Abstracts

Rationale: To study the characteristic behaviors and patterns associated with psychogenic nonepileptic seizures (PNES) and determine if common features can be identified and grouped accordingly.Methods: We studied epileptic (n=129) and non-epileptic (n=168) events recorded during video-EEG monitoring of 36 patients with confirmed pure-PNES and 21 patients with confirmed epileptic seizures. Event duration was recorded by noting time of onset and offset of irregular movement or behavior. The 27 most common behaviors exhibited by the patients were coded and analyzed with relation to event duration and fluctuation of intensity. Results: Compared to epileptic seizures, non-epileptic events were significantly longer in duration. More importantly, the higher variance of the length of events predicted reliably the type of event (non-epileptic events had a significantly higher variance). Furthermore, the most common feature observed across all non-epileptic events was the fluctuation of the intensity of the behavior exhibited during the events. Conclusions: Our findings suggest that the combination of duration and intensity-fluctuation of the behavior reliably predicts with high confidence if a patient's events are epileptic or non-epileptic in etiology.