• Rosanne S Naunheim, Matthew Treaster, Joy English, Teya Casner, and Robert Chabot.
• Washington University School of Medicine, St. Louis, MO 63110, USA. naunheir@wusm.wustl.edu
• Brain Inj. 2010 Jan 1;24(11):1324-9.

Primary ObjectiveTo validate a QEEG algorithm on traumatic brain injury in an Emergency Department (ED) setting.Methods And ProceduresEEG data were collected from 105 patients with head injury (53 CT+ and 52 CT-) and 50 ED controls. Ten minutes of eyes closed resting EEG was collected from five frontal locations. A discriminant index of the probability of belonging to the TBI CT+ group was computed. Analysis of variance was computed comparing this index across the three patient groups. Using ROC curves, the p < 0.05 confidence level was determined to compute sensitivity and specificity for the TBI CT+ population.ResultsCT+ patients had a mean TBI discriminant index of 80.4, CT- patients 38.9 and controls 24.5; F = 70.2, p < 0.0001. Sensitivity was 92.45% for the CT+ group and specificity was 90.00% for the control group.ConclusionsThe TBI discriminant index appears to be a sensitive index of brain function. It may be used to suggest whether or not a patient presenting with altered mental status requires a CT scan. This index may aid in the triage of such patients in the ED. Such an easy to use, automated system may greatly enhance the clinical utility of EEG in the ED.

26 keywords...

hide…