Getting Autonomy Right

My father will be 90 next month. He lives in a nursing home because he needs help with all the most basic tasks—eating, dressing, bathing, walking. But my father also has dementia, which has been getting worse, as dementia generally does. As a geriatrician, I know that one of the few ways to make a person with dementia better is to take away any medications that might be contributing to the cognitive impairment. So when I realized my father was still taking a small dose of the tranquilizer valium (diazepam) every day, I suggested it be stopped. I was stunned when my father’s doctor called me to say he felt stopping the valium posed an ethical dilemma.

An ethical dilemma? Valium is on the most widely recognized list of medications that are generally inappropriate for use in older people. The American Geriatrics Society recommends against prescribing valium in the elderly because the drug accumulates in the body, takes days to weeks to be eliminated  from the system, and is notorious for causing confusion and lethargy, especially in people with dementia. Now guidelines and recommendations are not absolute. Valium has not been taken off the market by the FDA. And my father is only on a low dose of the medicine. For years, he was on a far higher dose because it helped control his debilitating panic attacks. When I was growing up, he used to have panic attacks every few weeks, even with high dose valium, but the attacks were far worse without the valium. Multiple efforts to find another drug that was equally beneficial without potential side effects had failed, so my father stayed on valium. Over the last few years, his doctors have tapered the dose, but they never stopped the drug entirely. Since entering the nursing home 2.5 years ago, he has remained on valium even though he hasn’t had a single panic attack. And now he is sleepy much of the time—he can’t keep his eyes open when you try to talk to him—and is more confused than ever. 

Why, then, was stopping valium ethically problematic? It might seem, by contrast, that continuing a drug that every geriatric authority deems pernicious is what is ethically problematic. But it turns out that 1.5 years ago, when my father’s mental faculties were better than they are now, though hardly normal, he told his physician that he wanted to stay on valium. So in the view of that same physician, stopping the medication now would be a violation of the all-important ethical principle of respecting autonomy.

Now I’m all for respecting a patient’s right to make medical decisions for himself. But making a medical decision assumes the patient has the capacity to weigh the risks and benefits of the treatment. And I am quite confident that my father, who is unaware that he even has cognitive impairment, cannot understand that valium could be worsening his cognitive impairment. I am sure that my father, who does not remember what he had for breakfast or even if he had breakfast, has no recollection of when he last had a panic attack. He is simply unable to evaluate the situation, and to rely on the opinion he had 1.5 years ago (assuming he understood the issues then, which is another matter), is absurd. “Respect for autonomy” is not equivalent to “the customer is always right.” It is not respectful of a person’s right to determine what happens to his body for a physician to mindlessly follow the dictates of the patient. When a cancer patient declines potentially life-prolonging treatment, should his oncologist accede to his wishes without first being sure the patient understood both what would happen if he got the treatment and what would happen if he didn’t?

There’s another problem with the assertion that my father should continue to get valium today just because he previously requested the drug. Leave aside for the moment the issue of whether he has the capacity to make any medical decisions. Even a competent patient cannot demand non-beneficial treatment. This is a point that both patients and physicians seem to have trouble accepting. Medicine is a profession and the essence of a profession is that its members have a particular expertise. Physicians are uniquely positioned to make medical diagnoses and to offer treatment based on their clinical assessment and medical judgment. Patients can decline the treatment if there are several reasonable options, and they can select which of the various possibilities makes most sense to them, given their personal views of the side effects they are keenest to avoid and the benefits they are most eager to experience. But patients cannot demand antibiotics for treatment of a viral infection or laetrile for treatment of cancer. Sometimes the treatment the patient wants is possibly effective, though only very, very rarely, and the issue becomes whether patients have a right to insist on treatments that are exceedingly unlikely to work—especially if they are fraught with danger and very expensive. But in my father’s case, it just doesn’t make sense to assert that he could demand valium if his physician believes it is not medically indicated (and his doctor agrees it is not a good drug for him), even if he had the cognitive capacity to discuss medical issues.

But perhaps even more troubling to me than the failure to recognize that the right to make medical decisions presupposes competency, and even more disturbing than the confusion about whether patients can demand non-beneficial treatment, is the profound misunderstanding about the nature of advance care planning. Now advance care planning is something I think about a great deal. It’s a topic I write about on this blog and elsewhere. Advance care planning is a process in which a person who is of sound mind indicates what approach to medical care he would want if he became ill and was unable to make decisions for himself. If advance care planning is to work, it’s crucial that when a patient articulates his wishes about future care at one point in time, and then finds himself in exactly the situation he envisioned at a later time, his physicians abide by those wishes. What’s not right—and this is what my father’s doctor worried about—would be to cavalierly ignore or even reverse a thoughtful, carefully made decision once the patient loses the capacity to make decisions. It would make a mockery of advance care planning if, for example, a patient while clear-headed said he would under no circumstances want CPR attempted, and as soon as he developed dementia, his health care proxy or physician reversed the patient’s “DNR” order. 

So at issue is what exactly my father meant when he said, 1.5 years ago, that he wanted to continue taking valium. Did he mean that he wanted to take valium for the sake of taking valium? Did he meant that if his Parkinson’s disease became so severe that he could no longer swallow, he would want a feeding tube inserted just so he could continue to get valium? Did he mean that if a better drug came along, one that was more effective and with no side effects at all, he would nonetheless prefer to keep his valium? Or was he really saying that he wanted to continue taking valium as long as it effectively controlled his panic attacks, and he was willing to put up with some minor side effects in return for this benefit? If so, then it follows that if the valium is no longer needed for control of panic attacks, there is no reason to keep on taking it.

The story has a happy ending in that my father’s physician acknowledged my reformulation of the earlier statement “I want to keep taking valium” was plausible. The doctor will discontinue the valium on a trial basis. Now it’s conceivable that the panic attacks will return. Perhaps the reason my father hasn’t had any such episodes in the last 3 years is that the valium has been working wonders—even though for the previous 30 years, valium never ablated the panic attacks, it merely attenuated them. In that case, the doctor will again prescribe valium for my father. It’s also quite possible that discontinuing valium will have no discernible effect—either on panic attacks or on my father’s cognition. He will remain just as befuddled and sleepy as he is now. But at least we will not have uncritically adhered to something my father once said, in a misguided attempt to respect his autonomy.

While this particular “ethical dilemma” looks as though it will be resolved, I worry about all the other physicians who fail to appreciate the importance of ascertaining the patient’s understanding of the issues before blithely following his wishes; I worry about all the patients and physicians who do not distinguish between refusing unwanted medical treatment (a legitimate patient right) and demanding non-beneficial treatment (not a patient right); and I worry about how advance care planning is implemented. The latest variant of advance care planning is the POLST form (physician orders for life-sustaining treatment) and related efforts with similar names (in Massachusetts, it is MOLST), which is actually a medical order. In this approach, patients decide together with a clinician what specific treatments they would accept and which they would not (this can include CPR, ventilators, dialysis, feeding tubes, intravenous fluids or other interventions) and sign an order to this effect, an order that is supposed to be binding on all medical personnel: on EMTs in ambulances, nurses in nursing homes, and physicians in emergency rooms and hospitals. It’s a valiant attempt to make sure that patient autonomy is respected. But whether it will achieve this laudable aim depends on whether patients have the capacity to make decisions when they sign the form and whether they truly understand the underlying issues. Today, patient’s wishes are often not respected. Tomorrow, we may be imputing “wishes” to patients that they never had.