Abstracts

Rationale: The advent of therapeutic hypothermia has greatly improved survival for patients affected by cardiac arrest. In the cooling period and subsequent normothermic period, EEG background reactivity is a favorable prognostic marker of a good outcome. Some patients however will go on to die despite an initially favorable EEG record. We present 3 cases where such a progression to poor outcome was captured on continuous EEG monitoring and highlight their clinical and EEG data.Methods: Each cardiac arrest patient that is eligible for therapeutic hypothermia is placed on continuous EEG monitoring in our institution. This data is collected prospectively. Each patient is tested according to our protocol during the hypothermic period and immediately following normothermia (NT). During the normothermic period the FOUR score is calculated. A baseline EEG background rhythm post therapeutic hypothermia and rewarming is established and defined as the best background (based on organization and presence of faster frequencies) after achieving normothermia. We defined deterioration as any change from baseline that satisfied the following criteria: 1. Presence of generalized epileptiform activity, defined as GPD (+/- triphasic morphology, given poor inter-rater agreement) or GSW, according the new ACNS terminology 2. Emerging discontinuity if background suppressions lasted less than 49 % of the record 3. Burst suppression pattern if background suppressions lasted 50-89% of the record 4. Diffuse suppression if suppressions lasted >90% of the record > 50% loss of alpha OR beta frequencies, if initially present at baseline We included all 22 patients with cardiac arrest whom received therapeutic hypothermia between October 2014 and April 2015.Results: We identified 2/22 (9%) patients that died despite undergoing therapeutic hypothermia and showing an initially reactive EEG or clinical exam. 1/22 (4.5%) patient was not tested initially using our reactivity protocol however showed deterioration of EEG background rhythm emerging in normothermia. (Table 1)Conclusions: Deterioration of EEG background in cardiac arrest patients treated with hypothermia has been described though little data exists on this topic and there is no widely accepted definition of “background deterioration”. We hypothesize that there is delayed neuronal injury that can be evaluated through use of cEEG. Cardiac arrest patients treated with hypothermia are vulnerable to further injury and delayed deterioration despite being cooled and may represent a potentially salvageable population as prolonging the cooling period or re-initiating hypothermia could be envisaged in this situation. Further prospective studies are needed to determine which clinical factors or emerging EEG patterns place these patients at higher risk of deterioration and more importantly during which time frame this deterioration takes place in order to better inform clinical practice.