ABSTRACT. Early detection and differential diagnosis of psychogenic non-epileptic seizures (PNES) and epileptic seizures (ES) is a major clinical issue in comprehensive epilepsy centers. Using blind conditions with patients with PNES (N = 10) and ES (N = 31) before diagnosis, we tested the hypotheses that individuals with PNES would exhibit significantly greater dissociativity, hypnotizability, absorption, and history of early abuse than ES patients. Although PNES patients tended to show greater dissociative phenomena, only the last of our hypotheses was fully supported. Although absorption did not discriminate between the two diagnostic groups, it was significantly higher among those reporting early abuse. A logistic regression analysis using scores on dissociation, hypnotizability and absorption showed them to be poor predictors of diagnosis; however, other analyses revealed that female gender, reports of multiple trauma incidents lasting months or years, initial seizure onset in late teens or twenties, and daily seizure attacks significantly differentiated PNES from ES patients. Thus, demographic and seizure variables proved to be much better predictors of diagnosis Richard Litwin is in Private Practice, Cleveland, OH.

Background: Identification of clinical features that might distinguish psychogenic nonepileptic seizures (PNES) from epileptic seizures (ES) is of value for diagnosis, management, and understanding of both conditions. Previous studies have shown that patients' descriptions of their seizures reflect differences in content and delivery. We aimed to compare verbal descriptions of PNES and ES using a mixed-methods approach. Methods: We analyzed data from semi-structured interviews in which patients with video-electroencephalography (EEG)-confirmed ES (n = 30) or PNES (n = 10) described their seizures. Two masked raters independently coded the transcripts for relevant psychological categories and discrepancies that were noted and resolved. Additional analyses were conducted using the Linguistic Inquiry and Word Count system. The identified phenomena were descriptively compared, and inferential analyses assessed group differences in frequencies. A logistic regression analysis examined the predictive power of the most distinctive phenomena for diagnosis. Results: As compared with ES, PNES reported longer seizures, more preseizure negative emotions (e.g., fear), anxiety symptoms (e.g., arousal, hyperventilation), altered vision/olfaction, and automatic behaviors. During seizures, PNES reported more fear, altered breathing, and dissociative phenomena (depersonalization, impaired time perception). Epileptic seizures reported more self-injurious behavior. Postseizure, PNES reported more fear and weeping and ES more amnesia and aches. The predictive power when including these variables was 97.5%. None of the single predictor variables was significant. The few but consistent linguistic differences related to the use of some pronouns and references to family. Conclusions: Although no single clinical feature definitively distinguishes PNES from ES, several features may be suggestive of a PNES diagnosis, including longer duration, negative emotion (i.e., fear) throughout the events, preseizure anxiety, ictal dissociation, and postseizure weeping. Fewer reports of ictal self-injury and postseizure amnesia and aches may also indicate the possibility of PNES.