially limits the opportunities for such healing at the end of life.

There is clearly a tension between the fear and burden of dying and the fear and burden of living with suffering, but this tension should not be resolved by either extreme (deliberately ending life or prolonging it endlessly on life support), particularly when those decisions are made on the basis of unreliable calculations of benefit and harm.

2. Is PAS/E morally equivalent to withholding or withdrawing life sustaining therapy (WWLST)?

WWLST is generally held to be morally acceptable (1, 16–18). However, it is controversial whether PAS/E is morally equivalent to WWLST.

Position 1: There is no ethically significant distinction between PAS/E and WWLST

In both PAS/E and WWLST, the physician performs an act with the (usual) proximate consequence of the death of the patient. If a physician is allowed to withdraw a therapy (WWLST) where the benefit no longer outweighs the harm (consistent with the patient’s values) and the patient is allowed to die, then a physician should be allowed to provide an active therapy (PAS/E) for the same purpose. The key considerations in the discussion—namely the patient’s values and the idea of trading quantity of life for quality of life—are not affected by whether the death of the patient is achieved actively or passively.

The usual ethical distinction that is drawn between PAS/E and WWLST is the intent of the physician: deliberately ending life in the former while allowing the patient to die in the latter (19). But those of us in favor of PAS/E disagree with this analysis; we feel that intent of the physician is not always easy to define (20), and many intensivists report an explicit intent to shorten life when they perform WWLST (21). In both PAS/E and WWLST, the overall intent of the physician is to provide comfort; only the instrumental act differs. So we are left to compare the instrumental acts: deliberately ending life vs. deliberately ending life support. Are they morally equivalent? Importantly, we do recognize a difference between the two, but we do not feel that the difference is large enough to justify prohibiting one and permitting the other. WWLST does not always result in death, but death is so likely after withdrawing life support that a physician must accept some degree of moral agency when it occurs (18). Indeed some intensivists may intend to hasten death by WWLST (16). Physicians are responsible, both legally and morally, for all the foreseeable consequences of the actions that they perform, not just those that are intended (22). Certainly, if a person maliciously decided to turn off life support (as in IVE or NVE), he could not defend himself against a murder charge by saying that he merely allowed the victim to die, and that he did not intend to end his/her life (20). The medical team has moral agency and is responsible (along with the patient or family) for determining the timing, method, and rationale for WWLST. This responsibility is no different for a physician and patient/family who choose to determine the timing, method, and rationale for PAS/E.

Position 2. PAS/E and WWLST are ethically distinct because of differences in intent

PAS/E and WWLST are sometimes mistakenly differentiated as active vs. passive means of ending life. However, there is no ethically relevant distinction between active and passive means of deliberately causing death—either could be employed in euthanasia (23). Rather, WWLST is ethically distinct from PAS/E because of critical differences in intention, causation and other factors (24). First, while PAS/E necessarily requires intent to cause death, WWLST does not. Life support is not withheld or withdrawn in order to end the patient’s life. Rather, the intention of WWLST is to respect the patient’s decision that any given intervention is overly burdensome or disproportionate in his/her specific life circumstances and should thus be avoided in order to minimize suffering and maximize dignity (17). Death is a foreseen but unintended consequence of WWLST. While a minority of intensivists may deliberately aim to hasten death by WWLST (16), such intentions are neither necessary nor intrinsic to the practice. Death is not even a necessary consequence of WWLST: a small percentage of patients do not die in a course temporally related to WWLST (18, 25).

Second, death is not the mechanism by which goals of WWLST are achieved. Patients do not need to die in order to respect their wishes for discontinuation of unwanted and often burdensome interventions or to ensure that patient dignity and comfort are maximized during and following WWLST. Thus WWLST can be deemed successful irrespective of whether or when the patient subsequently dies.

The ethical significance of the distinction between intending and merely foreseeing consequences has been ably defended (26) and is widely recognized in end-of-life care decision-making (24, 27). Given this distinction, we argue that WWLST is categorically different from PAS/E, and we may embrace the former as an integral part of benevolent care while firmly acknowledging the latter as a breach of the patient-physician covenant (28).

Question 3. Is it morally acceptable for physicians to cause death intentionally?

Position 1. Intentionally causing death is morally acceptable

There are many important factors at play (aside from intent and premeditation) when one person is involved in the death of another. Supporters of PAS/E identify presence of patient consent as the key factor that distinguishes PAS/E from murder. Consent changes the fundamental nature of an act, which is why we pay people to perform surgery, but send them to jail for a knife assault. In contrast, we argue that patient consent renders PAS/E permissible. We already respect the right of patients to consent to treatment plans that will effectively end their lives (i.e. WWLST). We in favor of PAS/E cannot see any sufficient rationale for allowing patients to consent only to passive plans of treatment towards ending their life and not active plans of treatment towards ending their life, when both forms of involvements are intended to minimize suffering and convey respect for patient as a person (25).

There is also a harm-reduction argument to be made. In some cases, we know that a patient is certain to die of an incurable and progressive illness, and that there is a significant possibility that this patient will develop a terminal event that would cause suffering. We can see no adequate reason why that patient should not be allowed to die at a time of their own choosing, rather than being obliged to wait for the unpredictable terminal event with a possibility of suffering despite adequate/maximal palliative care. True, that person would be potentially sacrificing some of his/her quantity of life in the interest of quality, but we accept that patients are in the best position to find balance between their own desired quantity and quality of life and the degree of uncertainty that they can tolerate.

Position 2. Intentionally causing death is morally unacceptable

In the context of the provision of healthcare, and in the treatment of patients at the end of life, we hold that it is unethical for a healthcare professional to intentionally end a patient’s life because such an act runs counter to what we believe to be the moral foundation of medicine: the incalculable and intrinsic worth of the human person (26, 27). Medicine derives its moral greatness from its respect for the value of each individual, a value that transcends circumstance or preference. This value derives from a person’s status as a sentient and rational being capable of morally significant choices.

The wrongness of deliberately causing death lies in the value of the person himself. The deliberate act of ending a life makes a “somebody” into a “nobody” (29). Permitting killing out of “respect for preferences” renders the value of one’s person contingent upon those preferences. Importantly, the value of a person is not diminished even when that person thinks that his/her life is unworthy of living. A willingness to deliberately cause a person to become a non-person is deeply subversive to medical ethics because it renders the value of a person a matter of mere judgment. Put simply, a person is necessarily more important than her preferences, for her intrinsic value is the foundation of respect for her preferences.

Patients and physicians enter into a moral covenant grounded in beneficence, which Sir William Osler recognized as the distinguishing feature of our profession in a lecture to the Canadian Medical Association in 1902 (28, 30). The cornerstone of patients’ trust in medicine is the absolute certainty that physicians will always care for them out of respect for their intrinsic worth. Actively involving physicians in the act of killing irrevocably breaks the certainty of trust long held as the core of the patient-physician covenant. We conclude that it is unethical intentionally to cause death because it constitutes a profound violation of the intrinsic and incalculable worth of a person.

4. What is a reasonable accommodation between the right of patient access to PAS/E and the right of conscientious objection?

“The physician-patient relationship, like any ethical relationship, is a reciprocal relationship. In the justifiable concern for patient autonomy, we must remember that the physician is a moral agent as well as the patient. When the two are in conflict, the patient’s wish does not automatically trump the physician’s.” (31)

The American Thoracic Society recently published an official policy statement on conscientious objection in critical care (32). The policy statement’s recommendations are summarized in Table 2. Here we present our shared position on conscientious objection with respect to PAS/E in light of this policy statement.

Table 2.

Policy recommendations for managing conscientious objections in the intensive care unit as recommended in the American Thoracic Society 2015 Consensus Policy Statement (32)

Recommendation 1: Conscientious objections (COs) in ICUs should be managed through institutional mechanisms rather than ad hoc by clinicians. Healthcare institutions should develop and implement CO policies that encourage prospective management of foreseeable Cos and that provide a clear process to manage unanticipated COs.

Recommendation 2: Institutions should accommodate COs in the ICU if the following criteria are met: The accommodation will not impede a patient’s or surrogate’s timely access to medical services or information; The accommodation will not create excessive hardships for other clinicians or the institution; The CO is not based on invidious discrimination.

Recommendation 3: A clinician’s CO to providing potentially inappropriate or futile medical services should not be considered sufficient justification to unilaterally forgo the treatment against the objections of the patient or surrogate. Clinicians should instead use a fair process-based mechanism to resolve such disputes. A clinician may use the institutional CO management process to request a personal exemption from providing the medical service.

Recommendation 4: Institutions should promote open moral dialogue, advance measures to minimize moral distress, and generally foster a culture that respects diverse values in the critical care setting.

We unanimously agree that accommodation for the matter of conscience is necessary. Patients should respect the fact that PAS/E is an ethically controversial topic, and they should expect many physicians to be unwilling to provide it upon request. At the same time, they should be assured that discussions will be directed towards more aggressive palliation of pain, anxiety, and dyspnea, for example, even if such therapies suppress respiratory drive. All participants in the patient-physician covenant must recognize that conscientious objection is an instrumental means of promoting the integrity and quality of medical care (33).

Physicians should view requests for PAS/E as an opportunity to explore the suffering and fears of the patient, and discuss the legitimate options for addressing these issues. They should not prompt either an immediate granting of the request or an immediate transfer of care. Special constraints upon conscientious objection apply in the ICU because hospitalized patients often have little or no ability to choose their attending physician; a conscientious objection could therefore seriously obstruct access to PAS/E. As outlined in the ATS policy statement, we suggest that physicians who object to providing PAS/E in jurisdictions in which these practices are legal transfer the care of such patients to colleagues who are willing to consider referring or providing PAS/E. Such a transfer of care does not constitute a referral or render physicians morally culpable for ensuing events; rather it merely provides the patient with their right to an alternate attending physician. White and Brody put it well in saying that such an approach serves as a “shield” to protect physicians seeking to act within their core moral beliefs and never a “sword” to force beliefs onto patients (33).

Conclusions

As the debate about legalizing PAS/E continues unabated around the world, intensivists will be caught up with this important medical, legal, and ethical issue. The debate revolves around four key areas of ethical tension as summarized above. Despite “opposing” positions on three of these issues, we were able to reach unanimous agreement on the handling of conscientious objection. We hope that our discussion enables readers to reflect critically on their own position on PAS/E in order to care for critically ill patients and their families with ever greater compassion and humanity, and to discuss these issues among colleagues with clarity and respect.