Abstracts

Rationale: Several studies have described the use of continuous electroencephalography monitoring (cEEG) in the neurological intensive care unit (NICU). In these patient populations, cEEG has been shown to detect non-convulsive seizures in up to one third of patients. However, there have been few studies evaluating the use of cEEG outside of the NICU. We describe cEEG findings in critically ill patients in the post-surgical cardiothoracic intensive care unit (CICU).Methods: A total of 74 patients admitted to the CICU at Emory University Hospital between January 2008 and August 2010 underwent cEEG monitoring. EEG reports were retrospectively reviewed to assess for presence of clinical or subclinical seizures. Clinical data was also collected including type of cardiothoracic surgery, history of clinical seizures prior to monitoring, other medical and neurological complications as well as functional outcome and mortality based on Glasgow Outcome Scale (GOS) at discharge. Results: The majority of patients were admitted for cardiac valve replacement or major vessel aneurysm repair 47/74 (63%). Other surgeries included coronary bypass grafting (CABG) and pulmonary procedures. The median number of days of cEEG monitoring was 2. Witnessed convulsion was reported prior to cEEG in 57/74 (77%) and was the most common indication for cEEG while evaluation of altered mental status accounted for the remainder. There were no significant differences in age, gender, history of epilepsy, or the incidence of acute, hepatic failure or new structural lesions on neuroimaging between patients who experienced seizures and those who did not. (p > 0.05). Twenty-two percent of patients (16/74) experienced seizures during cEEG and the majority of these were completely subclinical 11/16 (69%). Status epilepticus was detected in 6/74 (8.1%) and one-third of these were entirely subclinical. Having a witnessed convulsion prior to cEEG was a strong predictor of subsequent clinical or subclinical seizures on cEEG (LR = 4.49, p = 0.034). Mortality at discharge (GOS=1) was higher in patients with isolated seizures (p = 0.015) as well as status epilepticus (p = 0.001). Status epilepticus was associated with poor outcome (GOS 1, 2 or 3) at discharge (p = 0.037) although the presence of isolated seizures (p = 0.25) was not. Conclusions: In the post surgical CICU, seizures are frequently seen in patients undergoing cEEG and the majority are purely subclinical. Having a prior clinical seizure was a strong predictor of subsequent clinical or subclinical seizures. Patients with status epilepticus had worse functional outcome (GOS 1, 2 or 3) while mortality rates (GOS 1) were higher in patients with either status epilepticus or isolated seizures on cEEG. CICU patients with witnessed clinical seizures should be seriously considered for cEEG monitoring not only for detection of subsequent seizures but also to aid in prognosis. Further studies of the utility of cEEG for detection of seizures and prediction of outcome in post surgical cardiothoracic patients are warranted.