Abstracts

Rationale: Rationale - Eletroencephalography (EEG) plays a central role in diagnosis and management of patients with epilepsy, but routine short-term EEG records present a low sensitivity, ranging around 50%, mainly regarding the presence of interictal epileptic discharges (IED). Background abnormalities of cerebral activity such as focal slow activity or attenuation may be more prevalente, although with a low specificity to seizure disordes. On short-term EEG, clinical seizures are even less frequently seen (5-7%), but if present they give great oportunity to diagnose and classify these patients. Our main objective is to identify eletroencephalographical markers of patientes more likely to experience a seizure during routine short-term EEG recording. Methods: Methods - Retrospective review of ... patients undergoing outpatient 30-minute routine EEG was done. Eletroencephalogram data were divided into interictal abnormalities including analysis if background activity, presence of IEDs (morphology, distribution and latency to first IED) and ictal findings (clinical and eletrographically manifestation). Results: Results - 11.657 recordings of 10.140 patients were reviewed. 6.597 patients were outpatients under investigation of epileptic disorders. Out of these patients, 302 (4.5%) had seizures during short-term EEG. 217 (71.8%) patients were under 18 years old. Focal seizures were recorded in 140 (46.3%) patients, and generalized seizures in 162 (53.7%). There were more pediatric patients with generalized seizures than adults, but there was no significant difference among boht groups (p=0.3). Abnormal background was seen on 130 patients, with no difference between those presenting focal seizures and those presenting generalized seizures (p=0.2), but focal slowing was more prevalente among patients with focal seizures (p=0.005). When compared to patients with epilepsy, but without seizures during routine EEG, background abnormalities and IED characteristics were not different. Conclusions: Conclusions - Eletroencephalographic findings did not help to predict seizures during routine EEG in our series.