Prediction of 'awakening' and outcome in prolonged acute coma from severe traumatic brain injury: evidence for validity of short latency SEPs

OBJECTIVE: To evaluate the prognostic value of somatosensory evoked potentials (SEPs) in severe traumatic brain injury (TBI) considering both 'awakening' and disability. METHODS: SEPs were recorded in 60 severe TBI with duration of acute coma>7 days. N20-P25 amplitudes, their side-to-side asymmetry and CCT were measured. SEPs on each hemisphere were classified as normal (N), pathological (P) or absent (A). 'Awakening' and disability were assessed after at least 12 months using Glasgow Outcome Scale (GOS). SEP predictive value was compared with GCS and EEG reactivity RESULTS: Seventy-five percent regained consciousness. 29/60 had a good outcome (GOS 4-5) and 31/60 had a bad outcome (GOS 1-3). According to the ROC curve, SEP findings were classified in 3 grades. Grade I (NN, NP) had PPV of 93.1% for 'awakening' and 86.2% for good outcome. Grade III (AA) had PPV of 100% for bad outcome and 72.7% for 'awakening'. Grade II (PP, NA, PA) was associated with the wider range of outcome. A multivariate analysis including SEP grading, GCS and EEG reactivity did not increase the percentage of cases prognosticated by SEP alone. CONCLUSIONS: We confirm the high predictive value of SEPs in TBI, which is greater than GCS and EEG reactivity. Indeed, SEP grades I and III were able to predict the correct prognosis in more than 80% of severe TBI. Therefore, SEPs should be used more widely in the prognosis of severe TBI. SIGNIFICANCE: Differently from post-anoxic, in post-traumatic coma the presence of normal SEPs has a favourable predictive value both for 'awakening' and disability. We think that in literature enough attention has still not been paid to this finding.