The prognostic value of the EEG in postanoxic coma

Abstract

Introduction: Several studies have been performed to assess the prognostic value of early neurological and neurophysiological findings in patients with postanoxic coma, but they have not led to precise, generally accepted, prognostic rules. This study was performed to assess whether it is possible to create a prognostic outcome table, using a combination of clinical variables and the electroencephalogram (EEG).

Methods: Clinical variables and EEG were registered on admission, at day 1-5, day 6-14, and day 15-42, and were related to the Glasgow outcome scale (GOS) at 30 and 180 days. The EEG abnormalities were categorized using the EEG classification system proposed by Synek (J Clin Neurophysiol 5:161-174, 1988) and Young et al. (Can J Neurol Sci 24:320-325, 1997). These EEG classifications were then further divided into prognostic categories.

Results: Age was a significant predictor of outcome. The early recorded clinical variables were the most predictive and the GCS showed a limited prognostic value. The first EEG registration proved to be the most predictive. The Synek-classification was divided into three prognostic categories: "benign," "malignant," and "fatal" and the Young-classification into four: "benign," "intermediate," "malignant," and "fatal." An outcome prediction table is proposed using the Young-classification stratified for age.

Conclusion: Age is an important variable determining the prognostic value of the EEG and should always be taken into consideration. The prognostic categories, especially when derived from the Young-classification, showed a good prognostic value. Although this is a pilot study, we believe that the revised prognostic categories have a good prognostic value in predicting outcome and are worth further investigation and validation.