Abstracts

Rationale: Emergent EEG is frequently ordered by emergency physicians. This study was performed to assess the yield of emergent EEG in case of correct indication. To determine the correct indications, we used criteria adapted from the Tour guidelines (Neurophysiol Clin 1997; 27: 373-405 and Neurophysiol Clin 1998; 28: 103-153): an emergent EEG is required in case of acute confusion or coma to rule out non convulsive status epilepticus (NCSE), or in case of a possible first seizure to increase the chance of picking up interictal epileptiform discharges. Methods: All EEGs with a correct indication requested by emergency physicians between 11-2009 and 05-2010 were prospectively included. Patient history, EEG findings and final diagnosis were reviewed. An EEG was considered useful if it showed epileptiform discharges in a patient with a final diagnosis of seizure/epilepsy or if it allowed confirming or excluding NCSE. Results: Ninety-three EEGs were reviewed. The referring diagnosis were possible first seizure (69/93), acute confusional state (17/93) or coma (7/93). In the possible first seizure group, there were 51/69 episodes of transient loss of consciousness and 18/69 episodes of transient neurological symptoms. Interictal epileptiform discharges were seen in 10/69 patients of the first seizure group, in 1/7 patient of the coma group and in no patient of the acute confusion group. Nine out of 93 EEGs showed focal discharges and 3/93 generalized discharges, including 2/93 with GPEDs (one patient had both focal discharges and GPEDs). Ictal activity, suggesting NCSE, was identified in 4 patients, all in the coma group. The most common final diagnoses in the possible first seizure group were first seizure (25/69), syncope (14/69), non-epileptic psychogenic seizure (6/69), stroke or TIA (5/69) and migraine (2/69). No diagnosis was found in 10/69 patients. In the acute confusion and coma group, the most common diagnoses were toxic/metabolic encephalopathy (6/24), dementia (6/24) and NCSE (4/24). Patients with a final diagnosis of first seizure or de novo status epilepticus were diagnosed with either symptomatic epilepsy (13/26), cryptogenic epilepsy (3/26), primary generalized epilepsy (2/26), provoked seizures (6/26) or seizure of unknown origin (2/26). Among them, 9/26 patients had interictal epileptiform discharges and 1/26 was in NCSE. No epileptiform discharge was seen in patients with a diagnosis of syncope. Of all the EEGs reviewed, 34/93 contributed significantly to the diagnosis by showing epileptiform discharges (9/93) or ruling in (4/93) or out (21/93) NCSE as the cause of coma or confusion. One EEG was misleading as it showed epileptiform discharges in a patient without a history of seizures but the correct diagnosis of stroke was evident from history and the CT-scan. Conclusions: Our data suggest that, when properly ordered, emergent EEG is an efficient tool to diagnose epileptic disorders, and to rule out NCSE. Better clinical assessment of a possible first seizure is probably needed to increase its yield. Further study is required to define criteria to select patient at risk of NCSE.