Defective directives? Struggling with end-of-life care

Making health care decisions for patients who cannot is emotionally wrenching for families and poses an ethical dilemma for physicians. In a medical system that puts a premium on individual autonomy, what is to be done when patients lose the capacity to decide on care?

For more than two decades, the answer has been to avoid ever getting to that point. Patients are encouraged to spell out in writing what kinds of life-sustaining care they would find acceptable. Living wills, also known as substantive or instructional advance directives, are intended to give patients prospective autonomy over decisions such as whether to be resuscitated, intubated or artificially hydrated.

But according to a growing body of research, there are serious shortcomings with this type of directive.

For one thing, few patients have advance directives. Although the Patient Self-Determination Act of 1990 requires most health care organizations to inform patients of their right to make advance directives, fewer than a third of Americans have a living will. Only half of patients with terminal illnesses have directives.

Moreover, physicians often lack access to advance directives. Three in five patients with living wills do not give them to their doctors, and families often are unaware of whether their loved ones have an advance directive. About 40% of hospitalized patients have some diminished decision-making capacity.

Even when patients have medical directives and doctors can access them, the instructions often are vaguely worded and fail to match the clinical realities. What does it mean, for example, to have "no chance" of survival?

Living wills "have an important place in advance care planning for the end of life, but they are an incomplete solution," said Alexia M. Torke, MD, assistant professor of medicine at Indiana University School of Medicine. "Even when patients do have them, they often don't clearly answer the questions at hand about life-sustaining care. So surrogates and physicians still have to work together to figure out what to do."

John D. Lantos, MD, agreed. "There's always a question of what exactly the person meant and whether their current clinical circumstances warrant a change in plans," said Dr. Lantos, John B. Francis Chair in Bioethics at the Center for Practical Bioethics, a Kansas City, Mo.-based think tank.

Many studies have found surrogates are about as accurate in predicting loved ones' wishes when they have no written instructions as when they do, raising questions about the value of advance directives.

For example, a Feb. 12, 2001, Archives of Internal Medicine article reported the randomized controlled trial of more than 400 patients, 65 or older, and their designated health care proxies. Surrogate-patient pairs filled out questionnaires about what kind of care the patient would want in various scenarios, such as being in a six-week coma after a stroke with no cognitive functioning and a "very slight chance of recovery."

The surrogates who had guided discussions with patients as they filled out advance directives were just as likely to predict correctly the patients' preferences as surrogates who had no interaction with patient partners who did not complete advance directives.

"There are so many contingencies in medical scenarios that you can't put them all down in a living will. You are putting a lot of undue pressure on surrogates" to correctly interpret advance directives, said Angela Fagerlin, PhD, associate professor of internal medicine at the University of Michigan Medical School. She co-authored the Archives paper and has used her training in experimental cognitive psychology to study medical decision-making.

"When I interviewed people about their treatment preferences, a lot of times they said, 'I don't want to be a burden on my family. This is not just a decision about me, but about how it will affect my whole family.' Most people don't have extremely held beliefs about what treatments they want," Fagerlin said. "They just want the person they love and trust the most to make good decisions for them."

Perhaps the biggest problem with this type of advance directive is that patients often change their minds about life-sustaining treatment. After illness and hospitalization, three in 10 patients adjust their views, desiring more or less aggressive care than they previously thought they wanted.

If patients' own preferences are so unstable, then how reliable are their advance directives as a guide to what they would have wanted?

"People have a hard time anticipating the care settings in which they'll face decisions in the future," said G. Caleb Alexander, MD, assistant professor of medicine at the University of Chicago Pritzker School of Medicine. "You can quote Yogi Berra: 'It's tough to make predictions, especially about the future.' "

Advance directives are "well-intentioned," Dr. Alexander said, "but they often fall short of the mark. There's a wealth of evidence, literally hundreds of studies, that document serious problems with the design and implementation of advance directives.

"When living wills and advance directives were first developed, there was a real hope that they would substantially revolutionize the care of the dying. But I think the past decade shows that they haven't."

Dr. Alexander, who has written about surrogate decision-making, said his "goal is not to have every patient complete an advance directive but to ensure that patients and families are on the same page and that I have a general sense of the patients' approach."

He does not tell patients advance directives are a waste of time. Rather, he emphasizes the value of advance directives as a conversation starter.

Surrogate solution?

Dr. Lantos, of the Center for Practical Bioethics, said patients with strongly held views about various life-sustaining treatments should pursue living wills. But, he said, "everybody ought to name a surrogate decision-maker," using another form of advance directive known as durable powers of attorney for health care.

The American Medical Association strongly supports advance directives and durable power of attorney for health care.

The durable power directive does not set out specific instructions about ventilation, artificial hydration and so on, but instead requires only that patients name loved ones they trust to make treatment decisions, should they become incapacitated.

The American Bar Assn.'s Commission on Legal Problems of the Elderly says "most people should have a health care power of attorney" and lists as a "myth" the notion that "everyone should have a living will."

Rebecca Dresser, professor of law and medical ethics at Washington University in Missouri, said that designating a health care proxy is the best of a bad set of choices.

"Having someone you trust, maybe a couple of people, at the bedside who know you and can deal with the up-to-date situation, it's not ideal," Dresser said. "Ideally, we'd have the patient tell us, but since we often can't have that, then that's probably the next best alternative that's workable."

When patient surrogates tell physicians what they believe their loved ones would have wanted, it is known as substituted judgment. Fagerlin's Archives study and others have found that surrogates' substituted judgment is correct in seven of 10 treatment scenarios. So while naming a loved one as a surrogate decision-maker may have advantages over instructional advance directives, it is not flawless.

Another way to approach the decision

With living wills often nonexistent or inapplicable and surrogates often wrong about patients' wishes, how should physicians respond?

"In the absence of advance directives, doctors and family members both are forced to fall back on the best interests of the patient," said Dr. Lantos. "They may temper them as much as possible with guesses about what the patient would have wanted, but it's kind of fictional."

Dr. Torke, of Indiana University, conducted a series of in-depth structured interviews with physicians at a Midwestern medical center about their experiences with surrogate decision-making, as reported in the Journal of Clinical Ethics' Summer 2008 issue. She found physicians tried to find out what surrogates knew about patient wishes, but also attempted to see what surrogates' own wishes were and made their own clinical and ethical judgments about patients' best interests. Both were departures from the strict patient autonomy model.

"Because there is so much focus on autonomy today, we have gotten away from a primary goal of medicine, which is to act in the best interests of patients," Dr. Torke said. "When physicians face surrogate decision-making, there is some ambiguity about which of those goals they should look at first."

In pediatrics, physicians often use judgment about treatments that are in patients' best interests and without recourse to the patients' own wishes, said Dr. Lantos. That kind of ethical decision-making also occurs in end-of-life care, he said.

Physicians should work with surrogates to learn more about patients' life stories, he said, in an attempt to take actions that respect patients' dignity and personalities.

The autonomy model is "much easier to live with," Dr. Lantos said. "I'm not letting mom die. I'm simply carrying out her wishes." Yet, he added, doctors and families muddle through as best they can, and it is time ethicists acknowledged the clinical reality.

"These are hard decisions for anyone to make."