Do-not-resuscitate orders were less common at for-profit and academic hospitals than at rural facilities and also varied widely by county, even after accounting for patient characteristics, a new study found.
The differences appeared to reflect institutional culture, technological bent and physician practices, according to the study in the Archives of Internal Medicine. The solutions proposed by experts include guidelines on appropriate use of do-not-resuscitate orders and better training for residents and attending physicians.
In the study, researchers at the University of California at Los Angeles reviewed records for nearly 820,000 patients age 50 and older at 386 hospitals across California in 2000. They found a tenfold variation by county in the use of do-not-resuscitate orders within 24 hours of admission, said study co-author David Zingmond, an assistant professor of medicine.
The variations did not correspond well to county population, hospital bed availability or population density, Zingmond said.
"It appears that there is not a consensus on how to implement do-not-resuscitate orders across the state," he said. "We don't have people in one part of the state being 10 times sicker."
Financial incentives may be leading to more aggressive treatment, Zingmond said.
In addition, patients often check into academic medical centers specifically because of the high-tech, sometimes life-saving procedures available. "They didn't come with the purpose and expectation that we aren't going to do everything," Zingmond said. "They came here with the opposite expectation."
Jeffrey Kahn, director of the Center for Bioethics at the University of Minnesota at Minneapolis, said while bedside caregivers have little to gain financially from inappropriately aggressive care, they may respond to an institutional culture and incentives that promote more aggressive care. "But that's a harder case to make," Kahn said.
Although implementation of do-not-resuscitate orders ultimately is up to the patient, physicians generally initiate the conversation, Zingmond said. One reason for higher rates of do-not-resuscitate orders at small and rural hospitals could be that doctors at those facilities were more likely to have existing relationships with the patients, making discussion of difficult topics easier, he said.
The low rates at academic medical centers surprised Arthur Derse, a professor of bioethics and emergency medicine at the Medical College of Wisconsin in Milwaukee.
Derse agreed one explanation could be that patients often go to such facilities for aggressive treatment. The low rates also may reflect the fact that responsibility for initiating conversations about do-not-resuscitate orders often falls to residents -- and residents may feel uncomfortable discussing the topic, he said.
"We teach them how to do it, but it's a difficult art to learn: How to talk to patients about end-of-life care," Derse said. "I think this means we may have to redouble our efforts to teach residents and attending physicians how to have these discussions and how to do it well."
Zingmond said it would help if hospital and quality guidelines addressed how and when physicians should discuss do-not-resuscitate preferences with patients and families. What's more, if doctors knew their quality scores and evaluations included whether they recorded discussing end-of-life preferences with patients (and subsequently whether the preferences were followed), doctors would be more likely to put aside their discomfort and talk about the subject, he said.
Also, residents and medical students should accompany physician-mentors when they discuss options with patients and families, he said.
Derse opposed tracking only the rate of do-not-resuscitate orders. "I think the quality issues are, was it considered and was it appropriate," he said. "But just measuring whether a person has a (do-not-resuscitate) order in place might be too much."
Wesley Smith, a senior fellow at the conservative Discovery Institute in Seattle and author of Culture of Death: The Assault on Medical Ethics in America, said guidelines on when such orders are appropriate might be beneficial, but do-not-resuscitate "orders should not be the goal in and of themselves."
The issue might offend some patients and families, and bringing it up at the wrong time could create mistrust, said Smith, a volunteer lay chaplain at a hospital in the San Francisco Bay area where he lives.
"My experience is that when someone knows someone is dying, they are willing to have (do-not-resuscitate) orders, and when they're not dying, they're not," he said. "If Uncle Charlie has a heart attack, you're not thinking (do-not-resuscitate) because two days ago, you were having dinner with him."