Research says living wills won’t guarantee patients’ wishes  

Michigan researchers say Patient Self-Determination Act a failure

Since passage of the Patient Self-Determination Act (PSDA) in 1990, the living will — a form of advance directive that spells out the signer’s wishes for end-of-life care and termination of care — has become an almost automatic subject in any discussion of death or resuscitative medicine.

But is the living will the useful tool that polls indicate most Americans believe it is? Two University of Michigan researchers say it’s not and recently published a report of their evidence-based research they say shows, when it comes to our faith in living wills, Americans are choosing a feel-good step that has no proven value.

The continued urging of adults to draw up living wills is "but the triumph of dogma over inquiry, and hope over experience," wrote Angela Fagerlin, PhD, a faculty member at UM’s Ann Arbor (MI) VAMC, and Carl E. Schneider, JD, the Chauncey Stillman Professor of Law at the UM Law School and professor of internal medicine at UM.

Fagerlin and Schneider published their report, "Enough: The failure of the living will" in 2004, and say it is the first evidence-based study of what they say many in health care and health care law already knew — that living wills do not accomplish what most people think they will accomplish.

An unfunded mandate

Living wills and other forms of advance directives became commonplace with the passage of the PSDA and soon were the source of a great deal of new paperwork and expense for hospitals and, ultimately, patients, Fagerlin says.

"One thing that concerned us was that, when the PDSA was enacted, no research was done to see if living wills would be an effective tool. And Congress didn’t fund the PDSA, so suddenly hospitals were required to [fund it]," she explains.

Living wills are recognized in all 50 states, but to varying degrees. Some state courts uphold living wills as legal documents, while others have no laws recognizing them as anything other than an expression of a patient’s wishes.

The PSDA encourages all people to make choices and decisions now about the types and extent of medical care they want to accept or refuse should they become unable to make those decisions due to illness. The PSDA requires all health care agencies receiving Medicare and Medicaid reimbursement to recognize the living will and power of attorney for health care as advance directives. Under the PSDA, health care institutions must ask patients whether they have advance directives and must provide patients with educational materials about their rights under that particular state’s laws.

All that costs money, and the cost issue is a key component to what Schneider and Fagerlin see as the failure of living wills.

"Living wills consume patients’ time and energy, [and] when doctors or lawyers help, costs soar," they wrote. One study determined that the PDSA imposed on all hospitals a start-up cost of $101,569,922; expenses continue as the program is administered.

"I wish this would see some policy change. We have 45 million people without health care, but we spend millions on making copies and processing forms," says Fagerlin.

If living wills were free, the authors say, their lack of effectiveness would not be such a problem. Conversely, if they were more effective, the cost of creating them and administering the PDSA would be more justifiable, Fagerlin says.

"But there is no convincing evidence that living wills reduce the cost of end-of-life care," she says.

Zero-for-five on the success scorecard

Fagerlin and Schneider devised five conditions to determine what the living will would need to accomplish to function as its creators intended. Out of those five, the authors say, evidence shows that living wills in general meet none. The conditions, and the authors’ findings, are:

1. People must have living wills. Most people don’t have them. Surveys indicate that while people think living wills are a good idea, only about 18% of adults actually have living wills. "People don’t have living wills probably because they trust their families to make decisions for them," Fagerlin says. "One study asked people if they had made an advance health care decision, and their family thought it was a bad one, would they trust their family to overrule them. Most said they would."

2. They must decide beforehand what treatment they would want should they become incompetent to decide. "People who have living wills, we found, hadn’t really thought through what their instructions are in a way that a life-or-death decision would demand," Fagerlin says. Furthermore, the authors found that patients’ wishes changed with their circumstances. What they thought they wanted before they became sick changed once they were actually in the hospital and facing a medical crisis.

3. The writers must accurately and lucidly state what their preferences are. Often, patients’ proxies or family members were not able to accurately state, from what was written in the living wills, exactly what the patient’s wishes would be in a given situation.

4. The living will must be available to the people making decisions for a patient. In many cases, living wills are drafted, signed, and stored away years before they are ever needed, and when a medical crisis arises, the person making decisions does not have the living will at hand. Also, 62% of patients with living wills never give copies of them to their physicians, so their physicians don’t always know their wishes.

5. People must grasp and heed the living will’s instructions. The Michigan researchers wrote that a 1998 study on effects of advance directives found that living wills do not alter care. Hospital and ICU lengths of stay, as well as health care costs, were similar for patients with and without advance directive statements.1 Another study found that in 30 of 39 cases in which a patient was incompetent, and the living will was in the patient’s medical record, the surrogate decision-maker was not the person the patient had appointed.2 In a third study, care received by one-quarter of the patients was inconsistent with the wishes spelled out in their living wills.3

Fagerlin says there are three primary reasons living wills are not successful in meeting their intended goals: 1) they are not interpreted correctly; 2) the patient is never recognized by his or her surrogate as being hopelessly ill, and 3) family members are not present or are unable to advocate for the patient.

The five conditions for a successful living will, Schneider and Fagerlin wrote, "are unmet and largely unmeetable."

Fagerlin says one question she continues to pursue after 10 years of research into end-of-life issues is, "Can we really know what we will want until we are actually in a situation?"

She says her studies have shown people’s beliefs change with time, change in life circumstances, and change in times of medical crises. So expecting someone to be able to accurately predict what they will want at the end of life, unless that time is imminent, is unrealistic.

"There’s something about what happens in a hospital that can make people change," Fagerlin says. "They may not think they’d want life support before, but when they’re in the hospital, they can change their mind. Then, afterward, after some time has passed and they’re out of the hospital, their beliefs return to what they were before.

"So how can you know what your true preference is? It’s a fascinating question."

Power of attorney may be more useful

Schneider and Fagerlin say they aren’t advocating the total elimination of living wills. They say living wills can be useful in times of crisis when the patient has strong, specific, delineative wishes, and has special reason to prescribe his or her own care.

Durable power of attorney for health care, however, is a broader, more flexible, and — in the eyes of the states — more weighty instrument for protecting patients. Fagerlin says durable powers of attorney are probably a better choice for most patients but still are reliant upon the designated proxy to know just what the patient would want done on his or her behalf.

Fagerlin and Schneider’s report concludes with a call to drop emphasis on living wills, promote powers of attorney, and repeal the PSDA.


1. Goodman MD, et al. Effect of advance directives on the management of elderly critically ill patients. Crit Care Med 1998; 26:701-704.

2. Morrison RS, et al. The inaccessibility of advance directives on transfer from ambulatory to acute care settings. JAMA 1995; 274:478-82.

3. Danis M, et al. Advance directives for medical care: Reply. N Engl J Med 1991; 324.

Further reading


  • Angela Fagerlin, PhD, Faculty, Program for Improving Health Care Decisions, Division of General Medicine, Ann Arbor VAMC, University of Michigan, 300 N. Ingalls, Room 7C27, Ann Arbor, MI 48109. Phone: (734) 647-6160. Fax: (734) 936-8944. E-mail: