Abstracts

Rationale: Cardiac surgery is associated with central nervous system morbidity and mortality. The literature suggests that seizures are strong independent predictors of poor neurological outcome and mortality after cardiac surgery, though current literature does not include continuous EEG data (cEEG). In this study, we present cEEG data on all patients in the cardiac surgery intensive care unit (CSICU) who suffered neurological problems after surgery.Methods: A retrospective chart review in the CSICU was performed from 11/1/2010- 6/1/2011 using a prospective database of all adults who underwent cardiac surgery at Mount Sinai Hospital. According to clinical practice in the CSICU any patient who does not regain consciousness, has a neurological deficit or a clinical event consistent with a seizure within 24h of cardiac surgery is consulted by neurology and connected to cEEG monitoring. We compared EEG results, imaging results and outcomes.Results: Over the 7 month study period, a total of 696 people underwent cardiac surgery, of which 12 (2%) met requirements for neurological consult. Of these 12 patients, 5 (42%) people did not regain consciousness, 3 (25%) had neurological deficits and 4 (33%) had clinical events consistent with seizure. The cEEG revealed 2 (16%) patients in electrographic non-convulsive focal status epilepticus (NCSE). The original consult reason in these patients was failure to regain consciousness. In the 10 remaining patients, no other electrographic seizures (ESZ) were captured on cEEG. Periodic epileptiform discharges (PEDs) were present in 3 (25%) cEEGs. One patient had both PEDs and focal NCSE. Interestingly, in this patient the PEDs were on the left, and the NCSE was on the right. Three people (25%) had rhythmic jerking of their body with no EEG correlate suggesting a subcortical source. The EEG was reactive to noxious stimulation in 5 people (42 %). Non-contrast head ct-scans were performed in all 12 patients. Six (50%) had new infarcts on their ct-scans, 1 (8%) had diffuse loss of grey white differentiation and cerebral edema, and 5 (42%) had no acute findings. All patients with NCSE or PEDs had an acute stroke on imaging. Five (42%) expired during the admission, 4 (33%) were discharged to home, and the remaining 3 (25%) were discharged to acute rehab. One of the patients in NCSE expired. The other patient who had both NCSE and PEDs received aggressive treatment for her subclinical seizures and ultimately was discharged to acute rehab. All of the patients with an EEG reactive to noxious stimulation were discharged to either home or acute rehab. Conclusions: In this retrospective study of CSICU patients with neurological complications following cardiac surgery, NCSE and/or PEDs occurred in 33% and was associated with an acute stroke on imaging. While connected to cEEG, 25% had rhythmic jerking that was not a seizure, but may have been treated as such without the EEG data. A reactive EEG appears to be a good prognostic indicator.