Home Q&A One researcher’s view of what’s wrong with living wills

One researcher’s view of what’s wrong with living wills

April 2005

Published in the April 2005 issue of Today’s Hospitalist

Talk to Angela Fagerlin, PhD, about the Terry Schiavo case, and you’ll hear what sounds like a contrarian view of living wills. While many pundits hold the Schiavo case up as a textbook example of the value of advance directives, Dr. Fagerlin has an entirely different take. In her mind, the case is yet another example not only of why living wills don’t work, but why the U.S. health care system needs to take a new approach to decision-making at the end of life.

In a report on living wills published late last year by The Hastings Center, in fact, Dr. Fagerlin bluntly describes living wills as a “triumph of dogma over inquiry, of hope over experience.” A research scientist at the Ann Arbor VA Health Services Research and Development Service who spends much of her time researching end-of-life care issues, Dr. Fagerlin says the problem is simple: Living wills don’t work. They don’t help surrogate decision-makers fulfill the wishes of their loved ones, and they may also contain incorrect information.

While that may sound like a harsh assessment, Dr. Fagerlin says that study after study proves not only that living wills don’t work, but that they don’t reflect the type of decision-making that most people say they want at the end of their lives. That, combined with the crushing cost of advance directives to hospitals, has led her “and a number of end-of-life experts “to conclude that living wills don’t work.

In an interview with Today’s Hospitalist, Dr. Fagerlin, who is also a researcher at the department of internal medicine at the University of Michigan, makes a case for why health care needs to move beyond living wills in end-of-life decision-making.

Why are you so convinced that living wills aren’t effective in end-of-life decision-making?

We outlined a number of steps that are necessary for living wills to live up to their mandate, which is to help surrogate decision-makers reach the same decisions for incompetent patients that the patients would have made for themselves.

One of the first steps is that people need to complete a living will. Unfortunately, the vast majority of Americans do not have living wills.

Even when people do fill out living wills, they have a really difficult time predicting what they want in the future. Think about when you go into the grocery store for one or two things, and you end up buying 10 different items.

We think we can predict what we’re going to do, but we often fail. When it comes to health care, we have an even more difficult time predicting what we’re going to want in the future because we are talking about things most of us have never experienced.

How often do living wills make it into the hospital?

A study in 1995 showed that when a group of patients who had living wills came to the hospital, only 26 percent had a correct notation in their medical record indicating that they had a living will. In addition, less than 20 percent actually had their living wills in their files in the hospital.

When living wills are both completed and available, do they help surrogate decision-makers honor patients’ wishes?
There has been no evidence showing that having a living will makes a surrogate decision-maker any more accurate at predicting someone’s life-support preferences. I worked on a study published in 2001 in the Archives of Internal Medicine, which was the only study I know of that examined whether living wills actually improve the decision-making of surrogate decision-makers. That trial showed that living wills made absolutely no difference in surrogates’ accuracy when it came to predicting peoples’ life-support preferences.

Studies have also shown that without a living will, surrogate decision-makers can predict someone else’s life-support preferences about 70 percent of the time. Living wills do not bring that number up at all.

Why don’t living wills help surrogate decision-makers?

Living wills tend to be vague and are often not related to the person’s current health condition, which makes it almost impossible for them to apply information in the living will to the current health situation.

In the Terry Schiavo case, for example, a living will using typical language “”I would not want heroic treatment in a persistent vegetative state” “probably wouldn’t have helped, because her parents claim she’s not in a persistent vegetative state. They could say they have to disregard the living will because it does not cover her health condition.

Do living wills offer any help to physicians, who may not know patients well enough to predict end-of-life care preferences?

A study in a 2001 issue of the Archives of Internal Medicine examined whether living wills could increase the ability of physicians to predict the preferences of patients. That study compared a group of family physicians and internists to a group of emergency room physicians and hospital-based physicians. It found that living wills didn’t improve the ability of general medical physicians to predict their patients’ end-of-life care preferences, but they did help the hospital-based physicians.

Because the hospital-based physicians wouldn’t have known these patients, the living wills helped them predict what the patient would have wanted. But because the general physicians knew the patients and their preferences, the living will didn’t add anything to help predict their preferences.

If living wills offer so little help, why are they so popular?

People want to hold out hope that living wills work because we have such an autonomy-based culture, particularly in medicine. We encourage patients to make most, if not all, of their health care decisions. In many aspects, that’s good, but we can’t imagine a situation in which somebody is not able to have their wishes respected.

When it comes to living wills, people think that if we just work harder, we can make living wills work. I fundamentally disagree, because you’re trying to change human psychology. We’re asking people to identify what they would want in a situation they’ve never seen before, so they have no experience on which to predict their preferences.

What’s an alternative to living wills?

There have been a number of studies in which the vast majority of Americans say that they don’t really care what decision is made for them. They say that their main concern is that the person they trust most in the world make a good decision and feel comfortable with that decision. They’re more concerned with the surrogates feeling good about the decision, because they trust the surrogate to make the decision that is in their best interest.

In those cases, all you really need is a durable power of attorney for health care, where one person is authorized to make the decisions that they think the patient would want.

What are the financial implications of living wills?

One concern with living wills is that they’re very costly, both for patients and hospitals. Patients can download them from the Internet, but you usually have to pay for them, or you have to pay a lawyer.

There are even bigger costs for hospitals. Every time a patient comes in, even if it’s only for an outpatient procedure, hospitals are required to ask if that person has a living will. Thousands of dollars are spent every year in the dissemination of living wills.

To put this kind of financial burden on hospitals when there’s no evidence that living wills will accomplish their goal is a waste of money. We could be using that money for things that we know are effective, like childhood vaccines or extending health insurance to the uninsured.

What advice can you offer hospitalists who are trying to care for patients with living wills?

Going back to the research, a majority of people have repeatedly said that they want their families to feel good about end-of-life decisions, that they trust their family members to make the right decision. So I would say that if there’s a little discrepancy between a living will and what the surrogate is saying, go with what the surrogate is saying.

Studies have shown that peoples’ end-of-life care preferences sometimes change within a year or even a couple of months. Because preferences can change so much, especially when a living will is a couple of years old, there’s no reason to believe that it actually represents what people want.

For more information

“Enough: The failure of the living will” appears on the Hastings Center Web site. (To find the report, type the name into the Web site’s “Search” box.) The report, which was co-authored by Angela Fagerlin, PhD, and Carl Schneider, JD, contains an extensive bibliography of end-of-life care readings.