Abstracts

Rationale: In the intensive care settings, seizures and status epilepticus are a common neurologic complication, which is attributable to primary neurologic pathology and secondary to critical illness. Seizures in ICU may be difficult to recognize as they often lack clinical manifestations, such as rhythmic convulsions, and require EEG for diagnosis. It is important to properly diagnose the seizure type and its risk factors to ensure appropriate therapy. Our aims in this study are to determine the Incidence of clinical and electrographic seizure in patients admitted to Memorial Sloan Kettering Cancer Center (MSKCC) ICU in 2016-2017 and to determine risk factors/predictors for clinical seizure. This is the first study that looked into the incidence of seizure in all patients admitted to ICU (not in the patients with EEG results). Previous studies evaluated the incidence of electrographic seizures among patients with EEG in ICU; our outcomes were a clinical and electrographic seizure. We had a wide variety of predictor variables in our data set compared to the other studies.  Methods: Among 1,059 patients admitted to medical ICU at MSKCC from 2016 to 2017, data for patients who had continuous EEG or routine EEG was obtained (n=184). Data reviewed, and we identified patients who had an electrographic seizure on routine or continuous EEG for clinical indications. We also identified patients who had clinical seizures. We used multivariate logistic regression in our analysis.  Results: Among all patients, the incidence of combined clinical and the electrographic seizure was 50 patients among 184 patients (incidence of 47.2 among 1000 patients). The incidence of clinical seizure, electrographic seizure, clinical seizure without EEG correlate, clinical seizure with EEG correlate, non-convulsive seizure (NSC), non-convulsive status epilepticus (NCSE), and convulsive status epilepticus (CSE) were 40.6, 10.4, 35.9, 4.7, 7.6, 2.8,  and 5.7 per 1000 respectively. In stepwise regression model history of seizure (OR:2.2, PV: 0.05),b brain metastasis (OR: 2.5, PV: 0.03), being of vasopressor in ICU admission (OR: 2.9, PV: 0.03) Age>65 (1.4, 0.02), length of VEEG (OR:3.8, PV:0.01) increased the risk of clinical and electrographic seizures while controlling for other confounding factors. Length of continuous EEG >24h significantly increased the risk of clinical and electrographic seizure detection (OR 2.5 and 10.8 respectively) while controlling for other confounding factors. Being on chemotherapy in the past 30 days increased the risk of electrographic seizure significantly by 10.4 folds (PV: 0.03) while controlling for other confounding factors.  Conclusions: The ICU patient population with a history of seizure, brain metastasis, being of vasopressor in ICU admission, age>65, and being on chemotherapy in the past 30 days are in an increased risk of clinical and electrographic seizures. Length of continuous EEG >24h significantly increased the odds of clinical and electrographic seizure detection, therefore we recommend to obtain continuous EEG >24h in the high-risk patient population. Funding: No funding