Subscribe to eTOC

EEG Can Help Predict Neurologic Outcomes After Cardiac Arrest

Article In Brief

Researchers found that EEG initiated rapidly following cardiac arrest and continued for up to five days can reliably predict the neurological outcome of a coma. The results can help with making decisions about withdrawal of care, but independent experts said it shouldn't be the only tool used in the decision-making process.

Electroencephalography (EEG) initiated rapidly following cardiac arrest and continued for up to five days can reliably predict neurologic outcome of coma, according to a study that was published in the August issue of Annals of Neurology.

Continuous EEG patterns at 12 hours predicted a good outcome while generalized suppression or synchronous patterns with at least 50 percent suppression were invariably associated with a poor outcome between six hours and five days after cardiac arrest.

The results can aid in decisions about withdrawal of care, the study authors said. However, experts who reviewed the paper agreed the test should never be used in isolation but rather as one tool in a complex decision-making process.

The authors of the study also underscored the threat of the “self-fulfilling prophesy,” in which test results that portend a poor outcome may influence treatment decisions in a way that leads to withdrawal of care.

In the paper, lead author Barry J. Ruijter, MD, PhD, of the department of clinical neurophysiology at the University of Twente, Enschede, in the Netherlands, and colleagues, wrote that to minimize the risk, decisions on treatment withdrawal were based on international guidelines including bilaterally absent somatosensory evoked potential (SSEP), absent or extensor motor responses, and absent brainstem reflexes.

“EEG recordings were intended for the detection and treatment of electrographic seizures, and none of the participating centers used recommendations to withdraw care based on early EEG findings,” they wrote.

Study Design, Findings

In the study, 887 patients at five teaching hospitals in the Netherlands who were comatose following cardiac arrest were started on continuous EEG as soon as possible after admission, preferably within 12 hours after arrest. Patients were treated according to standard protocols for comatose patients after cardiac arrest. Patients received propofol, midazolam, or both for sedation, and morphine, fentanyl, or remifentanil for analgesia.

The primary outcome measure was neurologic functional recovery at six months, expressed as the score on the five-point Glasgow-Pittsburgh Cerebral Performance Category, with “good” being defined as a score of 1 or 2, and “poor” defined as a score of 3, 4, or 5.

Results showed that continuous EEG patterns were associated with a “good” outcome, if present within 12 hours after cardiac arrest. Generalized EEG suppression and synchronous patterns with 50 percent or greater suppression were invariably associated with a poor outcome.

“With this prospective cohort study...we confirm that early EEG allows for reliable prediction of outcome of comatose patients after cardiac arrest,” Dr. Ruijter and colleagues wrote. “...We confirm that unfavorable EEG patterns and absent SSEP responses have complementary value for the prediction of poor outcome.”

Expert Commentary

Experts who reviewed the paper for Neurology Today agreed the study is a methodologically strong one that addresses a critical, if not, difficult subject. “No one wants to prolong life if there is no chance of meaningful recovery,” said Lawrence J. Hirsch, MD, FAAN, professor of neurology and chief of the division of epilepsy and EEG at Yale University. “For that reason, early predictors of long term outcome would be of great utility.”

Yet EEG alone as a predictor for long-term outcome is problematic, and at least one reviewer said the study is unlikely to change the standard of care in US centers.

“The major confounder here is a multitude of CNS depressing drugs in substantial doses and very little information about the patients' neurologic findings,” said Eelco F. Wijdiks, MD, FAAN, chief of the division of critical care neurology at Mayo Clinic in Rochester, MN. “This hampers the interpretation of the studied population. Prolonged EEG monitoring cannot and should not trump a serial neurologic examination, particularly in heavily sedated patients.”

“Moreover, the midazolam dose is at a ‘status epilepticus’ level rather than at a ‘sedation’ level,” he said.

Other experts said the results of the study merit replication in cohorts receiving aggressive treatment and therapeutic temperature management. “The findings convincingly demonstrate the utility of early EEG findings in helping to prognosticate in patients with post-anoxic coma, with maximal utility at 12-24 hours after arrest,” Dr. Hirsch said. “The authors provide specific EEG patterns that are easily recognized and have high predictive power at their centers, both for poor outcome, which included severe disability, and good outcome.

“One can convincingly state that early EEG findings can predict outcome, including predicting who will have care withdrawn, but whether some patients who had withdrawn life-sustaining treatment would have woken up in a delayed fashion cannot be answered from this type of study.”

Dr. Hirsch added that the results may not necessarily apply to US centers where patients are likely to receive a different standard of care. He noted in particular that at the five centers participating in the study, the standard was to withhold treatment of electrographic status epilepticus for patients not currently enrolled in a clinical trial. “This is not how we would treat patients here in the US in most centers, especially in light of recent publications showing good outcomes with aggressive treatment in many of those patients,” he said. “In addition, only one of the five centers used therapeutic hypothermia on a regular basis.”

Jennifer L. Hopp, MD, FAAN, director of the University of Maryland Epilepsy Center, agreed about the strength of the study and about the limitations regarding use of EEG for prognostication. “The study is important and was well designed with regard to consistency of clinical care, procedures, and EEG interpretation in a multicenter, prospective cohort study,” Dr. Hopp told Neurology Today. “The findings of continuous EEG patterns as a predictive finding of good outcome is a validation of prior work but is strengthened by the methodology and large number of patients in this particular study. Another strength of the study is the assessment of EEG at multiple points in the clinical course of patients following arrest. The inclusion of EEG interpretation at 36 hours and at five days post-arrest improved sensitivity of the findings.”

But Dr. Hopp said a host of factors can affect the EEG in patients who have suffered cardiac arrest, including hypothermia and medications which may affect the record. “The interpreter of the EEG needs to be acutely aware of these factors and take them into consideration when interpreting the record,” she said.

The take-away message is that early EEG can be one useful tool among others, depending on a host of other clinical factors, in making critical decisions about withdrawal of care, the experts agreed.

“It does appear that early use of EEG may yield findings that may be helpful in clinical decision-making, as long as factors such as hypothermia and sedation that may affect the EEG are considered,” Dr. Hopp told Neurology Today. “Clinicians should be aware of the concept and concern about ‘self-fulfilling prophecy’ when using EEG tools and continue to include other data, including clinical examination in the decision-making process.”

Dr. Hirsch added: “Not only can the EEG provide very useful prognostic information, as proven in this study, but it can also rule out seizures that require treatment.... Clinicians should include the EEG findings in their decision-making but should not use EEG by itself when deciding whether or not to withdraw life support.”


Dr. Ruijter had no disclosures. Dr. Hirsch received research support to Yale University for investigator-initiated studies from Monteris and Upsher-Smith; consultation fees for advising from Adamas, Aquestive, Ceribell, Eisai, Monteris, Marinus, Medtronic, Neuropace, and UCB; royalties for authoring chapters for UpToDate-Neurology and from Wiley for co-authoring the book, Atlas of EEG in Critical Care; and honoraria for speaking from Neuropace. Dr. Hopp has received a speaker honorarium and travel expenses for a course from the AAN, an honorarium from Continuum and UpToDate, consulting fees from Taconic Research, LLC, and salary support from NIH grants.

Link Up for More Information

• Ruijter BJ, Tjepkema-Cloostermans MC, Tromp SC, et al. Early electroencephalography for outcome prediction of postanoxic coma: A prospective cohort study Ann Neurol 2019;86(2):203–214.