Abstracts

Rationale: An increasing number of neuro-intensive care units (NICU) have recently started to perform continuous video-EEG monitoring (VEM) in order to capture epileptiform discharges (ED), non-convulsive seizures (NCS) and non-convulsive status epilepticus (NCSE) in critically ill patients. In this retrospective study we investigated the frequency of ED, NCS and NCSE during continuous VEM in patients admitted to our NICU.Methods: All the patients who had VEM in our NICU during the last 15 months were included in the study. Video-EEG recordings were obtained by a 32 channel digital EEG device (Grass Telefactor Comet). Filter settings were 1-70 Hz (including notch filter), recording sensitivity was adjusted accordingly. Recordings were analyzed by a spike-seizure detection software and visual inspection of at least 20-25 minutes of one-hour epochs of continuous EEG. Results: Fifty patients (25 F, 25 M; age:21-85 (mean 59) years) were included. Each patient had 1-4 EEG recordings, lasting 1-192 (mean 42) hours. Etiology was heterogenous (e.g. reversible leukoencephalopathy syndrome, limbic encephalitis, herpes encephalitis, etc.), however most of the patients (n=29, 58%) were admitted with a diagnosis of cerebrovascular disease. Eleven patients (22%) had ED, which were periodic in 82%. Five patients (10%) had NCS or NCSE. Fifteen patients had ?2 recordings, and NCS were captured in the second recording of two (2/15, 13%) of them. Six of 11 patients who died (54%) had ED (?2 test; p=0.003), while only two patients (2/11, 18%) had NCS/NCSE (Fisher s ?2 test; p=0.301). Conclusions: In our NICU almost 20% of patients had ED. Ten percent had NCS/NCSE and sometimes more than one EEG recordings were needed to record them. Patients with ED had a higher mortality rate. In our prospective studies, we are planning to investigate the effect of early recognition of ED or NCS/NCSE, and prompt initiation of anticonvulsant medication on the prognosis of patients.