(1024) Processed Electroencephalography to Evaluate Sedation in Patients Receiving Paralysis in the ICU

Introduction: Titrating sedation in patients receiving neuromuscular blockade (NMB) is challenging. Processed electroencephalograms (pEEG) may correlate with sedation depth, but there are a paucity of data evaluating pEEGs in the intensive care unit (ICU).

Methods: We prospectively enrolled patients receiving continuous NMB between April 2023 and July 2024 in a single academic ICU. We excluded patients with neurologic impairment or injury. We measured pEEG using SEDLine (Masimo Corporation) with four frontal EEG electrodes applied after NMB initiation and continued up to 72 hours. We recorded demographics, sedatives, and pEEG parameters (Patient State Index (PSI), and Suppression Ratio Percent (SR%)). PSI is calculated based on EEG frequency composition and amplitude and ranges from 0 (isoelectric) to 100 (awake). We categorized PSI into three categories: 0-24 (possible over-sedation), 25-50 (general anesthesia), >50 (possible under-sedation). SR quantifies the proportion of a given pEEG segment that is suppressed < 2uV, with SR >80% indicating potential over-sedation . We quantified the number of patients with PSI < 25 and the duration of time PSI was < 25. We also quantified the number of patients with PSI >50 for >1 minute. Providers were blinded to the data and sedation levels were not titrated based on pEEG.

Results: We enrolled 13 patients (7 women, median age 49 (IQR 34-67) years) and monitored pEEG for a median of 28 (IQR 18-47) hours. Patients received median 3 (IQR 2-4) continuous sedative infusions (CSI). All patients received fentanyl, 11 propofol, 8 midazolam and 7 ketamine via CSI. Median PSI was 22 (IQR 21-41). The median percentage of time the PSI was < 25, 25-50 and >50 was 79%, 15%, and 6%, respectively. Six patients (46%) had PSI < 25 for over 80% of the monitoring window (median duration 17 hours (range 3 to 69) and SR was > 80% in four patients. PSI was >50 for at least one minute for nine (69%) patients. There were no adverse events related to the application of the pEEG.

Conclusions: Processed EEG monitoring was safe and feasible in this study. Depth of sedation estimated by pEEG was variable. Our data suggests many patients may have been over- and under-sedated. Future studies evaluating pEEGs to titrate CSI during NMB are warranted.