ARTICLE IN BRIEF
✓In new AAN practice parameters, the Quality Standards Subcommittee reviewed and assessed the diagnostic value of the clinical examination and of ancillary investigations for poor outcome in comatose survivors after CPR, including these variables: circumstances surrounding CPR, elevated body temperature, neurologic examination, electrophysiologic studies, biochemical markers, monitoring of brain function, and neuroimaging.
Predicting patient recovery from coma, particularly after resuscitation from cardiac arrest, is one of the more vexing puzzles that neurologists face, and one of the most common reasons for consult requests from non-neurologists. In response, the AAN has released new guidelines to help guide decisions about care (Neurology 2006;67:203–210).
Eelco F. Wijdicks, MD, lead author of the guidelines, described them as a high priority for the AAN “simply because we really didn't know for certain how neurologists in the trenches prognosticate” on coma recovery following cardiopulmonary resuscitation (CPR).
“The assessment of neurologic prognosis has long been based mainly on the algorithms…which derive from a single cohort study and brings about substantial statistical uncertainty,” the authors wrote.
To improve this approach, an international team worked for two years to identify assessments that, according to the published literature, had the highest degree of accuracy in predicting a poor outcome; more than 400 studies published from 1996 to 2006 were reviewed.
POOR OUTCOME DEFINITIONS
“There were quite a few papers that we had to trash or not use because of poor outcome definition. There was no definition at all in some of the papers,” said Dr. Wijdicks, Chair of the Division of Critical Care Neurology at Mayo Clinic in Rochester, MN. He added, “there are few data on prognostication of good outcomes.”
“I think some of those studies were done with no neurologists on board. Even among more current research, there are few studies with sufficient neurological details,” Dr. Wijdicks said.
The guidelines team defined a poor outcome as death, persisting unconsciousness after one month, or severe disability requiring full nursing care after six months.
VARIABLES – PREDICTIVE AND NOT
The authors reviewed the diagnostic value of the clinical examination and of ancillary investigations for poor outcome in comatose survivors after CPR. They assessed these variables: circumstances surrounding CPR, elevated body temperature, neurologic examination, electrophysiologic studies, biochemical markers, monitoring of brain function, and neuroimaging.
Prolonged anoxia, duration of CPR, cause of the cardiac arrest, elevated body temperature, and presence of brain swelling were least accurate in predicting a poor outcome, the neurologists concluded.
Clinical findings that accurately predicted poor outcomes included myoclonus status epilepticus within the first 24 hours in patients with primary circulatory arrest; absence of pupillary responses within days one to three after CPR; absent corneal reflexes within days one to three after CPR; and absent or extensor motor responses after three days, the authors wrote.
More – and costly – technology isn't necessarily better at predicting poor outcomes. “Serial or continuous EEGs may appear more accurate and valid than single EEGs, but this remains to be adequately tested. The presence of EEG reactivity and variability has been suggested to favorably predict recovery of consciousness, but this has not independently demonstrated,” according to the guidelines.
The value of ongoing MRI and CT scans are also uncertain, the authors concluded, “because there is insufficient evidence to conclude they conclusively predict poor outcomes.”
EXPERTS' RESPONSE TO GUIDELINES
J. Claude Hemphill III, MD, Associate Professor of Clinical Neurology and Neurological Surgery at the University of California-San Francisco, and Director of the Neurocritical Care Program at San Francisco General Hospital, welcomed the guidelines.
Articles about coma recovery have typically focused on a single measure, such as the clinical examination or serum biomarkers. The strength of this paper is that it “really puts it all together in a coordinated approach,” Dr. Hemphill told Neurology Today.
“This is the first document that codifies this into a practice parameter,” he added. “That is extremely useful – and will also help us understand the shortcomings of our research.”
William M. Coplin, MD, Chief of Neurology and Medical Director, Neurotrauma and Critical Care at Detroit Receiving Hospital/Detroit Medical Center, agreed that the guidelines were needed.
These practice parameters “explore an area where people are often guided more by religion than by science,” said Dr. Coplin, who is also the Chair-elect of the AAN Critical Care and Emergency Neurology Section. “It is important to try and debunk the self-fulfilling prophecy…the patient looks terrible, the patient will do terribly. It is a systematic review. It helps us in terms of speaking to the family about resource utilization and level of care for patients who are not going to wake up.”
There are some shortcomings to the guidelines, Dr. Hemphill said. “The prognostic criteria provide good information for patients with essentially 100 percent likelihood of a poor outcome. What is not addressed are criteria for assessing patients who are facing a less determined outcome,” he said.
On a different measure, Dr. Coplin said he believed evidence for the usefulness of testing for creatine kinase brain isoenzyme (CKBB) is “a little stronger” than it appears in this paper. Echoing Dr. Hemphill's comment, Dr. Coplin said the guidelines “still leave us a little confused as to who is going to do well.”
He agreed that more research is needed on the long-term outcome of coma patients, and acknowledged the difficulty this poses. “It is hard to collect data,” Dr. Coplin said. “This is not exactly a disease state [for which] there are referral centers, where people can go to an academic medical center and be put in a registry.”
The authors themselves noted that more research on the “feasibility, costs, and predictive value of MRI,” as well as on outcomes, is needed.
Dr. Wijdicks said he does not think any of the conclusions will be viewed as controversial, and are not meant to substitute for individual clinical judgment and experience. But, he adds, “It is important to have an evidence-based review.”
Another stated purpose of the guidelines is to assist neurologists in discussing with family members the “need for life supportive care” when a poor outcome is anticipated.
The guidelines may prove helpful to physicians who are “dealing with walking, talking patients. This is something they see, but they are not that focused on it,” Dr. Coplin said. “Some physicians are afraid to use the D word…to say that the patient is going to die. That uncertainty is always there in making final decisions.”
A RESOURCE FOR INTENSIVISTS
Intensivists might find the guidelines informative as well, as they lay out the rationale for certain judgments, Dr. Wijdicks said. “The guidelines are clear,” he said. “They do not say, ‘Here are the guidelines, don't call a neurologist.’ They say neurologists should be involved in conducting the neurological assessment and in developing a prognosis. It is best done by a neurologist, and they should discuss that [finding] with the family.”
Dr. Hemphill agreed. “I think it will encourage non-neurologists to seek neurology consultation, rather than doing it on their own using a table from a journal article as a cookbook. It's not always that straightforward,” he said.
The guidelines should be incorporated in the neurology curriculum in medical schools and residencies, Dr. Hemphill said. He noted that this has already happened in a limited way at his institution. “I already saw several residents carrying around copies of it that they identified on their own,” he said.
While the guidelines will undoubtedly bring to mind the Terri Schiavo case, “the guidelines had none of this in mind,” and had been in process “long before” her case became a matter of public debate, said Dr. Wijdicks.
He emphasized that the purpose was to help “identify patients for whom there is futility of care, and to let certain decisions” flow from that. The team was careful not to go beyond the research, he noted. “We don't say what those decisions should be and there is a reason for that.”
Table. PREDICTORS FO...Image Tools