Can occurrence of convulsive status epilepticus during video telemetry be predicted? Observations from an epilepsy monitoring unit.

PURPOSE: The objective of the study was to evaluate the frequency, clinical, and video-electroencephalographic (VEEG) predictors of convlusive status epilepticus (CSE) in the epilepsy-monitoring unit (EMU). METHODS: The data of all patients who had CSE in our EMU between 2008 and 2017 were reviewed. For each case, two age- and diagnosis-matched subjects who underwent VEEG and did not develop CSE were taken as internal controls. Electro-clinical data of both the groups were compared. Predictors of CSE were assessed using logistic regression analysis. RESULTS: Out of 11,188 video-telemetries were conducted between 2008-17, forty-three events of CSE (0.38%) were recorded. On comparisons with 86 internal controls no differences were apparent on prevalence of cognitive impairment, structural lesion, number of baseline anti-seizure medications (ASM), ASM taper schedule, ictal patterns, and duration of VEEG monitoring. Inter-ictal rhythmic periodic patterns had significantly higher prevalence in cases (p = 0.028). Logistic regression analysis revealed that odds of CSE were higher with past history of SE [p = 0.008; adjusted odds ratio (OR) = 5.48 (confidence intervals {CI} 1.55-19.28)] and in presence of rhythmic spike and wave discharges [p = 0.016, OR = 33.518(CI = 1.93-581.4)]; the odds were lower if the first two seizures recorded did not evolve into CSE [p = 0.009, OR = 0.247 (CI = 0.08-0.70)] and if there was prior history of daily seizures [p = 0.02, OR = 0.250 (CI = 0.07-0.84)]. CONCLUSION: CSE is a rare yet important adverse event in EMU. Clinical predictors are more relevant in comparison to EEG variables. Extent of ASM withdrawal may not directly account for occurrence of CSE; factors inherent to a patient's epilepsy are deterministic.