Clinical relevance of long-latency SEPs and VEPs during coma and emergence from coma

Follow-up measurements of long-latency visual (VEP) and somatosensory (SEP) evoked potentials were performed on 30 comatose patients. Twenty-seven of the patients had severe head injury, 2 had encephalitis and 1 was in a posthypoxic state. For the SEP study a mechanical vibration stimulus was used, applied 60 times at intervals of 10 sec. The same rate was used for visual stimulation. The late EP components were classified by a signal-to-noise ratio (SNR), whereby an SNR of less than 2.6 is characteristic of a questionable or unmeasurable EP and an SNR greater than 2.6 is evidence of a clearly existing EP; the clinical state was scored using the Glasgow Coma Scale (GCS); the patient outcome was assessed 6 months after the injury using the Glasgow Outcome Scale. The highest correlation (r = 0.72) was found between the clinical state, represented by the GCS, and the SEP. A similar correlation (r = 0.66) was obtained between the GCS and the vertex VEP. The occipital VEP showed no correlation. Emergence from coma and recovery was accompanied by an increase of the SEP and an increasing spread of the VEP over the whole scalp. Most patients with a clear long-latency SEP in coma had a favourable outcome; a missing or a questionable SEP indicated a poor outcome.