Killing or Letting Die, a series on decision-making when death is at the table:
Part 1 - Intending or Pretending?
Part 2 - What's My Duty?
Part 3 - Causation or Conflation?
Part 4 - The Priority of Authority?
We started our safari with a paper by James Rachels. He described two cases to help make his point that there’s no moral difference between killing and allowing someone to die. In one case, a man forces his young cousin under water to drown him. In the other case, the cousin slips, hits his head, and slides under the water. The man watches on but doesn’t touch the boy. Rachels argued that if there were a true difference between killing and allowing to die, we would claim the man in the first case behaved worse than the man in the second. We don’t, and therefore Rachels concluded there’s no difference.
But is there?
We saw in the last issue how Rachels overlooked intention. We can also see how he overlooks duty.
We don’t know a lot about the men in these cases, but they at least have a duty not to kill their young cousin. The basis of that duty is found in the cousin’s right to live. We’ll touch on the limits of that duty shortly, but suffice it to say, in general, we all have a duty not to kill others.
But what about the man in the second case? He didn’t cause his cousin’s death! We’ll tackle causation and responsibility in an upcoming letter. For now, notice that while it’s true he didn’t cause his cousin’s death, it’s a pretty anemic society that would allow our duty to end with merely not killing others. In some cases, we may have a general duty to intervene to prevent harm to someone else (e.g., Peter Singer’s “drowning child” thought experiment, which he uses to illustrate a different point). But here, this adult man has a special duty to his young cousin to rescue him from drowning after he slips. He is a caregiver, as far as we can tell, by virtue of his age and familial ties. That role confers a special duty, or strengthens general duties. We’ll see if there are any limits to that.
What is duty?
When we tracked the concept of intention, we noted that a lot of folks have written about it. The same goes for duty, and I can’t do it all justice here. I’ll keep it narrow to this question of differentiating killing and allowing to die. However, if I’m at least able to provoke a little more curiosity and meaningful conversation about this, I hope I’ve done well.
Simply put, a duty is what we owe to someone else, an institution, or a standard. That claim might arise because of my role (e.g., caregiver, clinician, law enforcement officer), because of prior action (e.g., I created the circumstances that put someone in a position of need), because of a contract, or because of someone else’s rights.
Clinicians’ duties are informed by law, by professional codes of ethics, and by the informed consent process (a contract). Even within those seemingly concrete boundaries, there’s a lot that’s fuzzy. Consider: the strength of my duty to my patient may wax and wane depending on a number of factors. A surgeon may have a general duty to help their patient in ways that an operation might ameliorate. Once the operation is done, they may have a greater duty to help their patient with any adverse consequences that befall their patient post-operatively. However, if that patient develops, say, a vitamin deficiency a year later related to the surgery, the timing and type of complication may suggest that the surgeon doesn’t have the duty to address this problem. Someone else may. In fact, health systems diffuse duties across the bureaucracy of expertise. Clinical duties are shared, discharged, transferred, minimized, and expanded. None of this is quantifiable.1
It’s beyond me to produce even a short list of clinical duties here, but as you’ll see as I develop the idea, there are some things we need to have straight before we start making such a list. Clinicians have all kinds of duties, but at issue here is whether they have a duty to end someone’s life, to never end someone’s life, to treat a patient over their objection if the patient wants to stop life-sustaining therapy, or to stop life-sustaining therapy under certain conditions even if death is foreseen.
Who has a duty?
Well, I can say this much: machines don’t have duties.
Why? Because machines aren’t moral agents. Humans, including clinicians, are. Only a moral agent can have a duty. That will pull in all sorts of relevant considerations in a moment, so it’s important to establish it at the get-go: clinicians with duties aren’t machines.
There are innumerable quirks you could inject into Rachels’ cases to show how the moral evaluation might change based on gradations of duty. For example, in a third case, the man didn’t know the boy had slipped in the bathtub and later discovered him dead. We might claim that he had a duty in general to care for the boy, but in those specific circumstances, he couldn’t act on that duty in a morally relevant way. The time to act on his duty to save the boy’s life had passed.
In a fourth case, the man saved his cousin and called the paramedics, who then whisked him off the hospital. The boy died en route because a paramedic administered the wrong medication. The duty to care for this boy was transferred from the man to the paramedic. We can’t say the man committed a wrong in this case.
How we determine who has a duty hinges on our capacities, limitations, and roles, as well as the rights of others. These are negotiated in many little and large ways. It’s uncommon for any of these to be absolute, although not killing someone is pretty close.
Rosamond Rhodes remarks:
“Society … permits doctors an extraordinary set of powers, privileges, and immunities. Physicians are allowed to ask probing questions and examine nakedness (i.e., invade privacy), they are given license to prescribe medications (i.e., poisons), they are granted the privilege to perform surgery (i.e., assault with deadly weapons) with immunity from prosecution, and they are also empowered to assess individuals’ decisional capacity and sometimes override the expressed wishes of their patients and impose unwanted treatment on them. … physicians are given license and immunity to do what nonphysicians are prohibited from doing.
…
Because society’s trust is essential to the practice of medicine, society and the medical profession itself have developed explicit and implicit conventions, laws, and penalties to constrain physician behavior and ensure the trustworthiness of the profession.”
So, society grants powers and it also sets limits.
One key mechanism through which this occurs is via consent. Patients (or their surrogates) activate a clinician’s duty through consent. A clinician can’t be said to have duty to treat a patient who hasn’t consented to treatment (except in limited, emergent circumstances).2 However, clinicians, by virtue of their office, do have a duty to provide the information necessary for someone to determine whether they’ll consent to whatever is being offered.
Consent is (usually) necessary but insufficient. On the one hand, someone may consent for their uncle, who is not a clinician, to attempt a craniotomy on them. We can’t say that creates a duty for the uncle to perform the craniotomy, and in fact, the consent itself likely wouldn’t exonerate the uncle from whatever befell this poor person. On the other hand, they may consent (or want to consent, anyway) for their primary care physician to also be their chauffeur. That also doesn’t create the duty for the physician to drive this person around town. Consent activates a duty that is already, in a way, latent within the role of the person, but it doesn’t create duties. Consent without the other’s capacity to act is just a wish.
Because clinicians have unique duties to patients, they may be uniquely positioned to “let die” in ways others won’t, and therefore are also at greater risk of “killing.” We’ll see how that plays out.
What do clinicians owe their patients?
This is another way of asking, what are the duties of clinicians? It’s a big question, one that I can’t fully explore here. I just want to trace a few silhouettes of thought.
Rhodes argues that it’s society that confers on clinicians their powers, privileges, and immunities. This is because society recognizes our mortal bodies are frail and vulnerable, and we each have health-related needs. We set apart some people to learn about the body and how to help it. That doesn’t mean clinicians are the only professionals who need to consider health. Many others do - e.g., health safety inspectors, automobile manufacturers, chefs. What makes clinicians different is that they’re uniquely qualified, by their knowledge and skill, to intervene on the human body for a particular purpose. Other professions intervene on the human body too - e.g., tattoo artists, barbers, jewelers. What separates clinicians from these other professions is the aim of their work: clinicians aim to sustain and restore health.3
Our role makes promises to others about what we can do and what we might owe them. It wouldn’t make sense to say that clinicians owe patients whatever patients want. Someone might want you to pay for their lunch after their appointment today, but as a clinician, that’s not your duty. Even in a more narrowly circumscribed request, it also wouldn’t make sense to owe patients whatever they want that your unique expertise can provide. It may not accomplish what the patient think it will. But even if it would, there may be counterbalancing considerations that would keep you from fulfilling their request.
So, how might we approach this question of what we owe patients?
The role is voluntary. Clinicians voluntarily take on their profession, which means they take on its duties voluntarily. The role of the patient, on the other hand, usually isn’t voluntary. Of course consent is voluntary, but the fact that someone might find themselves in need of the services of a clinician isn’t itself voluntary (usually). This mismatch in role voluntariness helps to determine the direction and magnitude of the duties: patients have the needs in this relationship.4
The duties are received. Clinicians don’t get to pick and choose the duties they’re going to assume; they accept the duties passed down to them through the tradition of their profession. While the technical practices will evolve with time, the duties are usually static in reference to the goal of the practice. Recently these are changing: do clinicians have a duty to end the lives of their patients in certain circumstances?
The duties are necessary. The duties are a prerequisite of remaining in the role. There may be circumstances where you can’t forsake the duties without professional liability (e.g., loss of role). Indeed, sometimes forsaking one’s duty incurs criminal liability. Duties keep the clinician out of a trouble, and also help keep them on track toward the purpose of their practice.
The duties require technical expertise. The duties of all clinicians have much in common, but they also vary among specialists - the duty of a cardiologist is different from the duties of a respiratory therapist, radiologist, and nurse. This acknowledges that someone in a particular role “can do” something, but it’s insufficient to determine if they “must do” something or are “permitted to do” something. We can’t say what a profession must do until we know what they can do.
The duties require consent. As mentioned earlier, a clinician’s duty to act is only activated by a patient’s consent that they act on them (except in limited, emergent circumstances). Even before a patient offers formal consent for an intervention, their presence implies consent for you to explain to them what you might be able to do for them. You’d at least have the duty to offer truthful, helpful information to guide them toward informed consent. It isn’t reasonable to claim that a clinician has a duty to a patient who doesn’t consent to what the clinician can offer.
The duties require a purpose. If a clinician has a duty to help their patient, we can ask, help them how? For what? It can’t be to help them with anything, for any purpose. So the clinician’s duty is circumscribed in some way, generally by their technical expertise. At what should the clinician’s expertise aim? Here we come to a controversy. One might argue that, because the profession of medicine is socially constructed and conferred on clinicians because of the needs of society, society could re-write this contract to say, “We want clinicians to do what we want, not just what they think is best.” Society could expand this idea of “respect for autonomy” beyond a negative sense (i.e., avoid imposing on a patient something they don’t want) to include a positive sense (i.e., give to a patient something they want). This is a huge topic, probably best reserved for its own letter, but I’ll address it in closer detail as it relates to our question in a moment. For now, I’ll assert that the purpose required by the duties of the clinician is the pursuit of patients’ health. If we want to modify that purpose, we’ll also modify the clinician’s duties.
The duties are limited. Any number of circumstances might modify or eliminate the claim of a clinician’s duties. For example, if they’re not in the building when their patient dies, they might not have violated their duty to save their patient’s life - that is, unless they should have been in the building at the time. Clinicians are human: even within the narrow scope of their technical expertise, they can’t be accountable 24/7/365.
If the purpose of medicine is to support and restore patients’ health, then the foundational duty of the clinician is to help the patient in that pursuit. But why is this a clinician’s duty, rather than just a permitted (but unrequired) activity?
A simple answer is because law requires it. Even deeper than that, though, the patient is uniquely vulnerable by way of their need, and also by way of what is required for the clinician to evaluate and treat them. As Rhodes observes, clinicians may need to examine a patient’s naked body, take samples from that body, cause some degree of pain, give substances that change all kinds of physiologic processes, even create holes in the body and remove entire organs. Talk about vulnerability! And it requires tremendous trust. Patients, and society, need assurances beyond the understanding that clinicians “can do” these things. They need to know that, when the need arises, clinicians “must do” these things. Society grants the power, and demands the assurance that clinicians won’t flake out when the time to use that power comes.
But it’s not like (most) clinicians need their arm twisted to behave this way. Many willingly shoulder their duties, driven by compassion. Indeed, it might make us feel uneasy if we knew a clinician was only caring for us because they were compelled by duty to do so. Here’s where duty and virtue link hands in the pursuit of the good of health - a topic for another time.
Do clinicians have a duty to end the lives of their patients?
Alright, here we go. You’ve waded through my feeble attempt to introduce some of these concepts, now let’s move on to my feeble attempt to apply them to understanding the difference between killing and letting die.
It’s already clear that clinicians can leverage their technical expertise to end the lives of their patients. This is certainly something they “can do.” And in some jurisdictions, they’re permitted to do it within certain boundaries. But must they do it, if the patient requests it? These questions matter because they may help us establish a difference between killing and letting die.
It might seem straightforward: society grants to clinicians their powers, so society can modify those powers as it sees fit. One parallel example from the furious controversies of real life is on the matter of abortion in the United States. Due to a federal judicial change, abortion can now be outlawed at the state level. People balk at politicians inserting themselves into medicine, but in their frustration folks fail to appreciate that medicine’s self-regulation is granted by society via law. Society can extend or retract the power it grants clinicians based on its needs and values.
Clinicians and others can advocate for changes one way or another, in legal as well as academic, philosophical, ethical, and informal forums. The advocacy itself demonstrates we don’t believe law is foundational; something else undergirds the law to make it just. This is the basis on which Martin Luther King Jr. echoed Augustine when he wrote, “An unjust law is no law at all.”
Similarly, the law around assisted suicide and euthanasia (ASE) may go too far or not far enough when considering these deeper matters. The law is helpful in recognizing, protecting, and enforcing a duty, but the law needs to be justified. So, society might err in either direction: permitting the impermissible, or disallowing the allowable. Even worse, it might require the opposite of that which is required, or fail to require what’s required.
I’ve argued why death cannot be a goal of care elsewhere, but let’s just run with it for a bit. I’ve already made a (feeble, contentious) case that the goal of medicine is health. The clinician’s fundamental duty is to the pursuit of that end for their patients. It is not, as I’ve already claimed, to satisfy whatever their patient desires, even within the realm of the clinician’s technical expertise. We’ll further refine that point shortly.
To permit killing one’s patient tampers with the foundations of medicine. To make it a duty grounded in the patient’s request is to change the practice of medicine into something else entirely. Clinicians devolve into technicians of the body, tools in the hands of patients.5
Ironically, though, by treating clinicians like tools, we also degrade their moral agency to the point that it’s difficult to claim they have any duties at all. As mentioned earlier, machines don’t have duties. To treat clinicians like tools is to literally de-moralize them. A moral agent’s relationship with their duty is one that they appreciate with their conscience. But that doesn’t seem to be how some want duties to function for clinicians. Treating a duty like a lever instead of a claim on a moral agent’s conscience is a form of disrespect. We’d rather not deal with fallible, idiosyncratic humans. What we really want is a machine that satisfies our desires, even if those desires don’t coincide with our health. I’ve written before about this aspirational hope of health without people. Mere preference satisfaction de-moralizes medicine. One extreme variety of this is killing.
A practice of medicine that devolves to mere preference satisfaction de-moralizes the practice and the clinician by making individual patients the arbiter of good medicine. We hand over the keys of the profession to patients to decide what’s good. And not just the abstract crowd of “patients,” but the patient in room A will decide one way while the patient in room B decides another. Meanwhile, the clinician caring for them needs to have the qualities necessary not to pursue health, but to abide in doing so many things at once for so many different purposes. Their practice will be stretched thin across the banquet of what patients demand of them. They must be prepared to prevent one person’s suicide while aiding another’s. In so doing, they must adjudicate (because they are still human with a conscience, despite appearances to the contrary) whether the patient’s preference is valid. These expressions of preference still require the clinician’s authorization.
I don’t think anyone has adequately justified such a duty, and so I’d say clinicians do not have a duty to end the lives of their patients. If this duty were ever established, its creation would be concomitant with the total transformation of the purpose of medicine. We wouldn’t be able to call it “medicine” anymore.
Do clinicians have a duty not to kill their patients?
Unless we torture the definition of health, ending someone’s life cannot fall within its purview (as I previously reflected). Therefore, if clinicians owe patients assistance in pursuing their health, clinicians owe it to their patients not to kill them. Not only that, but the patient’s own right to life protects them, as do laws against murder.
Laws and professional statements that permit the killing of patients don’t obviate this duty. I’ve already explored how laws require deeper justification than their own power. The duty not to kill a patient is far better justified than the permission to kill a patient.
What is a clinician’s duty regarding the use of life-sustaining therapy (LST)?
We’ve had in view the “killing” part of our question, and now we move to the “letting die” part.
There are two corollary questions here. First, do clinicians have a duty to deploy LST? This is probably the least controversial question of the bunch. Still, it raises some degree of distress among clinicians because they may feel compelled by a request from family or patient to use these therapies even if the clinician feels like the likelihood of benefit is low. A full exploration of this question will take us too far afield of our main task, but in my experience, clinicians can overlook the value in permitting a time-limited trial of LST to allow space for grief to make an open appearance within a family. Clinicians need to expand their understanding of “beneficial” to include uses of LST to allow for that space. Using LST even in circumstances with physiologic benefit will be minimal may afford time to realize another goal.
Second, do clinicians have a duty to stop LST? I’ve previously explored this question as it relates to pacemakers, so I won’t repeat myself here. However, I want to frame a response in terms of our present inquiry. Do clinicians have a duty to let a patient die? The way this question is framed may draw your memory back to our discussion of intention. A clinician might say they have no duty to let someone die, but they do have a duty not to assault a patient with a medical intervention to which they don’t consent, even if the foreseen but unintended consequence of withdrawing or withholding that LST is death.
Nevertheless, isn’t withdrawing or withholding LST just the same as killing? If we have a duty to withdraw or withhold LST, what makes killing different? Joachim Asscher poses an interesting argument which we’ll explore from a different angle when we discuss causation. He describes this idea of “surrounding responsibility.” If you create the circumstances in which something else occurs, you bear this “surrounding responsibility” for that latter event or action. He doesn’t discuss duty, but let’s look at how surrounding responsibility might create a duty.
In a medical context, this surrounding responsibility belongs to clinicians who manipulate the body for any number of reasons. You might be hard-pressed to claim that a surgeon caused a post-operative blood clot in a leg. However, the surgeon created the circumstances that allowed for this to happen; they have this “surrounding responsibility.”
Asscher argues that there may not be a difference between killing and letting die because when a clinician has surrounding responsibility for the patient’s situation, there may be no extra responsibility to take for killing someone compared with letting them die. The clinician is responsible for the situation in which the patient is going to die, so it’s just the same as if they killed the patient. “Where medical staff have a high degree of surrounding responsibility, it is inappropriate to appeal to the distinction to justify letting die.” It doesn’t make sense to claim that letting die is different from killing some situations because clinicians have this “surrounding responsibility” for the entire circumstance in which this person’s life is in their hands.
Asscher doesn’t carry the argument in this direction, but if there’s a duty to let someone die by withdrawing or withholding LST because of this degree of surrounding responsibility, could there also be a duty to kill them?
I don’t think so. The duty to withdraw or withhold LST isn’t grounded on ending the patient’s life. As we considered in Part 1, treating someone over their objection when they have the capacity to decide otherwise is an attempt to live their life for them - an egregious evil. Even if killing and letting die are similar in a causative sense, that similarity doesn’t mean the duty to one is also the duty to the other.
Are there limits to duty?
I already suggested that there are almost innumerable ways responsibilities and duties can be modified, curtailed, or even eliminated. There are obvious logistical, physical limits (e.g., the barriers of time and space). There are also ethical limits. If consent activates duty, lack of consent limits duty. Whereas consent was a necessary but insufficient ingredient in directing the clinician’s action, lack of consent is both necessary and sufficient to limit duty. Even in those cases where the patient themselves cannot consent, someone consents for them - either a surrogate decision-maker or, via law or policy, a communal form of permission for emergency circumstances.
But why should this be the case? Why shouldn’t clinicians exert any and all effort in doing whatever they think is best in pursuit of health? After all, the patient has a right to live. That right isn’t one that the patient must claim for it to be active; the state and the profession of medicine collude to prevent suicide all the time.6 What if the patient is attempting suicide by requesting that we withhold or withdraw LST?
You really can’t talk about this one in the abstract. Consider this case:
A man was stricken with a severe bout of pneumonia requiring mechanical ventilation. His healthcare team believes he’ll be safely extubated in the next 24-48 hours. However, he requests that they remove the breathing tube immediately “so that he can die.” He writes frequently about how he hopes he’ll die and would end his own life if he could. Assume for the sake of argument that he has the capacity to make this decision.
Consider another case:
A woman with lung cancer that has spread to her brain, bones, and liver is stricken with a severe bout of pneumonia requiring mechanical ventilation. No further cancer-directed therapies are available to her. Her healthcare team believes she won’t survive this hospitalization. She requests that they remove the breathing tube immediately “so that she can die.” She writes frequently about how she hopes she’ll die and would end her own life if she could. Assume, again for the sake of argument, that she has the capacity to make this decision.
Both patients are acting with suicidal intent in requesting the healthcare team stop their LST. Most clinicians would probably attempt to cheerlead or otherwise negotiate with the man in the first case, whereas they may immediately acquiesce to the woman in the second case. Recognizing there are other ethically relevant facets to this case worth asking about, do the clinical differences between these two people modify the clinician’s duty to either withdraw or continue LST?
Let’s take the second case first. There are certain harms that befall a dying person that only a dying person can suffer. This woman is going to die, possibly soon, with or without the ventilator. Continuing the ventilator means she dies with the intrusion of an intervention designed to keep someone alive (even if they’re otherwise dying). All things considered, the clinician might do more harm than good by forcing the ventilator on this dying woman. Furthermore, while the treatment philosophy for managing acute suicidal ideation is sometimes involuntary hospitalization to allow for crisis stabilization, that is often not the treatment philosophy for chronic suicidal ideation. Even psychiatry acknowledges the limits of its own intervention on suicidality. That’s relevant for this case because the hope of remitting the suicidal ideation is virtually nil for this woman. There’s likely not enough time left in her life.7
Now on to the first case. This man may have been dying but for the ventilator and antibiotics he’s receiving. In a different era, he may have been dead already. So what’s different about his decision? Well, in this case, there may be a role for acute crisis stabilization. Believing that he’ll survive the medical insult, the clinician’s duty to preserve life in the face of a suicidal threat is greater. Without more information, we might surmise that, like most bouts of suicidal ideation, this too will subside with time, and will improve even more as the acute stressor of the ventilator is removed at the appropriate time.
It would be a mistake for clinicians to blithely override or acquiesce to either patient’s request. A decision requires careful deliberation and negotiation. The withdrawal of consent, stark and on its own, is just insufficient to tell a clinician what to do with their duty when it comes to LST.
However, the limit of the clinician’s duty, if there is one, doesn’t translate into a duty to kill the patient at their request. These actions aim at two different goals.
To sum up: clinicians have a duty not to kill their patients. They also have a duty not to assault their patients with LST, even if death is foreseen. This is sometimes called “letting die.” This doesn’t translate into a parallel duty to kill patients in similar circumstances. Furthermore, it would be too much to say clinicians have a duty, in any circumstance, to “let die.” The limits imposed on a clinician’s duty imply that they aren’t “letting die” but rather avoiding other forms of harm even if it means the patient will die.
But don’t clinicians cause a patient’s death by withholding or withdrawing LST? This will bring us to the next stop on our safari: causation.
Clinical teams may set up little economies attempting to quantify some of this. Clinicians might trade two regular call shifts for a single holiday call shift, for example. The duties of being on call over a holiday may count for more in this little economy than they would on a regular day.
For example, presenting unconscious to an emergency department without any documentation about care preferences. Or, presenting in the midst of a psychiatric crisis where the patient lacks the capacity to consent to or decline hospitalization.
Others argue that health is only one of several goals of medical care. A major goal we should consider, apart from health, is the patient’s own values and desires. Therefore, patient’s may want things that are irrelevant to their health, and perhaps even contradict their health, and experts in medical technology should be prepared to offer them that. I address why that might not be a good idea elsewhere.
That doesn’t mean clinicians don’t have needs (they certainly do), or even that patients don’t have duties (they may, like payment). This only means that the nature of role voluntariness lends weight to how we understand how duties are formed and directed within the relationship itself.
Doesn’t this happen in numerous other interventions - e.g., cosmetic surgery, some sports-related interventions, contraception? Possibly, probably. Each of those warrant reflection too. My focus here is on differentiating killing from letting die.
Some people, like Thomas Szasz, have taken issue with suicide prevention. That’s a topic for another time. I think it’s a good idea to prevent suicide.
Of course that doesn’t mean the clinician need throw up their hands and do whatever she says. Engaging a good faith conversation with her, her family, and other team members (e.g., mental health, chaplaincy) might assuage some of her existential distress and allow her to accomplish important tasks as she’s dying.
Joshua, this is helpful. I have a living will that says I do not want to be kept alive by heroic measures. However, that seems vague. My wife (a physician) and I have had conversations about specific scenarios (e.g., about pneumonia and ventilation) that have helped clarify what I want and do not want; these conversations have been very helpful.
So my question is: Is there a version of this essay--by you or someone else--designed for patients?