Abstracts

Rationale: Studies have been done to identify duration for which patients should undergo cEEG monitoring to exclude seizures. However, It is unclear how long cEEG should be continued in different patient population based on their etiology of seizures and their mental status in order to exclude subclinical seizures. The purpose of this research is to identify the patient population that is more likely to have delayed seizure onset and hence may benefit for longer cEEG monitoring in comparison to other patients. Methods: We identified all patients who underwent cEEG monitoring at Cleveland clinic between January to December, 2016(2425 patients). Baseline demographic, clinical, and EEG findings were recorded and Data was analyzed using Wilcoxon Ranksum Text. Results: Among the 2425 patients who had EEG monitoring, 334 (13.8 %) were found to have seizures. 46 (13.8%) of these patients were found to have status epilepticus on EEG.  25 of these seizure patients were excluded from analysis because of missing data. Among the 309 patients included, the median time to seizure onset was 3 hours with a wide range from patients who had seizure detected at the time cEEG was started to those who had their first seizure at 165 hours. 120 patients (38.8 %) had their first seizure within the first hour of initiation of EEG monitoring. 24 of these patients (7.8 %) were already having seizures at the onset of initiation of EEG monitoring. Seizures were detected within the first 24 hours of cEEG monitoring in 80.3% (n=248) of patients & within the first 36 hours of cEEG monitoring in 90 % of all patients with seizures. 93.5 % (n=289) had their seizures detected within 2 days of monitoring and in 6.5 % (n=20 ), the first seizure was detected after 48 hours of monitoring.Depending on etiologies, some patient populations were more likely to have delayed onset of seizure. For most etiologies, 50% of patients with seizures had their seizures detected in less than 6 hours of EEG monitoring. However, for patients with hemorrhage (including SAH, ICH and SDH) the time to detect seizures in 50 % of the patients was delayed to 15 hours (p value: 0.0004). Seizures were detected in less than 24 hours in 80 % of patients with seizures except those who had seizures secondary to hemorrhage, trauma and tumor. Seizures were detected in less than 36 hours in 90% of patients with seizures except those who had seizures secondary to hemorrhage, tumor and those with unknown etiology of seizures.Comatose patients were more likely to have early onset of seizures compared to noncomatose patients (median: 12 min, P value 0.0035). Among awake, lethargic and comatose patients, seizures were detected in less than 2 days of cEEG monitoring in 90% of patients as against more than 3 days of cEEG monitoring required to detect seizures in 90% of stuporous patients. Conclusions: cEEG monitoring detected seizures in 13.4% (n=334) of patients undergoing cEEG monitoring with 90% of seizures detected within 36 hours of monitoring. Patients with cardiac arrest and comatose patients were more likely to have their seizures detected early on.  However, Patients with stuporous mental status and those with seizures secondary to hemorrhage, trauma and tumors are more likely to have delayed onset of seizures and hence require longer monitoring to assess seizure risk. Funding: None