• Johanna Wennervirta, Tapani Salmi, Markku Hynynen, Arvi Yli-Hankala, Anna-Maria Koivusalo, Mark Van Gils, Reino Pöyhiä, and Anne Vakkuri.
• Department of Anesthesiology and Intensive Care Medicine, Helsinki University Hospital, 340, 00029, Helsinki, Finland. johanna.wennervirta@hus.fi
• Intensive Care Med. 2007 Jan 1; 33 (1): 133-6.

ObjectiveTo evaluate the usefulness of entropy and the bispectral index (BIS) in brain-dead subjects.Design And SettingA prospective, open, nonselective, observational study in the university hospital.Patients And Participants16 brain-dead organ donors.InterventionsTime-domain electroencephalography (EEG), spectral entropy of the EEG, and BIS were recorded during solid organ harvest.Measurements And ResultsState entropy differed significantly from 0 (isoelectric EEG) 28%, response entropy 29%, and BIS 68% of the total recorded time. The median values during the operation were state entropy 0.0, response entropy 0.0, and BIS 3.0. In four of 16 organ donors studied the EEG was not isoelectric, and nonreactive rhythmic activity was noted in time-domain EEG. After excluding the results from subjects with persistent residual EEG activity state entropy, response entropy, and BIS values differed from zero 17%, 18%, and 62% of the recorded time, respectively. Median values were 0.0, 0.0, and 2.0 for state entropy, response entropy, and BIS, respectively. The highest index values in entropy and BIS monitoring were recorded without neuromuscular blockade. The main sources of artifacts were electrocauterization, 50-Hz artifact, handling of the donor, ballistocardiography, electromyography, and electrocardiography.ConclusionBoth entropy and BIS showed nonzero values due to artifacts after brain death diagnosis. BIS was more liable to artifacts than entropy. Neither of these indices are diagnostic tools, and care should be taken when interpreting EEG and EEG-derived indices in the evaluation of brain death.

12 keywords...

hide…