Abstracts

Rationale: Non-convulsive status epilepticus (NCSE) is an underdiagnosed entity and EEG is an essential tool for the diagnosis and management. Decisions based on EEG findings is a real challenge, because the difficulties about the interpretation of EEG patterns. Salzburg Criteria (SC) were validate in patients with NCSE. Our objective was to evaluate if the application of SC was replicable in our population of patients with suspected of NCSE. Methods: We reviewed, retrospectively, the medical records of patients with suspicion of NCSE recruited since June 2015 to July 2017. They were divided into 2 groups according to the final diagnosis: NCSE and not in NCSE. EEGs were analyzed, applying the SC, and reclassified in NCSE, possible-NCSE (P-NCSE) or not in NCSE. Etiology and abnormalities on MRI were evaluated too. Results: Were included 66 patients with suspicion of NCSE. Thirty-three had a final diagnosis of NCSE: 63.6% males, mean age 57 years, SD 19.6. The etiology was acute symptomatic in 60.6% and 81.8% had an epileptogenic lesion on MRI. When we applied the SC on EEGs, 75.8% of patients fulfilled criteria for NCSE and 24.2% for P-NCSE. For those patients not in NCSE 90.9% fulfilled criteria for not in NCSE and 9.1% for P-NCSE. In our analysis, SC scale had a sensitivity of 100% (if definitive NCSE + P-NCSE) and a specificity of 90.9%, with a positive predictive value of 91.6% and a negative predictive value of 100%. Conclusions: In our population, SC showed a high diagnostic accuracy for patients with clinical suspicion of NCSE. False positives may be found in P-NCSE and should be analyzed case by case. Funding: None