End-of-Life ICH Treatment Highly Variable Among Physicians

Caroline Helwick

October 17, 2013

NEW ORLEANS — End-of-life treatment for patients with intracerebral hemorrhage (ICH) is highly variable, and much of this variability can probably be attributed to heterogeneity among physicians in their treatment recommendations, a survey from the University of Michigan shows.

The survey asked physicians how they would treat 3 different theoretical patients whose severity was standardized. The goal was to determine what role — if any — physicians play in the variability in end-of-life decisions after ICH.

"We found tremendous variability in physicians' initial treatment recommendations, despite standardized severity and patient wishes," said Darin B. Zahuranec, MD, from the Stroke Program and the Center for Bioethics and Social Sciences in Medicine at the University of Michigan, Ann Arbor.

Thirty-day mortality for ICH is around 30%, and most early deaths occur after a decision to withdraw or withhold life-sustaining treatment. Do-not-resuscitate (DNR) orders are strongly associated with mortality after ICH, and there is substantial variability across physicians and institutions as to these practices, he said.

The findings of the survey were presented at the American Neurological Association (ANA) 2013 Annual Meeting.

Theoretical Cases

The researchers surveyed neurologists and neurosurgeons from a single academic institution. Of 115 who received the surveys, 58 responded (50%) and formed the basis of this pilot survey.

The respondents were presented with 3 cases of adults with spontaneous ICH, aged 61 to 83 years, whose Glasgow Coma Scale (GCS) scores ranged from 4 to 11. Each case included examination findings, computed tomography (CT) images, and standardized background information that indicated the patient was independent at baseline, had no advance directive, but had stated that he or she "did not want to be a vegetable."

After each vignette, the respondents were asked a series of questions. They had to estimate 30-day mortality and 90-day functional independence, and they had to give a recommendation to the family for initial treatment intensity on a scale of 1 (comfort care only [ie, DNR order], with withdrawal of any existing intensive life-sustaining measures) to 6 (full intensive care [ie, all resuscitative measures, any indicated invasive diagnostic or therapeutic procedure]). Collapsing the scale, the researchers constructed 3 categories: comfort care (1 to 2), moderate care (3 to 4), and full care (5 to 6).

One case involved an 83-year-old receiving warfarin who had right hemiparesis and aphasia, with a GCS score of 11. The median probability of 30-day mortality, according to the participants, was 30% (interquartile range [IQR], 20%, 50%), and median for 90-day functional independence was 10.5% (IQR, 10%, 20%).

Their recommended treatment intensity ranged widely; 14% recommended comfort care, 36% recommended moderate care, and 50% advised full intensive care, Dr. Zahuranec reported.

Another case involved a 76-year-old with decreased responsiveness and a GCS of 7T. He had extensor posturing of the left arm and flexor posturing of the right arm, and he withdrew both legs to pain. For this case, median probability of 30-day mortality was 55% (IQR, 34%, 75%) and median estimated 90-day functional independence was 10% (IQR, 5%, 25%).

The recommended treatment intensity was almost evenly split across the care levels, with 43% advising comfort care for this patient, 28% advising moderate care, and 29% recommending full intensive care.

"There was wide variability in 30-day mortality estimates, possibly due to uncertainty in the ICU prognosis and different assumptions about the course of treatment," he said. "Estimates in 90-day functional outcome were less variable, which likely reflects the dichotomous outcome of return to independence in patients with moderate-to-severe haemorrhage."

"We think there may be intrinsic physician biases on quality of life in severe stroke, and it's possible this is responsible for some of the observed variability in ICH treatment at the end of life," he concluded, adding that these decisions clearly affect disease course. "These decisions are made early. One third of DNR orders are given on the first day."

The researchers are now conducting a national survey to better understand the factors underlying this variability. In this larger survey, they will examine physician factors (training, experience), patient factors, and the impact of formal prognostic models.

"We know that more work is needed to understand the impact of this [variability] on patients and families," he said.

Brisk Discussion

After the presentation, discussion of the topic was brisk among the attendees at the neurocritical care session.

James Meschia, MD, professor of neurology at the Mayo Clinic in Jacksonville, Florida, said he would like the survey to include postoperative vignettes. "I can almost guarantee that for the same predictive mortality, once the patient is operated on the decision to be aggressive changes."

"There's a tendency to hold on. You might have been reluctant to take the patient to surgery, for example, but once that decision is made there's a tendency to latch on. You can go from a middle range of 'enthusiasm' to suddenly becoming extremely aggressive in your care. Aggressive therapy tends to snowball," he followed up in an interview with Medscape Medical News.

He further maintained that physicians do bring their own ideologies and philosophies to the bedside of the patient with ICH, as do families and also institutions. "These things will be hard to sift out," he said.

Alesandro Rabinstein, MD, professor of neurology at the Mayo Clinic in Rochester, Minnesota, said he has led discussions about end-of-life treatment of ICH and has observed little agreement among neurologists.

"I came up with real-life extreme cases where we had all the information on the patient, including the patient's wishes. There was never a situation where we all agreed that withdrawal of life-sustaining treatment was the right thing to do. We never had uniformity in the room, even with just a few dozen physicians in the discussion."

Dr. Zahuranec, Dr. Meschia, and Dr. Rabinstein have disclosed no relevant financial relationships.

American Neurological Association (ANA) 2013 Annual Meeting. Abstract #T1803. Presented October 15, 2013.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.