S23. Mortality prediction with the status epilepticus severity score

Introduction Status epilepticus (SE) as a cause of impaired level of consciousness in intensive care unit (ICU) is reported in the international literature between 5% and 49%, probably underdiagnosed. The many different forms and stages of SE (Refractory SE, RSE and Super-Refractory SE, SRSE) are associated with neuronal injury and require optimal treatment. The Status Epilepticus Severity Score (STESS) was recently developed to predict outcome of SE. AIM: To determine the association between STESS scale and mortality in patients with epileptic status assisted in the ICU of a public hospital. Methods Analytical, observational and retrospective study of cases of nontraumatic coma with clinical or electroencephalography criteria of SE (classified according to the International League Against Epilepsy, ILAE) admitted to the ICU in a tertiary care hospital in Buenos Aires, Argentina, over a three year period (2014–2017). The STESS scale was used. EEG was recorded using 21 electrodes placed according to the International 10–20 System, at least 60 min of artifact-free recording. We excluded cases with extra-axial brain compression lesions. Results The cohort consisted of 15 subjects who presented some type of SE, mean age was 49 years ( ± 18 years), 46% were women, time from ICU admission to first EEG: 2.5 days ( ± 1.7 days). A total of 53% of patients had tonic-clonic SE, SRSE (40%) and RSE in the rest of the population. In-hospital mortality rate was 26%. The 86% presented acute brain injury (included stroke, hypoxic-ischemic encephalopathy, metabolic causes, autoimmune encephalitis and infectious encephalitis) and 13% was progressive causes (two patients with brain tumor). A total number of 13 cases (86%), presented with new onset SE without a history of epilepsy and 13% patients had active epilepsy on treatment. STESS scores were significantly higher in non-survivors compared to survivors ( P -value = 0.0035). No statistically significant association was found between STESS scores and injury. Patients with STESS scores ⩽ 3 were found to survive. Follow-up with EEG was useful in 73% patients with SE who recover consciousness once the seizure has been aborted. Conclusion Despite of the small sample size used in this study, the STESS scale at the admission could predict the evolution of SE. The STESS scores were found to be good predictor of mortality in UCI patients with SE. SE is a very heterogeneous condition in terms of clinical presentation and causes, the application of this scale can help to design better therapeutic regimens for the SE.