Abstracts

Rationale: Neurological prognostication in post-anoxic comatose patients remains a challenge. Continuous EEG monitoring (CEEG) with background and reactivity assessment provide important information. However, it is still unclear how best to perform reactivity testing and which method is superior. We aimed to determine which stimuli are associated with a better chance of identifying EEG reactivity and with a better inter-rater reliability. In a second way, we aimed to assess the prognostic value of EEG reactivity using a standardized stimulation protocol and standardized criteria, and to compare its inter-rater reliability and prognostic value to those of background assessment. Methods: Prospective cohort of comatose patients after cardiac arrest admitted to the intensive care unit of an academic medical center between January 1, 2017, and December 31, 2018, and receiving CEEG. Reactivity was assessed using a standardized sequential multimodal stimulation protocol by EEG technicians and standardized definitions. Background was assessed using pre-defined categories based on a standardized EEG terminology. Gwet’s AC1 and percent agreement were used to measure inter-rater agreement (IRA). Ability to predict unfavorable outcome (defined as a Cerebral Performance Category of 3 to 5: severe disability to death) was measured with sensitivity and false positive rate (FPR). Results: One-hundred and fifteen patients were included and 242 EEG epochs were analyzed. Several clinical variables (age, initial rhythm, myoclonus, pupillary reflexes) were associated with unfavorable outcome. The IRA for reactivity to individual stimuli varied from moderate to good (AC1:58-69%), while the IRA for background assessment was very good (AC1:81%). All stimuli elicited reactivity with a similar rate (23-42%). The prognostic value of background assessment for unfavorable outcome was overall better than reactivity, for both sensitivity (60% vs. 48%) and FPR (4% vs. 8%). Conclusions: Despite standardized stimulation protocol and definitions, background assessment shows better IRA and predictive value than reactivity testing in post-anoxic comatose patients. Funding: No funding