• M Matousek, E Takeuchi, J E Starmark, and D Stalhammar.
• Department of Clinical Neurophysiology, Sahlgren Hospital, Göteborg, Sweden.
• Acta Anaesthesiol Scand. 1996 Aug 1;40(7):824-31.

BackgroundThe aim of the study was to investigate the correlation between EEG indicators and clinical scores based on the RLS85 (Reaction Level Scale 85) in comatose patients. The results of a simple visual assessment of the EEG, using an arbitrary scale with typical EEG patterns, were compared with those obtained by quantitative electroencephalography (qEEG).MethodThe RLS85 scores were examined in 34 patients with impaired consciousness due to brain tumours, vascular lesions or head injuries. The EEG was recorded shortly before or after the clinical examination. The semiquantitative assessment was made by visual inspection of the tracings, using an arbitrary scale where 12 EEG patterns with increasing proportion of slow activity were displayed. Parallel to the visual analysis, the EEGs were processed by means of EEG spectrum analysis and the power/amplitude in slow frequency bands was used as an indicator. The results were based on correlation between various types of EEG variables and the RLS scores which were obtained in the same patients.ResultsThe correlation between the visual EEG indicators and coma scores ranged between 0.53 and 0.57 (P < 0.01). As regards the computerised EEG analysis, the correlation between the clinical scores and various EEG spectrum values did not exceed 0.45 (P < 0.01). A higher correlation could be obtained by combining eight EEG variables; the multiple correlation coefficient was then 0.68.ConclusionsThe amount of EEG slow activity is significantly correlated to the RLS85 score. This means that the EEG also provides information on the level and not only on the changes of the coma degree. Surprisingly, the indicators based on quantitative EEG, as used in commercially available instruments, did not give better results than the visual assessment. However, the results of the computerised analysis could be improved using multivariate statistical methods. The study also showed a way to improve communication between the neurophysiologist and clinician by presenting the EEG findings in terms similar to those used in the clinical scales. However, the clinician should be aware of the fact that the "EEG score" and the clinical score is not the same: the intention is to supplement rather than to simulate the clinical observation.

12 keywords...

hide…