Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

10.1 Introduction

Assisted suicide and voluntary euthanasia deviate from ‘normal’ death in two crucial respects. First, they are instances of self-killing, practices in which a person willfully brings about her own death. Second, they involve soliciting the aid of others to bring about one’s death. The second category encompasses the first, for in soliciting another’s aid to bring about one’s own death, a person brings about her own death even if, as is arguably the case in voluntary active euthanasia, that person’s death is caused by another person’s acts instead of by her own. However, as the following example shows, it is possible for a suicidal person to involve others in her self-killing without soliciting their aid:

R. has been suicidal for nearly all of her adult life. Now in her late forties, R. was diagnosed with manic depression in her early twenties. She has been subject to a number of therapies over the past quarter century, including psychotropic drugs and ongoing counseling. Twice her symptoms have become severe enough to require hospitalization. R. has engaged in three previous suicide attempts. Two of these involved her taking large doses of over-the-counter painkillers, the third a clumsy attempt at cutting her throat.

R.’s desire to end her life of course ebbs and flows, but it is recurrent. When in her suicidal frame of mind, R. is not the slightest bit ambivalent about wanting to end her life. But one problem has been her choice of method. R. does not live in a jurisdiction where physician-assisted suicide is legally available, and even if she did, she would probably not qualify because her medical condition is neither terminal nor ‘futile’. She finds her past suicide attempts a bit embarrassing: Only a coward uses methods known to be so ineffective. R. knows that guns are likely to be the most effective method, but she does not know the first thing about guns or how to get ahold of one—and even if she did, her history of hospitalization for mental illness would probably serve as a legal obstacle to obtaining a gun permit. R. is also reluctant to use violent methods or methods that would inflict physical trauma on her body because she does not want to be found in a bloody or brutalized state by her family. R. is thus in a quandary: Many of the most effective suicide methods are either unavailable or unattractive, but readily available methods are unreliable at best. Furthermore, she does not want to upset anyone through her suicide. She desires, above all else, a ‘normal’ death.

R.’s sister invites her to housesit while the sister and her family enjoy a vacation. R. recalls that her sister recently had the powerful painkiller oxycodone prescribed to her after major back surgery. Again feeling like life is futile, R. rummages through the medicine chest and finds the bottle of oxycodone and swallows ten times the dose suggested on the bottle. R. then calls a taxicab and directs the driver to deliver her to the nearest emergency room.

Within minutes after her arrival, R. collapses, unconscious. Hospital staff witness her collapse, find that her breathing has stopped, and ready her for emergency interventions. R. is fitted with an artificial ventilator. While her condition remains critical and her prognosis uncertain, R. can be kept alive indefinitely with the ventilator’s assistance. R.’s sister, listed on R.’s admission form as her medical contact, arrives at the hospital. One hour later thereafter, medical personnel find in R.’s purse an advance directive she completed two years prior. R.’s advance directive states that she does not wish to be revived or resuscitated if her breathing or cardiovascular activity stops, nor does she wish to receive artificial lifesaving measures. The attending physicians honor the wishes stated in her advance directive and remove R.’s ventilator. R. dies within five hours of consuming the overdose of pain medication.

Assuming that R. completed the advance directive, with its ‘do not resuscitate’ clause, in anticipation of taking the drug overdose, then R. seemed to have knowingly and willfully implemented a plan that culminated in a death she endorsed. She seems then to have ended her life via suicide. But the example of R. is an atypical instance of suicide, a case of what I will call medically enabled suicide. Medically enabled suicides have four distinctive features:

  1. 1.

    They are instigated by actions of a suicidal individual, actions she intends to result in a physiological condition that, absent lifesaving medical interventions, would be otherwise fatal to that individual. In R.’s case, the painkiller overdose was aimed at putting her in a physiological state that is fatal. However, this state can nearly always be treated so as to prevent the patients’ death.

  2. 2.

    These suicides are ‘completed’ due to medical personnel acting in accordance with recognized legal or ethical protocols requiring the withholding or withdrawal of care from patients (e.g., following an approved advance directive). R. dies in part because medical personnel honor an advance directive proscribing her receiving lifesaving care. Medically enabled suicide could also occur when competent patients refuse lifesaving care ‘at the bedside’.

  3. 3.

    The suicidal individual acts purposefully to ensure that medical personnel will act on these protocols. R. filled our her advance directive recognizing that its provisions, if honored, would make it the case that her drug overdose would lead to her death.

  4. 4.

    These suicides do not involve medical personnel providing aid in dying in the standard sense, either through (a) active voluntary euthanasia, or (b) assistance by means of prescriptions, etc. Medically enabled suicides involve the participation of medical personnel, but not their “assistance,” as that is standardly understood. Medical personnel, acting on protocols requiring the withholding or withdrawal of care, participate in patients’ suicidal plans without consenting to participating in those plans insofar as those plans aim to end the patient’s life. The patient’s underlying physiological condition ends up killing her, rather than (say) the subsequent administration of a lethal medication.

The frequency of medically enabled suicides cannot be estimated with any accuracy. In the United States, nearly half a million people are treated in emergency rooms for self-inflicted injuries or harm each year (US Centers for Disease Control 2012). Recent surveys indicate that about one-third of adults have a living will or other provision for their end-of-life care. About half of all Americans indicate that there are some circumstances in which they would refuse or cease treatments even if this would result in their deaths (Pew Research Forum 2013). Of course, no reasonable inferences can be made from this data about the prevalence of medically enabled suicides. But the number of self-inflicted injuries, combined with the fact that many adults have concluded that they would sometimes prefer not to receive life sustaining treatments, suggests that medically enabled suicides, while rare, do occur. In a survey of English clinicians treating self-poisoning (reported in Kapur et al. 2010), an advance directive was mentioned in 2.5 % of cases. The case of Kerrie Wooltorton is a real life example where an advance directive came into play in the treatment of suicidal injuries (Dresser 2010; Szawarski 2013).

Regardless of how common medically enabled suicides are, they raise compelling questions at the interface of medicine and individual choice. After first clarifying the concept of medically enabled suicide and exploring some of the reasons why it might be attractive to suicidal individuals, I then investigate two apparent dilemmas that medically enabled suicides raise for medical care providers. The first alleges that medical care providers may not contribute to harming their patients, and so they may not contribute to their patients’ suicides. The second alleges that if care providers, as a matter of personal conscience, believe that suicide is wrong, then they may not be compelled to contribute to their patient’s acting wrongly by assenting to the wishes of a patient pursuing medically enabled suicide. Both dilemmas arise from the fact that while medical personnel are bound by widely accepted precepts of medical ethics to honor the competent wishes of their patients, medically enabled suicides entangle them in their patients’ suicidal plans in ways that result in their contributing to those suicides. I conclude that neither dilemma should be resolved in the direction of medical personnel having the right to refrain from involvement in medically enabled suicides. Thus, while we may find medically enabled suicide distasteful or exploitative, a strong case cannot be made that medical personnel refusing to involve themselves in such suicides is ethically permissible.

10.2 The Concept of Medically-Enabled Suicide

Before considering the ethical obligations of medical caregivers as regards medically enabled suicides, we must first answer worries to the effect that medically enabled suicide is a conceptually suspect category.

First, are the actions of R. suicide at all? One reason for doubt is that in order for R. to die, others must act in specific ways. In this case, R. will not die unless medical personnel withdraw the ventilator in accordance with the wishes R. stated in her advance directive. Suicide is self-killing. But an act is not precluded from being suicidal on the grounds that should death occur, an agent besides the suicidal individual had to act in a specified way. In the phenomenon known as ‘suicide by cop’, an individual commits a crime in the hope that police officers will be provoked to kill her. Here the individual acts with the intention of bringing about her own death, but is not the cause of her own death (or at least is not the immediate cause of her own death.) Likewise, in medically enabled suicide, an individual initiates a plan of action whose success hinges upon another’s act. Both acts are necessary conditions of her suicide succeeding in bringing about her death, but the fact that her plan of action may not be sufficient on its own to bring about her death does not imply that her action is not suicidal. Furthermore, on standard definitions, suicide is intentional self-killing. But acting with an intention that one bring about one’s own death need not entail that one’s acts are the cause of one’s own death (Cholbi 2011a, 21–26). In cases such as R., an individual wishes to be dead, initiates a plan of action a success condition of which is (from her point of view) her own death, and dies as a result. Her actions are most coherently described as trying to die by virtue of her rational endorsement of dying as a means to her presumed end, relieving her own suffering or anguish. Medically enabled suicide thus clearly qualifies as suicide.

Some will resist the claim that such suicides are medically enabled, i.e., that they are in any way brought about by the involvement of medical personnel. After all, in acceding to the R.’s advance directive, medical personnel do not necessarily share R.’s intention. For they may not endorse the end she seeks to achieve. In examining the example of a suicidal individual “Tony,” with a similar clinical history and suicidal intention as R., Salter (2014, 46) argues that such examples do not constitute “aiding or abetting suicide”:

The decision to attempt suicide (by Tony) and the decision to withdraw care (by other relevant decision makers) are two separate — although certainly connected — issues. The hospital and clinicians did not aid, or even support, Tony’s original actions. They were not actors in that decision. … a subsequent, post-stabilization decision to withdraw treatment is separate from the suicidal act by the initial response of the medical team, which is to act contrary to the original goals of suicide. While Tony’s intention to cause his own death can be assumed, this need not be the intention of the clinicians in order for treatment to be withheld. Indeed, as is the case for all treatment refusals, it is assumed that the intention is not death, but instead to not prolong death or to relieve pain and suffering …Thus, the clinicians’ initial response to the suicidal patient and their divergent intentions sufficiently dissociate a decision to withdraw treatment from the original suicide attempt.

Salter is certainly correct that in such cases, clinicians do not aid or abet suicide inasmuch as they do not provide care with the intention of helping patients end their own lives. However, the partition between that intention and intentions such as relieving pain, etc., is more porous than Salter recognizes. For one need not intend what another intends in order to act knowingly so as to enable their intention to be realized. The cashier forced to turn over money at gunpoint does not share the thief’s intention of robbing the merchant. Yet in turning over the money, he does help the thief succeed in this intention. (This does not entail that the cashier is rightfully blamed when the thief makes off with the money.) Similarly, Salter’s argument does not enable clinicians to fully dissociate themselves from patients’ suicidal intentions, leaving them with ‘clean hands’ in cases such as these. Perhaps they ought not be blamed for their patients’ suicide, but that does not mute the ethical questions concerning their non-consensual involvement in them.

A more complex worry occurs when medically enabled suicide happens not as a result of ceasing life sustaining treatment but as result of never starting it. Suppose that R.’s advance directive had been discovered before the introduction of the mechanical ventilator, and medical personnel honor the directive. Might one argue that in this case, medical personnel do not enable R.’s suicide simply because involvement in her suicide requires them to act, and in this instance, the directive mandates that they not act? My opponent might rest this argument on the premise that medical personnel do not act to bring about her death but merely let R. die.

This conclusion is not plausible though. Action cannot be equated with behavior. And while it is true that in this case, there may not be discrete bodily movements that count as medical personnel not introducing the ventilator to R., their omission nevertheless counts as an act. For to refrain from doing what would otherwise be obligatory is an act—a mental act of choice. To deny this seems to invite sophistical conclusions. A bad Samaritan cannot assert that his not helping was not a wrongful act because he did not act, i.e., there is not some identifiable bodily movement that counts as his ‘inaction.’ Rather, he enables the death of the individual whom he opts not to help just as medical personnel enable R.’s suicide: by rationally choosing not to perform an act.

A final conceptual worry regarding medically enabled suicide is that it cannot be suicide because it is not a species of killing at all. One might argue that in R.’s case, she was not killed by anyone. Because R. was provided the ventilator, her suicidal act (her effort to kill herself) was interrupted by the actions of medical professionals. According to some physicians and medical ethicists, the subsequent removal of the ventilator results in R. dying from untreated respiratory failure. No person kills R. Rather, untreated respiratory failure kills her. Since R. is not killed by anyone, she does not kill herself and therefore cannot count as an instance of suicide.

This argument rests on the contentious claim that in removing R.’s ventilator, her doctors were not killing her but merely allowing her to die. I find this distinction implausible, an ethical shibboleth to sustain the fiction that doctors may not kill their patients (Miller et al. 2010). However, I will not contest that here. Instead, I draw attention to the apparent false dichotomy presupposed in this objection: Either R. is killed by someone (herself or someone else) or she is killed by her underlying condition. This objection seems to presuppose that if underlying respiratory failure kills the patient, then no one killed the patient. This presupposition reflects a crude picture of causation and the role of human agency in causation. For what distinguishes non-natural deaths is precisely a kind of dual causality: In cases of suicide or homicide, there are always two ‘killers,’ the individual who sets in motion a causal chain intending to lead to a person’s death, and the condition within that causal chain that proves fatal. Both the agent who kills and the condition that kills can be said the kill the person. In R.’s case, it is true both that R. killed herself (with medical professionals enabling this) and that untreated respiratory failure kills her. Indeed, it was R.’s intention in fashioning her advance directive and then taking the overdose that the withdrawal of treatment would kill her. She intended to kill herself by putting herself in a condition that would kill her once medical professionals responded appropriately to her advance directive. That the causal chain from her overdose to her death was interrupted by medical interventions does not make it any less the case that she killed herself. This is apparent when we envision a parallel case of homicide: Suppose that, in the course of an attempted robbery, R. was viciously attacked by a physician who just happened to have a portable mechanical ventilator. R.’s injuries are grievous enough that she will die without the ventilator. The physician fits R. with the ventilator and her condition stabilizes. Rummaging through the victim’s pocket, the physician finds an advance directive, mandating that mechanical ventilation not be administered. The attacking physician complies with the directive and R. dies as a result. No one would deny here that the physician killed R., despite its also being true that an underlying condition created by the physician killed R. Likewise, there is no basis for claiming that R. (in our original case) was not killed by anyone because her underlying physiological condition killed her.

10.3 The Attractions of Medically Enabled Suicide

Again, I make no assertion that medically enabled suicides are common. But it does represent a conceptually coherent category, and I would not be surprised if its numbers are significant and growing.

Why might someone seek out medically enabled suicide, and more specifically, why might someone opt for medically enabled suicide rather than suicide simpliciter? What is gained via a suicidal act intended to engage with medical care providers? These questions can be answered by noting that medically enabled suicide occupies a middle ground between more orthodox acts of suicide and the forms of assisted dying—physician-assisted suicide or voluntary euthanasia—available in a few jurisdictions (Dresser 2010). For while medically enabled suicide involves medical personnel in the death of an individual, it does not occur with the consent of those personnel. Hence, the attractions of medically enable suicide become more apparent when we consider the attractions of assisted dying.

First, many suicidal persons are deeply concerned for the effects that their deaths will have on others (Cholbi 2011b). In fact, according to Thomas Joiner’s well-developed theory of the causes of suicide, most suicidal persons are moved to end their lives from an essentially altruistic motive, the belief that they are burdensome to others (Joiner 2007). Most suicidal individuals also want to spare their loved ones the trauma of discovering their corpses or of encountering a badly damaged or brutalized corpse. Yet some of the most effective methods for ending one’s life (notably, guns) do tend to leave a damaged or brutalized corpse, and in many jurisdictions they are difficult to obtain in any case. Assisted suicide or euthanasia, because they are pre-arranged and use hygienic lethal methods, avoid such traumas. Moreover, assisted dying, while controversial, is also likely to be less psychologically distressing than ordinary suicide. Given the stigma of suicide, including popular beliefs associating suicide with madness or irrationality (Joiner 2010), individuals who wish to die have strong motivations to normalize their life-ending choices in ways that give those choices a patina of societal approval. By “medicalizing” patient self-killing, assisted suicide and euthanasia render self-killing less deviant, particularly in societies where over half of all people die in hospitals or other medically licensed facilities (US Department of Health and Human Services 2010, 43). Assisted suicide or euthanasia are less secretive or clandestine than standard suicides. Assisted suicide and euthanasia thus enable patients to have deaths with more of the features of a “good death”: peacefully, with family members and loved ones at one’s bedside. Finally, assisted suicide and euthanasia remove uncertainty regarding death as an outcome. Fortunately or not, many suicide methods, such as cutting or overdoses of over-the-counter medications, are not especially effective in bringing about death. The methods and techniques deployed for assisted suicide and euthanasia sometimes fail to end a patient’s life swiftly and without complication. One study of assisted dying in the Netherlands found that complications occur up to 16 % of assisted suicides or euthanasias. Rarely, however, do these complications postpone a patient’s death to a later date (Groenwoud et al. 2000). In contrast, ordinary suicide has a ‘success rate’ (i.e., a suicide attempts actually kills the suicidal person) in the single digits (World Health Organization 2006). Of course, the explanations for why suicide attempts prove not to be lethal are diverse. A medically enabled suicide in which an individual chooses a method with the very low probability of being lethal is not more likely to kill that individual than when an individual chooses that same method for a standard, non-medically enabled suicide. But a medically enabled suicide does circumvent one factor that prevents some suicidal acts from resulting in death, namely, medical interventions. Medically enabled suicides, by removing one significant barrier to a suicide attempt’s being lethal, have a greater likelihood of killing the suicidal individual.

Thus, because assisted suicide or euthanasia essentially medicalize the process of self-inflicted death, they offer suicidal patients a number of advantages over standard attempts at suicide. A death that is less traumatic, more conventional, more transparent, and more certain will be more desirable for most suicidal individuals. But of course, few suicidal persons live in the half dozen or so nations or five U.S. states that legally permit assisted suicide or voluntary active euthanasia. And even if a suicidal person lives in one of these jurisdictions, she is not likely to meet the legal criteria for assisted dying. R., suffering from bipolar disorder, does not suffer from a condition that is standardly classified as terminal, nor would her condition be judged futile (Cholbi 2013). Indeed, given that the vast majority of suicidal persons suffer from mental disorders (Joiner 2007, 192–202) rather than conditions such as cancer, few would satisfy the criteria to be eligible for assisted suicide or euthanasia. Finally, few suicidal patients will want to subject their care providers to legal risk, but at the same time, they may have difficulty identifying physicians willing to assist in their suicides, especially given (again) that they are not likely to be suffering from a condition that is terminal or medically futile. In most jurisdictions where assisted dying is available, the involvement of medical personnel is voluntary. The number of personnel involved in assisted dying is sometimes minute. In Oregon, whose Death with Dignity Act provides a legal protocol for assisted suicide wherein physicians can write prescriptions for lethal medications, only about 60 physicians generally write such prescriptions in a given year (Oregon Health Authority 2014). In a state with an active physician population of nearly 11,000 (Association of American Medical Colleges 2013, 191), that entails that only 0.5 %, or one in 200, of physicians have agreed to write such prescriptions. Consequently, even patients legally eligible for aid in dying may be unable to identify willing medical partners.

Assisted suicide or voluntary euthanasia thus offer considerable advantages to suicidal individuals, but also present considerable obstacles. My suggestion is that medically enabled suicides represent the next best alternative for many patients. Medically enabled suicide enables suicidal individuals to compel the engagement of medical personnel with their suicidal plans, but unlike in assisted suicide or euthanasia, without securing medical personnel’s assent to their suicidal plans. Medically enabled suicide also allows individuals to have a ‘normal,’ even ‘good’ death, in a culturally sanctioned clinical setting. Their deaths are less violent and more tranquil. Medically enabled suicides are culturally sanitized, cleansed of many of the popular negative associations with suicide. Suicidal individuals may also be able to exploit the ambiguity, both emotional and legal, of their courses of death. Their families and loved ones are less likely to associate their deaths with ‘suicide,’ given that it occurs in a care facility under the observation of medical personnel.

For patients like R., medically enabled suicide represents a viable middle ground between ordinary suicide—often more uncertain, dangerous, and psychologically harrowing—and a fully supervised ‘medical’ death. This is not to say that medically enabled suicide is without risk. For instance, an individual like R. may, due to ignorance or carelessness, take a fatal dose of a drug, thus resulting in an ordinary, rather than a medically enabled suicide. Still, the strong deference to patient choice regarding medical treatment seems to allow suicidal individuals to pursue a medically supervised, though not medically sanctioned, suicide.

10.4 Dilemma 1: Self-determination Versus Harming the Patient

But should medical personnel who are aware that they are being enmeshed in a medically enabled suicide assent to doing so? One rationale for their having a right not to participate in the care of someone with a clear intention to engage in medically enabled suicide stems from the claim that medical personnel should not knowingly harm their patients:

  1. (1)

    For a patient to end her life is a harm to a patient.

  2. (2)

    Medical personnel may not knowingly contribute to harming a patient.

  3. (3)

    So medical personnel may now knowingly contribute to ending a patient’s life.

    Therefore, medical personnel may not medically enable a patient’s suicide.

This argument is valid: (3) is a proper inference from (1) and (2), and the conclusion is inferred from (3) (along with the definition of suicide). However, the soundness of the argument is questionable.

To begin, (1) assumes that suicide is a harm to patients. Suicide can harm individuals by killing them, but it does not designate a class of harms. Rather, suicidal acts (should they succeed in killing a person) are harmful only if death is harmful. According to the most widely accepted account of the possible harms of death, the comparativist account, whether death is a harm to an individual at a given time is determined by comparing (a) the overall well-being contained in the individual’s life if she dies at that time with (b) the overall well-being contained in the life the individual would have had if she had not died at that time (more specifically, the overall well-being her life would have had if she had died at the next most likely time for her to die) (Feldman 1991). In other words, death is bad for us when, on balance, we would have enjoyed more intrinsic goods by surviving longer. According to comparativism, was R.’s death bad for her? Of course, this is a complex question, requiring us to think about the course of her life thus far, her likely future well-being, and the overall shape of the life she had with that of the life she would have had. Our concern here is not with what the correct answer is in R.’s case, or in any particular case. Rather, our concern is with the universal proposition implied in (1)—that suicide is always a harm to an individual. Clearly, if the comparativist account is correct, that will be a contingent matter. And it seems plausible to suppose that at least sometimes a person’s dying prematurely due to suicide results in her having a better (albeit shorter) life overall than the life she would have had if she had continued to live.

As for (2), it is clearly not true in its unqualified form. For one, there will be many risky medical treatments for which it is not apparent whether medical personnel are knowingly contributing to harming patients. Or at the very least, whether they contribute knowingly to harming a patient can only be judged in retrospect. Furthermore, medical personnel are clearly not barred from contributing to harming a patient when the evidence suggests that such harms are essential to a larger course of treatment that will prove beneficial to a patient. Many treatments have harmful side effects. But surely medical personnel are permitted to administer or recommend such treatments, despite their harmful side effects, precisely because of the reasonable expectation that the patient stands to benefit on the whole. It would remarkable if, for example, (2) disallowed medical personnel from administering chemotherapy to early stage cancer patients simply because chemotherapy is known to subject patients to identifiable harms.

Supposing that (2) could be amended to meet this worry, other worries become apparent when we consider whether medical personnel may contribute to harming their patients in non-suicidal cases. If a competent patient opts not to receive (or to forego) a life extending treatment, few medical personnel believe they are entitled to recuse themselves from the provision of care because of their belief that such care contributes to the patient being harmed. Their belief that the patient’s choice is harmful to her may well be true (and true according to the comparativist account of death’s badness). But medical personnel routinely involve themselves in courses of care or treatment that harm patients via the withholding or withdrawal of treatment. It would, I propose, be a dramatic reinterpretation of extant medical ethics if medical personnel could refuse to provide care or treatment that they believed to be on balance harmful to patients, even when their belief is true.

This illustrates that (2) intersects with (1) in thorny ways: To assume that medical personnel may not contribute to harming patients is to prioritize non-maleficence over other core bioethical principles, most notably, patient autonomy. It seems rather to be the case that either (a) medical personnel must honor patient autonomy and provide care that is objectively harmful to patients, or (b) patient autonomy rests on the thesis that a patient’s competent judgment regarding whether a course of action is harmful to her is determinative, that is, a judgment to that effect is one to which medical personnel ought to defer, perhaps on the grounds that only the patient is epistemically positioned or entitled to make such a judgment on her own behalf. Either way, medical personnel who refuse to provide specified forms of care in order to avoid doing what they believe harms the patient thereby consign patient autonomy to a rather marginal role in their understanding of their clinical obligations.

I conclude, then, that this dilemma dissolves under further scrutiny. It rests either on contestable claims about the harmfulness of suicide or is difficult to make consistent with the value the medical community ascribes to patient autonomy. The apparent dilemma is therefore best resolved in favor of medical personnel being obligated to participate in medically enabled suicides.

10.5 Dilemma 2: Self-determination Versus Contributing to Suicide

A second dilemma caregivers may face with respect to medically enabled suicides arises from the conviction that suicide is morally wrong. That conviction is clearly controversial, and I will not attempt to decide that matter here (Cholbi 2011a, 39–69). But if, as some believe, it is reasonable to permit medical personnel who conscientiously object to abortion to forego participating in abortion procedures or to forego the provision of abortifacient drugs, then it is reasonable to permit those who conscientiously object to suicide to forego involvement in medically enabled suicides. Indeed, conscientious objection to involvement in medically enabled suicide may rest on similar grounds to conscientious objection to abortion (that it kills the innocent, etc.). If so, those with conscientious objections to suicide may believe that medically enabled suicides place them in a dilemma, demanding that they either ignore their own consciences or ignore patient autonomy.

This second dilemma can be resolved by addressing an importantly different case from R.’s:

S. is injured in a gruesome industrial accident. Though S. survives her injuries and could likely live for a significant period of time post-hospitalization, many of her injuries are permanent. S. had one leg and one arm amputated. In addition, S. suffered injuries to her chest, abdomen, and lungs, and as a result S. has lost the ability to speak and requires supplementary oxygen. S. will be unable to feed herself and will never be employed again. Although S. may be able to return home, the regular presence of a home health aide will be necessary.

A few months after returning home, S. suffers an episode of severe respiratory distress. Once stabilized, S. tells nurses present that she no longer wishes to live with such poor quality of life, at constant risk of complications or hospitalization. She asks that her respirator and feeding tube be removed later that day. S.’s family congregates at the hospital. S. permanently loses consciousness several hours after the respirator and feeding tube are removed. She is declared brain dead 16 h later.

The most morally salient difference between R.’s situation and S.’s is the causal role played by their respective desires to die. In R.’s case, her desire to die is a cause of the state of the affairs produced by her self-injury, a state of affairs she directs others to intervene in so that she will die. In S.’s case, her desire to die is an effect of the state of the affairs produced by her injury, a state of affairs she directs others to intervene in so that she will die.

Those who believe that medical personnel should be entitled to abstain from involving themselves in medically enabled suicides must put a great deal of argumentative weight on this contrast. For it seems uncontroversial that S.’s competent request to cease life sustaining measures must be honored. Thus, if R.’s advance directive is not to be honored but S.’s request is, this pair of judgments must rest on plausibly ascribing inherent moral significance to the different roles played by the desire to die in each case. It will not suffice to assert that the difference is that R.’s desire led her to suicide, whereas S.’s did not. For one, it is arguable that both are suicides, inasmuch as both R. and S. undertake courses of action intended to result in a death that has their rational endorsement in the circumstances. But even if we embrace the contrary view that R. engaged in suicide while S. merely allowed herself to die, we do not yet have a principled basis for a conscientious objection to not medically enabling R.’s death while acceding to S.’s request, aware that it will end her life. As LaFollette and LaFollette (2007) observe, a claim of conscientious objection is subject to a number of criteria, among them that the objection rests on a “core belief” of the objector, “consistent with other things he says or does” in the course of his professional practice. In my estimation, the objector bears the burden of proof here to show that his objection(s) to the treatment in question can be coherently squared with his other attitudes. No doubt some avenue is available to justify disparate treatment of R. and S., but I cannot ascertain an avenue that would succeed in showing that the differing causal roles played by their respective desires to die justify their disparate treatment by medical personnel.

I propose, then, the contrast between a desire to end one’s life being a cause of an injury and its being an effect cannot bear the weight necessary to justify conscientious objection to medically enabled suicide. Furthermore, advocates of such an objection must address the apparent comparative unfairness faced by R. Both R. and S. are victims of bad moral luck. R.’s bad luck is a combination of circumstantial and constitutive luck, luck emanating both from her surroundings or environment and from her unchosen traits or dispositions. S.’s bad moral luck is circumstantial and resultant, emanating from her surroundings or environment and from variations in outcomes (i.e., S. or other workers had been in the same workplace situation many times in the past without being injured) (Nagel 1979). I cannot address the large scale questions about justice raised by how victims of bad moral luck should be treated. Luck egalitarian theories of justice, for example, requires that differences in individuals’ well-being, opportunities, etc., should hinge wholly on their choices and not on the sort of bad luck that befell R. and S. But we need not embrace such theories wholeheartedly to conclude that, absent some compelling argument to the contrary, victims of bad moral luck who are otherwise alike should not be treated differently. As we saw above, there is one notable difference between R. and S. that does not depend on luck, namely, that R. put herself in a life-threatening medical condition through her own conscious acts (granting that those acts would not have been performed were it not for facts about R., such as her illness, that arguably are the product of luck). But this difference, not dependent on luck, does not warrant differential treatment of them. Yet no luck-based differences exist between R. and S. to warrant differential treatment of them either. Again, that R.’s desire to die causally prompted her injury, whereas S.’s desire to die was effected by her injury, is not a difference that warrants differential treatment of them. R. may rightfully complain that if, her advance directive is not honored and her medically enabled suicide attempt stymied, but S.’s request to end life sustaining interventions is honored, then she is treated differently without any moral basis. She would thus be twice victimized, first by her own moral luck, and second, by those medical personnel who unjustifiably refuse to honor her directive. It does not seem reasonable for our moral obligations to differ from patient to patient simply because the patient has a self-inflicted condition (Salter 2014, 44).

10.6 A Note About Competency and Mental Illness

To this point, I have shown that medically enabled suicide is attractive to those in particular medical and social circumstances. Moreover, the apparent dilemmas that medically enabled suicide raises for medical caregivers are just that: apparent. Upon further analysis, there are not plausible arguments, appealing either to the wrongfulness or harmfulness of suicide, that make sense of caregivers justifiably disregarding the wishes of those suicidal persons who enlist those caregivers in enabling their suicides. Hence, one horn of each dilemma turns out to be specious. Suicidal persons have a right to have medical personnel enable their suicides.

In reaching this conclusion, I do not claim that medically enabled suicide is in no way morally problematic. One might think that medically enabled suicide amounts to exploiting or taking advantage of ethical rules or protocols designed for a very different purpose, to wit, to ensure that individuals can choose the courses of treatment that they judge best for their health and well-being. But we do not have to admire R. or her choices to believe that she has the right to make them and that medical personnel are obligated to enable her to achieve her chosen ends.

One possible reaction to the arguments I offered in Sects. 10.3 and 10.4 is to claim that there are other bases for medical personnel disregarding R.’s advance directive besides the harmfulness or wrongfulness of suicidal conduct. R. collapses into unconsciousness soon after arrival, rendering her an incompetent decision maker. And of course those are precisely the circumstances in which an advance directive is applied: when an individual is unable to render competent decisions on her own behalf. A defender of medical personnel having the right to disregard the advance directive might argue that, in light of R.’s ongoing mental illness, the advance directive should be disregarded. But this assumes that the mere fact of R.’s being mentally ill (or having been ill when she fashioned her directive) is sufficient to overcome the strong burden of proof normally associated with challenges to advance directives (Appelbaum 2007). Note that it is rarely assumed that those who create advance directives while suffering from the travails of ‘physical’ disorders are creating invalid directives. Indeed, there remain pervasive prejudices about mental illness, prejudices that deny that these conditions are genuine or that the suffering they produce can measure up to other forms of suffering. We should therefore be very wary of those willing to dismiss R.’s advanced directive solely on the grounds that her condition, being mental, is ‘all in her head’ (Cholbi 2013). I am thus sympathetic with the conclusion reached by Brown et al. (2013, 10–11) that the cause of a patient’s condition, including whether that cause is psychiatric in nature, should not affect clinicians’ willingness to forego or withdraw life support, and that such acts are the “rough moral equivalent of withdrawal after comparable critical illness or injury.” In any case, the determination of R.’s competency at the time when the advance directive was already performed by other professionals, and it would be imperious indeed for medical personnel to later decide that the directive was itself incompetently fashioned.

Those advocating that the advance directive may be disregarded may then appeal to the claim that R. engaged in irrational suicide behavior. Even if the advanced directive was competently fashioned, R. only needs medical attention because of an irrational act on her part. But this argument is even less appealing: As we saw in Sect. 10.3, the comparativist account implies that ending one’s life prematurely is not necessarily irrational. Furthermore, to allocate medical care on the basis of whether a person’s condition stems from an irrational choice on her part is anomalous, even cruel. Medical personnel do not deny care to motorcycle riders who irrationally choose not to wear their helmets, nor do they deny care to those whose injuries result from irrational alcohol abuse. It is simply no part of recognized medical ethics to determine which patients deserve care by appeal solely to clinicians’ judgments about the wisdom of those patients’ choices. Lastly, the central rationale for advance directives is to enable individuals to exercise a form of penumbral autonomy—to ensure that medical decisions made while one is incompetent reflect the values or preferences one can express while competent. Among R.’s purposes in creating the advance directive was to provide guidance to medical personnel on how she ought to be treated while in the incompetent condition that she herself created through her suicidal act. Yet an unsettling precedent would be set were medical personnel to set aside the advance directive purely because the patient’s own act triggered the condition under which the directive became salient.

Admittedly, competency is a more difficult matter if, unlike R., the patient is conscious and expressing wishes regarding her care. Here familiar, but not simple, problems arise concerning the determination of competency. Medical personnel would be advised to consider not just the suicidal patient’s expressed wishes in the moment, but also other evidence concerning her wishes (an advance directive should one exist, suicide notes, prior statements to family members, and so on). But determining competency is an ongoing feature of clinical practice, to be pursued on a case by case basis, and so cannot offer a principled, or even presumptive, basis for medical personnel refusing to involve themselves in medically enabled suicides.

10.7 Conclusion

In arguing that these two dilemmas are specious, I have sought to undermine whatever principled moral ground there might be for permitting medical personnel to refuse participation in medically enabled suicide. An implication of my arguments is that the fact the individual arrives at a life-and-death point via an act of suicide is not per se ethically relevant to the treatment obligations that medical personal bear toward such an individual (Lowenthal 2002). Again, concerns about patient competency are relevant, but no special issue of clinical ethics is raised by the facts of the patient having engaged in suicide or having done so in the hope of achieving a medically enabled suicide. For neither fact is more than contingently related to morally salient facts that do shape treatment obligations. Desperate times call for desperate measures, and so long as most suicidal individuals lack access to assisted dying, it appears likely that at least a handful will seek what they believe is a good death for themselves. Medically enabled suicide offers such individuals the advantages of a death with medical personnel acting as death’s non-consensual guarantor. If I am correct, medical personnel must fulfill this role, however distressing that may seem.