Living Wills Don't Work, Study Concludes

Living wills don't work -- and can't work -- for their intended purpose of allowing people to say in advance how they would want to be treated if they became too sick to choose for themselves, a new University of Michigan study says.

In an article in the March/April 2004 issue of the bioethics journal the Hastings Center Report, an internal medicine researcher and a professor of law and internal medicine report some provocative findings: The documents designed to help people choose the treatments they would like when dying fail to meet five key criteria for success. Worse, the evidence suggests they don't work. The researchers base their conclusions on a comprehensive review of hundreds of studies of living wills, end-of-life decisions and the psychology of making choices.

The best patients can do, the researchers argue, is to use a "durable power of attorney for health care" to appoint someone to make decisions for them when they can no longer make their own decisions.

Many experts recommend living wills to avoid the kind of problems raised by the Terry Schiavo case in Florida, a case in which a patient's husband and parents have fought a long legal battle over whether she should be kept alive.

The University of Michigan researchers say living wills may be useful for patients who are imminently facing death, who know their medical circumstances and who have strong and specific beliefs about them. But they conclude that living wills offer a false promise of control over end-of-life treatment.

"Our review shows that the evidence about living wills demonstrates that they fail all five tests that would have to be passed for them to work," says co-author Angela Fagerlin, Ph.D., a research scientist with the University of Michigan Medical School and the VA Ann Arbor Healthcare System.

"First, most people don't even have living wills. Second, those who do rarely know what care they would truly want in some hypothetical future. Third, it's surprisingly hard for people to state their wishes accurately and understandably. Fourth, the document is often unavailable when decisions need to be made. Fifth, even when it is available, surrogate decision makers usually cannot reliably apply its instructions to the patient's current health condition."

Co-author Carl Schneider, a University of Michigan Law School and Medical School professor comments, "Living wills don't fail for lack of effort, education, intelligence, or good will. They fail because of basic traits of human psychology."

For instance, studies show that people have great trouble predicting their own preferences about even simple, everyday things, like what snacks they will want or what groceries they will buy next week. "If they have trouble predicting what is familiar," asks Schneider, "why should we expect them to succeed when they are predicting what they will want in circumstances they have never experienced and can't foretell?"

The functional illiteracy of many Americans, and the difficulty even for skilled writers of expressing their wishes well, add to the problem, Fagerlin notes. And, people's preferences often change as their situations or medical technology changes.

For most people, Fagerlin and Schneider say, a durable power of attorney (DPA) for health care may be adequate. DPA's allow people to name someone whom they would want to make decisions about their care. That person is usually the patient's spouse or child, but it can be any trusted individual.

"DPA's only require a few simple choices, and they don't differ significantly from the existing system of allowing family members to make medical decisions about incompetent patients," says Schneider. "They also allow the decision-maker to use the information about the patient's condition that's available at the time a decision is needed, rather than asking the patient to guess about something far in the future. And they're inexpensive."

Download a copy of the Hastings Center Report article in PDF format.

Learn more about health care proxies.