Live Free or Guardianship: The Appropriate Standard
Note: We aren’t lawyers and Canmore Ethics does not give legal advice.
Summary: The Supreme Court of New Hampshire ruled that a guardian should disregard a patient’s stated wishes and make medical decisions according to the guardian’s view about the patient’s best interests. This goes against the consensus view in bioethics that wishes ought to be respected.
The Supreme Court of New Hampshire recently broke with bioethical consensus in a case about substitute decision-making. While this will have little effect outside of New Hampshire, it’s a significant change with little ethical justification.
D.E.
The case concerns D.E., a 79-year-old man who lacks some decision-making capacity because of a major neurocognitive disorder and an anoxic brain injury. D.E. was assigned a guardian because, in the court’s opinion, he is unable to make decisions about his healthcare (i.e., he is unable to consent to or refuse treatment), finances, or estate.
While D.E. was in an intensive care unit last year, and before he had a guardian, his children executed a Do Not Resuscitate (DNR) order for him. (It is implied that this occurred when D.E. was unconscious or otherwise unable to make his own medical decisions.) Later, D.E. improved and told his children that he wanted the order removed, though it is unclear what his capacity was at this point. A guardian typically has the ability to consent to or refuse treatment on behalf of their ward. However, when the lower court assigned D.E. a guardian, it prohibited the guardian from executing a DNR order for D.E. because of his consistent refusal of DNRs.
Since then, D.E. has been moved to a nursing home, where he remains conscious and continues to refuse a DNR. He has also repeatedly refused dialysis, which will quickly lead to severe health outcomes. Because of this, the guardian asked the court to allow her to execute a DNR. Since D.E. is at high risk for a cardiac arrest, a DNR will prevent the considerable pain of a resuscitation attempt that will most likely have little benefit.
The question considered by the Supreme Court of New Hampshire is whether the guardian should be given the authority to execute a DNR, and, if so, what standard she should apply in making this type of decision.
Two Standards
Ethically, there are two standards substitute decision-makers (SDMs) should use to make decisions on behalf of the person they are representing. In most places, SDMs are instructed first to make decisions in line with the patient’s wishes, if known. This is often called the ‘substituted judgement’ standard. If there are no known prior wishes, then SDMs are to use the best interest standard.
It is established clinical ethics practice to give the most weight to the patient’s known wishes. A lot of the work ethicists do in cases of this sort is to find out about those preferences and then guide SDMs and the medical team to act on them, if they’re justified. The reason for this emphasis is that both bioethicists and the general public broadly agree that people should be able to make decisions about what will happen to them and their possessions even when they’re no longer able to make or voice these decisions for themselves. Advance directives, living wills, and estate wills all fulfil this purpose. It’s only when one’s preferences aren’t known that the best interest standard takes over.
The Supreme Court of New Hampshire ruled that only the best interest approach should apply, meaning that guardians should not make decisions based on the patient’s previously expressed wishes.
It’s worth noting that, despite consensus, there’s a case against prioritizing prior wishes. Suppose on my 18th birthday that I hope to get married before I’m 30, but when I’m about to turn 30, I’m not ready to get married. Is there anything bad about my hope going unfulfilled? It seems it would only be bad if I still want to get married by 30, and see my bachelorhood as something to regret. But it doesn’t matter what 18-year-old me thought; he didn’t know what he was talking about. All that matters is my current situation. People change in big and small ways, so, on this view, we should focus on the present.
Perhaps the same is true for healthcare. People are bad at forecasting how they will feel in different situations, and healthcare situations are particularly hard to forecast. From the outside, we might say “I never want to live like that”, but when we’re actually in that situation, it might not be that bad. The classic example of this is a study that compared recent lottery winners with recent victims of paralysis, and found that both groups adapt to the situation and report similar happiness levels. Therefore, perhaps we should give less weight to prior wishes and go with best interest.
The problem with this approach is that it means we are to ignore all prior wishes. Advance directives make no sense on this view. For most people, that seems wrong. Another problem, more pertinent for D.E.’s case, is that the court is saying that even his current refusals lack importance. This goes against consensus, according to which even uninformed refusals (sometimes called ‘dissent’) should be recognized some of the time.
An additionally striking feature of this case is how it deviates from current legal approaches.
What the Law Says
In many jurisdictions, substituted judgement is given priority, though there isn’t always a clear delineation. For example, in Alberta, SDMs must decide in the patient’s best interest, but consideration must be given to the patient’s wishes, values, and beliefs. In Texas, guardians are instructed to put the patient’s preferences first, then act in their best interest. In this case, it’s important to note that prior wishes no longer apply if following them would cause “substantial harm”:
When making decisions on behalf of a ward, a guardian must strive to act in accordance with the known or ascertainable preferences of the ward, including preferences expressed in valid estate planning documents, unless to do so would cause substantial harm to the ward’s person or property, in which case the guardian should make decisions that are in the best interest of the ward.
Remarkably, New Hampshire itself has a similar provision for SDMs:
After consultation with the attending physician or [advanced practice registered nurse] and other health care providers, the agent or surrogate shall make health care decisions in accordance with the agent’s or surrogate’s knowledge of the principal’s wishes and religious or moral beliefs, as stated orally or otherwise communicated by the principal, or, if the principal’s wishes are unknown, in accordance with the agent’s or surrogate’s assessment of the principal’s best interests and in accordance with accepted medical practice.
The situation with D.E. seems to turn on the fact that he has a guardian, not an agent or surrogate. D.E. did have a surrogate when he was in the ICU—his two children who executed the DNR—but they were replaced by a guardian. (It isn’t clear why D.E. was assigned a different guardian instead of one or both of his children.) He doesn’t have an agent because he never completed an advance directive.
The reason the court gives for this distinction is, they say, that the state’s power in this kind of situation is grounded in its power and obligation to protect the vulnerable. Since we aren’t legal experts, we can’t weigh in on why this would justify guardians in ignoring stated wishes. The distinction between guardians, agents, and surrogates seems minor. The ruling suggests that, were D.E. to have a surrogate instead of a guardian, or were he to have made an advance directive, then his wishes should be followed. This is weird.
What Should We Make of This?
Here’s how one could argue that it’s ethically acceptable to execute a DNR order for D.E., though it depends on some details that aren’t clear in the ruling. Although he has consistently rejected a DNR, it appears that these refusals have only taken place following his loss of capacity. Since this capacity loss is significant enough to justify guardianship, D.E. lacks the ability to assess the risks and benefits of resuscitation. In other words, the whole point of guardianship is to protect D.E. by making decisions that are in his interests, which he is unable to do. His insistence on resuscitation is just one example of this.
The issue with this approach is that it fails to account for any of the reasons D.E. might be refusing the DNR, or how they might align with his wishes from before he lost capacity. It would be different if he were unconscious, but the case description indicates that he consistently answers questions about code status, so perhaps he has some insight on this point.
One response in favour of ignoring his wishes is that D.E. appears to be acting inconsistently by refusing dialysis, which will quickly lead to his death, but also refusing to accept a DNR, which will harm him with no clear benefit. Since D.E. lacks capacity to the extent that he requires a guardian, inconsistency shouldn’t be surprising. However, we don’t know why he’s refusing dialysis. In any case, even if we decide that some particular request shouldn’t be granted, that’s still a long way from the court’s claim that his wishes are irrelevant.
The details on D.E.’s situation might move the needle one way or the other, but the Supreme Court of New Hampshire isn’t focused on these details in their ruling. Instead, they’re making a general claim that the best interest standard is the appropriate approach for guardians. By framing the role of guardians in this way—the court says that it’s an ‘objective’ approach—there is no room for the patient’s wishes, values, and beliefs.
This means that a person who was a dedicated vegan her whole life can be given animal products if it’s deemed in her best interest, if D.E. were a Jehovah’s Witness he could be given blood, a patient who never wanted to be kept alive on a ventilator could be so treated, and so on. It just depends how ‘best interest’ is unpacked.
Sometimes a patient’s preferences are unknown. In those cases, the best interest standard is the correct approach. But the rest of the time it’s wrong to ignore them.
