When Advance Directives Are Ignored

In the final two years of her mother’s life, Lisa McAllister started carrying copies of her mother’s do-not-resuscitate and power-of-attorney forms in her purse and her car. Her mother was cycling between the hospital, assisted living and rehab — “a round robin of craziness,” said Ms. McAllister, a retired copy editor — and was clear about not wanting to be kept alive with substantially diminished quality of life.

One night in 2012, after her mother had fallen in a nursing home in Silver Spring, Md., Ms. McAllister looked at her mother’s medical chart and saw that someone had checked the box that said, “No D.N.R.”

“I said, ‘There’s a D.N.R.,’ ” Ms. McAllister recalled. “The floor nurse said, ‘No, there isn’t.’ ” Ms. McAllister got testier. “There is, it’s in the file. I’ve seen it there.”

They went to the director of nursing, who checked the records — and yes, there was the D.N.R. form.

It wasn’t the only time Ms. McAllister had to intervene to stop aggressive treatments. “It was a common occurrence,” she said. “There’s a lot of bodies and a lot of files and people coming and going.”

Health care professionals, ethicists and advocates continually urge older people to document their preferences about end-of-life medical decisions, and a growing proportion do. A recent large national study, tracking more than 6,000 people over age 60 who died between 2000 and 2010, found that the proportion with advance directives climbed to 72 percent from 47 percent.

But how much does it matter? How often are people’s advance directives ignored or overridden? Do they really ensure that the elderly get the care they want or ward off procedures they don’t?

That large national study concluded that having advance directives had little effect on whether people were hospitalized and how often, or whether they died in hospitals. Yet where someone dies has considerable impact on how.

“I don’t want to demean advance directives,” said Susan Tolle, director of the Center for Ethics in Health Care at Oregon Health and Science University. “They lift an emotional burden from family members. Appointing a person to make decisions on your behalf really does have benefits.”

But when it comes to end-of-life wishes, Dr. Tolle said, citing Dartmouth Atlas findings, patient preferences aren’t the dominant factor. “It’s your ZIP code” — how aggressively most local physicians practice.

Though many of us have heard stories about a surgeon who insists a patient won’t die on his watch, “I don’t think that’s the norm,” said Charlie Sabatino, chairman of the American Bar Association’s Commission on Law and Aging.

What happens more often is that “these things aren’t getting where they’re supposed to be,” Mr. Sabatino said. Directives remain in safes or drawers and don’t reach hospitals; the family doctor has a copy, but a new specialist doesn’t. “People with complex medical conditions have a physician for every organ,” Mr. Sabatino pointed out.

Sometimes, he added, people fill out forms but never have thoughtful conversations with family members or doctors. They choose ineffective proxies. Family members disagree about treatments. Administrative errors — the wrong box checked on a form — interfere.

Researchers at the University of California, Los Angeles, surveyed more than 800 seniors and found that advance directives were not available in the medical records of more than half of the subjects who said they had completed and given them to a health care provider.

Calling 911 can cause particular problems, Dr. Tolle said. “Emergency medical personnel want orders,” she said. “Nothing else has been shown to be effective.”

That means a Polst program, the exception to this discouraging picture. Oregon, which pioneered Polst (Physician Orders for Life Sustaining Treatment) 20 years ago, in 2009 created an electronic Polst registry giving emergency workers and hospitals round-the-clock online access.

In a new study of 58,000 Oregonians who died in 2010 and 2011, the registry’s first two years, Dr. Tolle and her colleagues found that with Polst “you are extremely likely to have your wishes respected,” she said. Of those who specified “comfort measures only,” only 6.4 percent died in a hospital. Because some terminally ill patients can’t be kept comfortable at home, that percentage will never be zero. Those who specified “full treatment” to prolong life, on the other hand, were more apt to die in the hospital (44.2 percent) than those with no Polst in the registry.

Alas, most of us don’t live in Oregon (or West Virginia, which also has a state registry). Although more than 40 other states have some sort of Polst program, they’re not nearly as advanced and may not extend to every region or hospital or ambulance squad.

What do we do, then, to maximize the chances that our elders’ wishes — or our own — prevail? Eventually, electronic health records may help. Meanwhile, the best advice I can find is that everyone should have an advance directive, discuss his or her decisions with doctors and family members, and appoint a strong health care proxy. Then that person has to physically shove the appropriate forms under people’s noses at every hospitalization, every visit to a new physician, every transfer to a nursing facility.  (This smartphone app from the A.B.A. can help.)

“It requires you to be persistent and give them the advance directive again, every single time,” Mr. Sabatino said.

Even then, things can go wrong. In upstate New York, a nurse was arrested this month for failing to administer CPR to an 80-year-old nursing home resident who did not have a D.N.R. order in effect. Florida just fined a nursing home whose staff resuscitated a 75-year-old who did have a D.N.R. directive.

So when people get more sick and frail, and have likely reached their final year or two, they need Polst forms. And we all have to push our states to put in place full coverage and establish state registries. At the end, Polst seems our best bet.