Abstracts

Rationale: Non-convulsive seizures (NCS) and non-convulsive status epilepticus (NCSE) are common in critically ill patients, may be detected by continuous EEG (cEEG) monitoring, and detection often prompts initiation of anticonvulsants. However, there are little data to guide cEEG implementation or NCS/NCSE treatment. We aimed to describe the current practice of cEEG and NCS/NCSE management in critically ill patients. Methods: An anonymous survey was distributed by email to members of the American Epilepsy Society, American Clinical Neurophysiology Society, and Child Neurology Society. Descriptive statistics are utilized. Results: 346 surveys were completed. Participants were pediatric neurologists (55%), adult neurologists (32%), and both (13%). Areas of practice included epilepsy, neurocritical care, and general neurology. 84% worked at large academic or tertiary care hospitals. 81% had EEG and cEEG available 24/7, and 52% monitored 6 or more ICU patients per month. The most common indication for cEEG was altered mental status with recent convulsive seizures; next most common indication was subtle abnormal movements and altered mental status without convulsive seizures. 68% would monitor patients with any type of acute encephalopathy, while 32% had more specific criteria, including encephalitis, hypoxic ischemic encephalopathy, and traumatic brain injury. In patients at risk for NCS, 52% would order a 30 minute EEG initially, while 32% would begin with cEEG. 73% would call in a technologist if needed. If no NCS were detected, monitoring would be discontinued after 1 hour (16%), 24 hours (46%), or 48 hours (17%). If PEDs were present cEEG would continue for 24 hours (38%), 48 hours (29%) or 72 hours (16%). If NCS are identified 79% initiate treatment, with 68% aiming to terminate all NCS, 27% tolerating up to 5 NCS per day, and 7% tolerating up to 10 NCS per day. The most commonly utilized initial medications for NCS were fosphenytoin/phenytoin, lorazepam, and levetiracetam and for NCSE were fosphenytoin/phenytoin and lorazepam, with substantial variability in 2nd and 3rd line medications (Image 1 and 2). For NCS, 47% would utilize coma inducing medications after a 3rd medication failed while 17% would never induce coma. For NCSE, 56% would utilize coma inducing medications after a 3rd line medication failed, and 7% would never induce coma. The most commonly utilized coma inducing medications were midazolam, pentobarbital, propofol, and phenobarbital. Conclusions: Substantial variability exists in use of cEEG and duration of cEEG. If NCS/NCSE are identified, most physicians would utilize anticonvulsants and eventually coma inducing medications if refractory and most would manage NCSE more aggressively than NCS. However, there is great variability in medications utilized and treatment goal. Many issues related to cEEG in critically ill patients and NCS/NCSE management are in a state of equipoise, suggesting study is both needed and ethical. Multi-site studies will need to account for substantial variability in current clinical practice to ensure enrollment goals are feasible.