Processed Electroencephalogram-Based Monitoring to Guide Sedation in Critically Ill Patients: A Systematic Review and International Expert Panel-Based Consensus Recommendations

Abstract Background The literature related to the use of processed EEG (pEEG) for depth of sedation (DOS) monitoring is increasing, however it is unclear how to use this type of monitoring for critical care patients within the intensive care unit (ICU). Methods We performed a systematic review of the literature according to the Grade of Recommendation assessment, Development, and Evaluation (GRADE) approach. The modified Delphi method was utilised by a team of experts to produce statements and recommendations derived from study questions. Three separate online rounds discussing 89 statements categorized into four domains were formulated. The panelists rated the appropriateness of each statement and were able to suggest modifications or addition of statements. An analysis of anonymised ratings of the statements by part of the panel followed each Delphi round and previously validated criteria were used to define appropriateness and consensus. Results Level of evidence regarding the four domains was very low. Fourteen panelists participated in the Delphi rounds and consensus was reached for 28 out of 89 statements, from which the reccomendations were created. The main findings were that DOS monitoring should be performed in critically ill patients whenever clinical evaluation is not possible, it should be performed by continuous pEEG techniques and the resulting data depicted with graphical tools to facilitate detection of excessive sedation, a potential cause of burst-suppression, and finally, structured training is suggested to achieve a basic pEEG competency. Conclusions Although evidence on using DOS monitors in ICU is scarce and further research is required in order to better define the benefits of using pEEG, the results of this consensus highlight the general agreement that critically-ill patients would benefit from this type of neuromonitoring.