Early electroencephalogram does not reliably differentiate outcomes in post-hypoxic myoclonus.

Abstract

Prognostication in patients with post-hypoxic brain injury remains difficult; yet, clinicians are commonly asked to guide decisions regarding withdrawal of life support. We aimed to assess whether electroencephalogram (EEG) is a useful tool in predicting neurological outcome in patients with post-hypoxic myoclonus (PHM). This study was conducted as part of an internal hospital audit assessing therapeutic hypothermia in patients with hypoxic cardiac arrest. We identified 20 patients with PHM and evaluated their initial routine EEG. Three blinded neurologists independently assessed EEGs and scored them using the standardised critical care EEG terminology from the American Clinical Neurophysiology Society (2012 version) and the EEG patterns identified by the Target Temperature Management (TTM) trial group. Glasgow Outcome Scale (GOS) scores were used to assess neurological outcome at 30 and 90 days. Mortality rates at these time points were also documented. We found that the majority of patients (18/20) with PHM had an initial EEG that was "highly malignant" or "malignant", but outcomes at 30 and 90 days were not universally fatal. Six patients were alive at 30 days, and five at 90 days. Of the latter, two patients had moderate disability (GOS score 4) and one improved from a GOS score of 3 to 5, with only low disability. Two patients with "benign" EEGs had unchanged GOS scores of 3 at 30 and 90 days, indicating severe disability. This study shows that PHM is associated with a poor but not universally fatal prognosis. Early EEG does not reliably distinguish between good and poor outcomes.