Related Experiment Video
Updated: Jun 22, 2026
A Bedside, Single Burr Hole Approach to Multimodality Monitoring in Severe Brain Injury
Published on: March 26, 2019
Continuous EEG-SEP monitoring in severe brain injury
A Amantini1, S Fossi, A Grippo
1Unit of Clinical Neurophysiology, DAI Neurological Sciences, University of Florence - Az., Ospedaliero Universitaria Careggi, V. le Morgagni 85, 50134 Florence, Italy. amantini@unifi.it
Aims:
To monitor acute brain injury in the neurological intensive care unit (NICU), we used EEG and somatosensory evoked potentials (SEP) in combination to achieve more accuracy in detecting brain function deterioration.
Methods:
Sixty-eight patients (head trauma and intracranial hemorrhage; GCS<9) were monitored with continuous EEG-SEP and intracranial pressure monitoring (ICP).
Results:
Fifty-five patients were considered "stable" or improving, considering the GCS and CT scan: in this group, SEP didn't show significant changes. Thirteen patients showed neurological deteriorations and, in all patients, cortical SEP showed significant alterations (amplitude decrease>50% often till complete disappearance). SEP deterioration anticipated ICP increase in 30%, was contemporary in 38%, and followed ICP increase in 23%. Considering SEP and ICP in relation to clinical course, all patients but one with ICP less than 20 mmHg were stable, while the three patients with ICP greater than 40 mmHg all died. Among the 26 patients with ICP of 20-40 mmHg, 17 were stable, while nine showed clinical and neurophysiological deterioration. Thus, there is a range of ICP values (20-40 mmHg) were ICP is scarcely indicative of clinical deterioration, rather it is the SEP changes that identify brain function deterioration. Therefore, SEP have a twofold interest with respect to ICP: their changes can precede an ICP increase and they can constitute a complementary tool to interpret ICP trends. It has been very important to associate SEP and EEG: about 60% of our patients were deeply sedated and, because of their relative insensitivity to anesthetics, only SEP allowed us to monitor brain damage evolution when EEG was scarcely valuable.
Conclusions:
We observed 3% of nonconvulsive status epilepticus compared to 18% of neurological deterioration. If the aim of neurophysiological monitoring is to "detect and protect", it may not be limited to detecting seizures, rather it should be able to identify brain deterioration, so we propose the combined monitoring of EEG with SEP.
