EEG in Cardiac Surgery – Moving Past The Obvious
Various strategies for intraoperative cerebral monitoring have been evaluated over the decades, with electroencephalography (EEG) monitoring as one of the oldest. There are reports in the literature from as early as 1937 comparing EEG (referred to as electro-encephalogram) to an electrocardiogram with evaluation of this technique for the detection of changes in brain activity with drug administration.

Researchers investigate the use of processed EEG (pEEG)-guided anesthetic management in cardiac surgery patients. They present a retrospective cohort analysis in patients undergoing cardiac surgery with cardiopulmonary bypass (CPB) comparing 150 with intraoperative pEEG with 150 historical controls without implementation of pEEG. Practice change for implementation of pEEG during cardiac surgery occurred in 2017 at the author's institution allowing comparison between the two groups. The authors assessed if pEEG-guided anesthetic management led to ease in CPB separation and decreased use of vasoactive and inotropic requirements (as measured by vasoactive and inotropic score [VIS]) in the intensive care unit (ICU). Although pEEG use did not impact successful separation from bypass it did lead to a significantly lower vasoactive and inotropic drug requirement upon arrival in the ICU.

There is no gold standard for determining the “best anesthetic” that balances adequate depth while limiting the risks of anesthetic drugs. Given that the brain is the target organ for inhaled and intravenous anesthetic agents, cortical brain activity analysis can potentially assist with goal-directed titration. Advances in monitoring and diagnostics in cardiac anesthesia are unparalleled, yet reliable, real-time assessment of brain function and perfusion is lacking. Traditional EEG monitoring during cardiac surgery utilized scalp electrodes in a particular configuration and required an understanding of pattern recognition for detecting abnormal brain activity. The sheer complexity of these systems limited routine clinical use. Technological advancements have led to modifications that simplify the broader applicability for implementation in clinical care. Algorithms reduce raw EEG data into pEEG indices that produce a dimensionless number that is easy to interpret. However, problems with precision and reliability are not uncommon with these new systems. For instance, some studies suggest that bispectral index (BIS) decreases with neuromuscular blockade in awake patients to levels congruent with general anesthesia. Additionally, baseline cognitive dysfunction, age, and metabolic perturbations can influence readings produced by pEEG. These findings suggest that pEEG can be highly variable, unreliable, and potentially lead to erroneous management.