Prognostic Value of EEG Monitoring After Status Epilepticus
Prognostic Value of EEG Monitoring After Status Epilepticus
ABSTRACT & COMMENTARY
Source: Jaitly R, et al. Prognostic value of EEG monitoring after status epilepticus: A prospective adult study. J Clin Neurophys 1997;14:326-334.
Modern digital technology has made continuous EEG monitoring in the neurological intensive care unit easy and widely available (for review, see Jordan K. J Clin Neurophys 1993;10:445-475). However, the significance of certain EEG patterns remains uncertain, particularly in status epilepticus (SE) patients (Pohlmann-Eden B, et al. J Clin Neurophys 1996; 13:519-530; Treiman D, et al. Epilepsy Res 1990;5:49-60). Jaitly and colleagues now present evidence that three EEG monitoring patterns (burst suppression, subclinical electrographic seizures, and periodic lateralized epileptiform discharges [PLEDs]) are associated with a poor prognosis when present after SE.
The authors performed continuous EEG monitoring for at least 24 hours on 180 adult patients admitted to the Medical College of Virginia Hospital in SE between 1989 and 1992. In 112 patients, monitoring was begun during SE; in the other 68, it was started within 30 minutes of cessation of clinically overt seizure activity. Outcome at 30 days after cessation of SE was classified according to whether the patient was deceased, was care-dependent, or was independent. "Deceased" and "care-dependent" were considered "poor outcome." The overall mortality rate was 28%, and the poor outcome rate was 38%. The authors correlated these outcome measures with the EEG patterns they observed.
The occurrence during the monitoring of the following EEG patterns was significantly associated with mortality and poor outcome: burst-suppression, electrographic ictal activity without clinical accompaniment, and PLEDs. The highest mortality rate (68%) occurred among the 27 patients exhibiting EEG burst-suppression. The following EEG patterns were not associated with mortality or poor outcome: focal slowing, generalized slowing, interictal epileptiform discharges, and attenuation.
In only 17 patients did a normal EEG appear during the initial 24 hours after SE ended. However, there was no mortality or residual neurological disability in this group. Thus, if the EEG normalizes rapidly after SE, the prognosis is good.
The authors went on to carry out a multivariate logistic regression analysis with SE etiologies and EEG monitoring patterns as possible predictors of prognosis. Their intent was to determine if EEG patterns conferred additional prognostic value, beyond the prognostic value of the underlying causative processes. Their multivariate analysis did indicate that burst-suppression, electrographic ictal activity without clinical accompaniment, and PLEDs continued to be associated with poor outcomes, even when underlying etiologies were controlled for.
COMMENTARY
All neurologists would agree with Jaitly et al that continuous EEG monitoring is critical for assessing the efficacy of treatment of severe SE, particularly when pentobarbital coma is employed. However, the authors have informed us that EEG patterns can also be used for prognostication purposes, in addition to being used as guides to therapy.
Several interesting questions were not addressed by the research design employed by Jaitly et al. They presented no data concerning treatment of SE and how treatment might affect both EEG patterns and outcome. Moreover, they used a multivariate logistic regression analysis to study prognostic value of the EEG patterns independent of underlying causative conditions. However, they did not inform us about which EEG patterns correlated with which underlying disease processes.
One wonders by what mechanism the EEG patterns of burst-suppression, electrographic ictal activity without clinical seizures, and PLEDs confer worse prognoses than is conferred just by the underlying diseases that caused the SE. In this series, strokes were the most common cause of SE. Strokes followed by PLEDs had worse outcome than strokes without PLEDs. The more severe strokes may put the patient at greater risk for secondary fatal complications such as aspiration pneumonia and pulmonary embolus. Unfortunately, Jaitly et al reported causes of SE but did not report causes of death.
Among their 180 SE patients, Jaitly et al found 96 with electrographic ictal discharges continuing after clinical SE had ended. Mortality in this group was 41%. Similar comatose patients in SE were described by Lowenstein and Aminoff (Neurology 1992;42:100-104). In their series, 38 patients were comatose but had electrographic-only SE or had electrographic SE with only subtle clonic face or upper extremity movements. "Most" of Lowenstein and Aminoff’s patients had previously been in clinical convulsive SE. The mortality rate in their group was 61%. Taken together, these studies suggest that the prognosis in such electrographic SE patients is particularly grim. However, we described one similar patient who survived (with incomplete neurological recovery) after going in and out of clinical and electrographic SE (and iatrogenic coma) for nearly two months (Labar D, et al. J Epilepsy 1993;6:170-173). The extent to which electrographic SE after convulsive SE should be treated remains a controversial issue among epileptologists and intensive care neurologists. drl
EEG monitoring of patients in status epilepticus:
a. indicates a poor prognosis when slowing and attenuation is present.
b. is useful for titration of antiepileptic medication.
c. is useful only in children.
d. indicates more than 80% mortality when PLEDs are seen.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.