Routine EEG monitoring in all comatose patients is not yet generally accepted practice. However, three significant findings have emerged from EEG studies of coma: firstly, the presence of cyclical changes in EEG amplitude and frequency, reminiscent of sleep-wake cycles, and the better prognosis in patients showing these EEG features compared with an unvarying pattern; secondly, the prevalence of unsuspected epileptiform activity, particularly in post-traumatic and post-hypoxic coma; thirdly, the association between inadequate sedation and poor control of intracranial pressure, with EEG arousal often coinciding with lability of blood pressure and intracranial pressure. In addition, patients with brainstem lesions show abnormal EEG slow-wave responses to stimulation, while extensive brainstem hemorrhage or infarction is associated with an unresponsive pattern of widespread rhythmical activity at a frequency of about 8 Hz ('alpha coma').
In addition to the general applications of EEG monitoring mentioned above, specific indications include the detection of cerebral hypoxia and ischemia in the first few days after head injury, particularly in patients with multiple injuries, and monitoring progress in patients remaining in coma after resuscitation ( Fig.2) and during recovery from sedative drug overdose or profound hypothermia. In these situations EEG activity often reappears before there is any clinical evidence of improvement.
Fig. 2 Cerebral function monitor (CFM) trace in a patient who had been resuscitated after cardiac arrest and was having repeated fits, shown by the marked increase in amplitude. During a period of hypotension, beginning at about 23.00, the EEG amplitude fell below 5 pV, indicating severe cortical ischemia, which persisted until a further cardiac arrest 12 h later. Calibration for cerebral function monitor: time, vertical lines every 10 min. Amplitude as marked (linear from 0 to 10 pV, logarithmic from 10 to 100 pV). BP, blood pressure.
EEG and evoked potential monitoring can affect clinical decision-making in a high proportion of comatose patients ( Jordan 1993), but whether routine neurophysiological monitoring in coma (regardless of the cause) improves the outcome has yet to be determined.
Was this article helpful?