If you land in an intensive care unit sick enough for doctors to consider withdrawing life support, be warned. Whether and when to pull the plug may depend in large part on the practices and culture of the ICU itself -- perhaps more than your needs or wishes, a new study finds.
That may be especially true if you’re so ill or incapacitated that you can’t make decisions about your own care, according to research being presented Tuesday at the American Thoracic Society International Conference.
After accounting for patient factors such as age, function, gender and race, the probability of having life support withdrawn ranged from 3.5 percent in some ICUs across the nation to 20.6 percent in others -- a six-fold variance.
“The really important message for patients and their families is, before you end up in an ICU, talk to your loved ones about what you would want,” said Dr. Caroline M. Quill, lead author of the study by researchers at the Perelman School of Medicine at the University of Pennsylvania.
Quill and her team analyzed records of more than 269,000 patients treated in 153 ICUs in the United States between 2001 and 2009. Overall, nearly 12 percent of patients had a decision made to go from a “full code” -- an all-out effort to save lives -- to some kind of limit on care.
That could have included: a DNR or do-not-resuscitate order; an order to withhold CPR or cardio-pulmonary resuscitation plus removing mechanical ventilation; dialysis or other life-saving treatments; or simply an order to provide only comfort measures or hospice care. About 59 percent of the patients died in the ICU and another 41 percent survived to discharge, the study found.
Particular patient characteristics accounted for most of the variability in decisions to withdraw life support, Quill acknowledged. But even after age, illness, functional status and other factors were analyzed, the variation among ICUs to authorize a DFLST -- decision to forgo life-sustaining therapy -- was striking.
“The finding of a six-fold variability among ICUs strongly suggests that the ICU to which a given patient is admitted strongly influences his or her odds of having a DFLST, regardless of personal or clinical characteristics,” the authors write.
The study didn't find particular variance by geography and the decision to withdraw care wasn't related to regional differences in cutting costs, Quill said.
End-of-life experts have known for years that individual doctor and ICU practices can affect the decision to withhold care, said Dr. J. Randall Curtis, director of the Palliative Care Center of Excellence at the University of Washington in Seattle.
In some ICUs, there’s a kind of heroic standard, an atmosphere in which doctors don’t talk about dying and every effort is made to sustain life, he said. In others, there’s an early effort to acknowledge the likelihood of death and to talk about the risks and benefits of care and how it fits into what a patient would want.
“I think there are still parts of our medical culture that say it’s our job to keep people alive no matter what,” he said. “As opposed to talking with patients and families about their values.
In the absence of instructions from the patient or their family members, the ICU’s culture is more likely to sway the decision about whether to prolong care -- or not.
Quill emphasizes that the study didn’t draw any conclusions about the rates of ICU decisions to withdraw support. “We don’t necessarily think that the 3.5 percent rate is too low or the 20.6 percent rate is too high,” she said.
Surprisingly, perhaps, there actually are no hard-and-fast guidelines in the critical care field about when to withdraw life support. That’s mostly because it would be too hard to envision every potential scenario, said Dr. Douglas White, an associate professor of critical care medicine and director of the program on ethics and decision making in critical illness at the University of Pittsburgh.
“These decisions about whether and when to withdraw life support are not scientific decisions,” he said.
In an ideal scenario, a patient would have made his or her wishes about end-of-life care known well in advance. In practice, however, although the number of people who spell out their wishes is growing, only about 10 percent of the general public has completed an advance directive or obtained a durable power of attorney for health care, Curtis noted.
That means that the final decision often is left to family members or other “surrogate” decision makers, and to doctors and clinicians.
“If the patient hasn’t articulated a clear preference, it’s very, very hard for doctors and families to work together to make the decision, “ White said. “(They’re) left to do the best that they can that generally reflects who the patient is as a person.”
It would be far better, experts agree, if more people anticipated the possibility of end-of-life decisions and left clear instructions. Ten years ago, that might have been preferences for -- or rejection of -- specific treatments such as CPR or mechanical ventilation, experts say.
Today, the discussion centers much more on the overall goals of care and the values of the patient. Take the end-of-life instructions drafted by Dr. John Luce, an emeritus professor of anesthesiology and medicine at the University of California, San Francisco.
Luce, who has researched and written about end-of-life care, said he values a “sentient existence” in which he retains the ability to “think, read, speak, write and communicate with people.”
“Anything that could not restore me to this, I would not want it,” he said.
So that’s how Luce’s advance directive reads, if, heaven forbid, he should need it. The new research that underscores the variability of ICU practices should also emphasize the need for ordinary people to make their wishes known, Luce said.
“The really important thing is to get people talking with those who could become their surrogates,” he said. “The written document is less important than the conversation that leads up to the directive.”
Related: