Earlier this year the Supreme Court of the State of New York issued an opinion that reminded me of an old joke.

A young law student asks his father, an attorney in practice for decades, to give him some expectation of what he will learn in the upcoming course on ethics. His father offers an illustrative example. You and a classmate pass the bar and go into practice together. A lady comes in to see you about a minor adjustment to her will. She returns the next day to receive her papers and hands you a crisp, new $100 bill in payment for this little bit of work. As she is leaving the office you discover that she actually gave you two brand new $100 bills stuck together. Now you have an ethical dilemma. Do you tell your partner about the extra $100 you made?

A man sought treatment for alcoholism in a hospital emergency department. He then changed his mind and decided to leave. He was apparently quite intoxicated (and the record included a “very high” blood alcohol level). He left, wandered onto a highway, was struck by a vehicle, and was left quadriplegic as a result of his injuries. He sued, claiming he should not have been permitted to leave the emergency department.

The Court said the defendants (physician and hospital) had neither the authority nor the duty (under New York law) to detain the patient.

This decision has been hailed by some of my colleagues as a victory for common sense and for the principle that there is such a thing as personal responsibility. And I must admit to being pleased that the doctor and the hospital were not held legally responsible for this tragedy. Further, I can easily imagine circumstances under which medical personnel could not possibly be responsible. For example, if the patient seemed cooperative and fully desirous of treatment for his addiction and then slipped out of the building undetected, it would be quite unreasonable to hold them to account for subsequent events.

Nevertheless, this case offers an excellent example of how ethical and legal analysis can lead to quite different conclusions.

Often when we’re trying to figure out the right thing to do, we can frame the question as an ethical dilemma. Sometimes that means there are competing interests that must be weighed against each other. Sometimes, as in a case such as this, there are principles of biomedical ethics in conflict, and we must decide which takes precedence.

‘While I consider it important to avoid violating the law in my practice of medicine, I consider it even more important to avoid violating principles of biomedical ethics.’

The two principles here are beneficence (the duty to help the patient) and autonomy (the patient’s right to make his own decisions about what happens to him). In the United States (and some other Western societies) autonomy is placed at the top of the hierarchy. This means if one is going to do something that violates patient autonomy, there must be a very good reason. The most common such reason is that the person is mentally ill and intends to harm himself or others. In such a situation, there is agreement that the duty to protect the patient (and possibly others) from harm takes priority over the duty to respect his individual autonomy.

Another common situation is one in which a patient wishes to refuse treatment or to leave a hospital against medical advice. Sometimes such a decision may place the patient in peril of serious harm. So, for example, if someone is having a heart attack, there is a very real risk of sudden death or severe, permanent disability attendant upon a decision to refuse treatment and leave the hospital. But we do not violate the patient’s autonomy and prevent him from leaving, except….

We make an exception for lack of decisional capacity. And to do that we must understand what decisional capacity is and how to assess it.

Let us use the example of the heart attack. If you are having a heart attack and decide to refuse treatment and leave the hospital, my first priority is to try to change your mind. I will explain my treatment recommendations and the risks you are assuming by rejecting them. I will enlist the aid of those whose opinions or feelings mean more to you than mine: your family or friends, or your personal physician. (Maybe even your nurse, because she has impressed you as a warm and caring person, and you have no suspicion that anything she tells you is motivated by pecuniary gain.)

I will try to find out why you want to ignore sound medical advice and whether you have concerns that can be effectively addressed. If you say you cannot stay in the hospital because there will be no one to feed your dog, I will try to figure out how we can get somebody to feed your dog. I will point out that if you die, your dog will be worse off than going hungry for a little while and tell you that in the wild, dogs go for days without eating, and that’s why even domesticated dogs typically eat like they’ve had nothing to eat for a long time, and don’t know when the next meal is coming, every time a bowl of food is put in front of them.

One of the things I must do when someone is making a really bad (meaning high-risk) decision is assess decisional capacity. Does the patient understand the reason for my recommendations? Does the patient understand my explanation of the risks of not following those recommendations? Is the patient genuinely willing to accept those risks? Can the patient explain why s/he is rejecting the advice and why s/he thinks it makes sense to accept the risks?

It should be apparent that, at the same time this process enables me to assess decisional capacity, it also allows me to make sure there have been no misunderstandings of the information I’ve conveyed and to correct any there may be.

All of this requires significantly time-consuming verbal communication. I must be satisfied that the patient really understands my advice, and the risks, and that requires that the patient explain them back to me in his own words. I also need him to explain his reasons for rejecting the advice. His reasons for rejecting the advice don’t have to satisfy my sense of what is logical or rational or makes sense. So, for example, if he gives me a list of recently deceased friends and relatives and tells me they all died in hospitals, and so there is no way he is going to stay in a hospital, and he rejects all my attempts to demonstrate the logical fallacies in his reasoning, that doesn’t mean he lacks decisional capacity.

If, on the other hand, he tells me if he stays in the hospital, as soon as he falls asleep tonight, space aliens or demons from Hell will appear in his room and abduct or incinerate him, I will be inclined to think he is delusional and mentally ill and does not have decisional capacity. (For the sake of clarity, I should note that a delusion is not a belief that is merely false. Rather, it is a belief that could not possibly be true.)

If his decision is based on values that are different from mine, I must respect that. For example, he may be perfectly comfortable with the risk of death and prefer to take his chances and die at home rather than be hospitalized and subjected to modern medical technology. I may then wonder (and even ask) why he came to the hospital in the first place, and my values may be completely different, because if I were similarly situated, I would want to live to see my grandchildren graduate from college and get married.

But those are my values, and I have neither the right to impose them on him nor any reason to allow that difference to cast doubt on his decisional capacity.

Getting back to the particular case in New York, it is important to understand that a high blood alcohol level does not necessarily mean that a patient lacks decisional capacity. People who are habitual drinkers are often lucid and able to carry on coherent and rational conversation with blood alcohol levels that would render me unconscious.

They are also often able to walk with a steady gait and cross streets carefully and safely. So, when we are deciding when a person who has been brought to the ED drunk can be safely discharged, we base that on whether the person can “walk and talk” normally (among other things), not on any number.

So let us assume, for the sake of argument, that the man in the New York case made his intent to leave known to ED personnel rather than deceiving them and slipping out of the building unnoticed. Did they consider his decision to be against medical advice? Did they think it was unsafe for him to leave? Did they assess his decisional capacity? Did they consider restraining him (either with tranquilizing drugs or with physical restraints), thus violating his autonomy in order to protect him? (Such decisions are not always straightforward, because restraints of either kind carry their own risks.) These are all the questions I would ask if I were reviewing the case and judging it for its adherence to principles of biomedical ethics.

While I consider it important to avoid violating the law in my practice of medicine, I consider it even more important to avoid violating principles of biomedical ethics.

If I obey the law but violate professional ethics – and I’m not saying that’s what happened in the New York case, as I think should be clear by now – that will not help me to sleep at night if there is a tragic outcome.

Dr. Solomon has served on hospital ethics committees for 25 years and on the ethics committee of the American College of Emergency Physicians for 15 years. He teaches emergency medicine to residents at Allegheny General Hospital in Pittsburgh and is Medical Editor in Chief of ACEP News. He is a social critic and political pundit and blogs at www.bobsolomon.blogspot.com.