Auditory event-related potentials (P300) in the identification of psychogenic nonepileptic seizures

The feasibility and conceivable value of postictal event-related potential (ERP) recordings were studied in patients with nonepileptic seizures (NES) admitted for long-term video/EEG monitoring. Ten patients with NES underwent preictal (on hospital admission) and postictal (less than or equal to 6 hours after seizure) ERP recordings of an auditory oddball paradigm. Additionally, 10 temporal lobe epilepsy (TLE) patients with partial seizures and secondary generalization underwent preictal, postictal (<6 hours after seizures), and interictal (7-48 hours after seizure) ERP recordings. We recently reported that ERPs recorded in TLE patients with partial epilepsy undergo a temporary change postictally, while returning to their preictal state during interictal recordings. In the current study intraclass correlations, transformed into z scores, are used to determine test-retest validity of repeated ERP recordings. An independent sample t test with z scores for the comparison of preictal and postictal recordings showed that ERP activation differed between NES and TLE patients (P = 0.009). More specifically, ERP recordings in the preictal and postictal states were similar in NES patients, but dissimilar in TLE patients. On the other hand, this dissimilarity in ERPs disappeared when comparing z scores for the preictal and postictal recordings in NES patients with z scores for the preictal and interictal recordings in TLE patients. This further supports the notion that identical waveforms during preictal and postictal recordings in NES patients reflect nonepileptic seizure activity. The current findings suggest that postictal ERP recordings are useful in the diagnosis of NES and differentiate TLE from NES. (C) 2004 Elsevier Inc. All rights reserved.