Brief CommunicationSalzburg Consensus Criteria for Non-Convulsive Status Epilepticus - approach to clinical application

•ACNS terminology was implemented into Salzburg Consensus criteria for NCSE (SCNC).•This provides unambiguous instructions and significant increase in specificity.•Based on these data and other findings, a modified SCNC (mSCNC) was proposed.•In mSCNC, improvement after intravenous AEDs is no longer a precondition for NCSE.•Patients with treatment-refractory seizures are adequately diagnosed and treated.

BackgroundSalzburg Consensus Criteria for diagnosis of Non-Convulsive Status Epilepticus (SCNC) were proposed at the 4th London-Innsbruck Colloquium on status epilepticus in Salzburg (2013).MethodsWe retrospectively analyzed the EEGs of 50 consecutive nonhypoxic patients with diagnoses of nonconvulsive status epilepticus (NCSE) at discharge and 50 consecutive controls with abnormal EEGs in a large university hospital in Austria. We implemented the American Clinical Neurophysiology Society's Standardized Critical Care EEG Terminology, 2012 version (ACNS criteria) to increase the test performance of SCNC. In patients without preexisting epileptic encephalopathy, the following criteria were applied: (1) more than 25 epileptiform discharges (ED) per 10-second epoch, i.e., > 2.5/s and (2) patients with EDs ≤ 2.5/s or rhythmic delta/theta activity (RDT) exceeding 0.5/s AND at least one of the additional criteria: (2a) clinical and EEG improvements from antiepileptic drugs (AEDs), (2b) subtle clinical phenomena, or (2c) typical spatiotemporal evolution. In case of fluctuation without evolution or EEG improvement without clinical improvement, “possible NCSE” was diagnosed. For identification of RDT, the following criteria were compared: (test condition A) continuous delta-theta activity without further rules, (B) ACNS criterion for rhythmic delta activity (RDA), and (C) ACNS criteria for RDA and fluctuation.ResultsFalse positive rate in controls dropped from 28% (condition A) to 2% (B) (p = 0.00039) and finally to 0% (C) (p = 0.000042). Application of test condition C in the group with NCSE gives one false negative (2%). Various EEG patterns were found in patients with NCSE: (1) 8.2%, (2a) 2%, (2b) 12.2%, and (2c) 32.7%. Possible NCSE was diagnosed based on fluctuations in 57.1% and EEG improvement without clinical improvement in 14.2%.ConclusionThe modified SCNC with refined definitions including the ACNS terminology leads to clinically relevant and statistically significant reduction of false positive diagnoses of NCSE and to minimal loss in sensitivity.This article is part of a Special Issue entitled “Status Epilepticus”.