[Early prognostic assessment using evoked potentials in severe craniocerebral trauma].

103 patients in acute posttraumatic coma were assessed during 72 hours following severe head injury by clinical examinations (documented with a modified Glasgow-Coma-Score and a brain-stem-score and brainstem auditory evoked potentials (BAEP) as well as short latency somatosensory evoked potentials (SEP). Patient outcomes were classified at 6 months or more, according to the following categories: good recovery, severely disabled or vegetative, and brain dead. Patients who had died by systemic complications were excluded from the study. The Glasgow-Coma-Score was reliable in forecasting a favorable outcome, but tended to produce false-pessimistic predictions. The brainstemscore and the BAEPs were reliable predictors of an unfavorable but not a favorable outcome. SEP data however, performed well as a prognostic indicant in predicting an unfavorable as well as a favorable outcome. Using the BAEP, the amplitude ratio Wave V/Wave I is more sensitive to detect a lesion than the interpeak-latency Wave I-Wave V, and--using the SEP--the amplitude ratio N20/N13b (over C2) is more sensitive than the central conduction time (N13a (over C7)--N20).