The Association of Preoperative Cognitive Dysfunction to Common Intraoperative Electroencephalographic Parameters and...

Older cardiac surgery patients have a higher prevalence of cognitive dysfunction and elevated risk of perioperative neurocognitive disorders (PND), both independently related to adverse postoperative outcomes. Neuromonitoring using electroencephalogram (EEG) and cerebral oximetry (CO) may predict PND. However, preoperative factors influencing intraoperative neurophysiological characteristics are not well understood. We conducted a study in a cardiac surgery cohort to better understand the relationship of preoperative cognitive dysfunction to intraoperative burst suppression (BS), spectral edge frequency (SEF), cerebral hypoxia/desaturation, and dual cerebral events involving both BS and cerebral desaturation to potentially link preoperative cognitive dysfunction to intraoperative neuromonitoring variables associated with PND.

This is a secondary analysis of a triple-blinded, ongoing, multi-center randomized trial assessing the efficacy of postoperative intravenous acetaminophen to reduce postoperative delirium (POD) in older cardiac surgery patients. We studied 110 patients ≥60 years who underwent CABG and/or valve surgery under general (inhalational) anesthesia at a single academic center. Preoperative cognitive status was assessed using the Montreal Cognitive Assessment (MoCA) and classified as normal (MoCA score≥26) or impaired (MoCA <26). Intraoperative frontal electroencephalogram data were recorded using the EEG monitor (SedLine, Masimo Inc). BS was detected using a recursive variance estimation algorithm, quantifying burst suppression duration (BSD). SEFs were derived through multi-taper spectral analysis. Intraoperative cerebral oxygenation was measured via cerebral oximetry (O3, Masimo), identifying cerebral desaturations (CO values <60%). Univariate analyses assessed associations between preoperative cognitive dysfunction and BSD, SEF, cerebral desaturation, and dual cerebral events of BS and cerebral desaturation. Multivariable regression analyses for these variables controlled for demographics and intraoperative confounders.

Baseline characteristics were comparable between the groups. There was no statistically significant correlation between preoperative cognition and BSD in cognitively impaired individuals (Cohen's d = 0.33; P = .09; remaining insignificant on adjustment, P = .8). Adjusted analyses showed those with abnormal MoCA scores had an average of 1.4 Hz lower SEF values (95% confidence interval [CI], 0.07-2.6; P = .03). Cognitively impaired patients demonstrated no significant increase in time spent in cerebral desaturations (55.4 [12.4-119] vs 46.3 [19.2-81.9] minutes; P = .6). No disparities were observed between the groups regarding concurrent and nonconcurrent abnormal EEG and CO values.

Preoperative cognitive dysfunction was associated with significantly lower SEFs, indicating increased isoflurane sensitivity without affecting BS or correlating with CO. SEF shows potential as a marker for cognitive vulnerabilities, but further studies are needed to validate its clinical utility and establish thresholds to optimize perioperative care.