A self-administered screening instrument for psychogenic nonepileptic seizures

• Published in: Neurology Vol. 72; no. 19; pp. 1646 - 1652
• Main Authors: SYED, T. U, AROZULLAH, A. M, LOPARO, K. L, JAMASEBI, R, SUCIU, G. P, GRIFFIN, C, MANI, R, SYED, I, LODDENKEMPER, T, ALEXOPOULOS, A. V
• Format: Journal Article
• Language: English
• Published: Hagerstown, MD Lippincott Williams & Wilkins 12-05-2009
• Subjects: Adult; Bayes Theorem; Biological and medical sciences; Brain - physiopathology; Diagnosis, Differential; Electroencephalography - methods; Epilepsy - diagnosis; Epilepsy - epidemiology; Epilepsy - physiopathology; Epilepsy - psychology; Female; Headache. Facial pains. Syncopes. Epilepsia. Intracranial hypertension. Brain oedema. Cerebral palsy; Humans; Logistic Models; Mass Screening - methods; Medical sciences; Nervous system (semeiology, syndromes); Neural Networks (Computer); Neurology; Neuropsychological Tests; Predictive Value of Tests; Prospective Studies; Reproducibility of Results; Self-Assessment; Surveys and Questionnaires; Videotape Recording - methods; Instrument; Psychogenic nonepileptic seizure; Nervous system diseases; Medical screening
• ISSN: 0028-3878; 1526-632X
• DOI: 10.1212/WNL.0b013e3181a55ef7

Delay in distinguishing psychogenic nonepileptic seizures (PNES) from epilepsy may result in significant health and economic burdens. Screening tools are needed to facilitate earlier identification of patients with PNES, thereby maximizing cost-effective use of video electroencephalography (VEEG), the expensive gold standard for differentiating PNES from epilepsy. We developed and prospectively validated a self-administered PNES screening questionnaire using variables known to distinguish PNES from epilepsy patients.
Adults referred for inpatient VEEG monitoring at two epilepsy centers were prospectively invited to complete a preliminary 209-item questionnaire assessing demographic, clinical, seizure-related, and psychosocial information that appeared in the literature as potentially useful indicators of PNES. A hybrid neural-bayesian classifier was trained to predict PNES using a sample at one center, and was prospectively validated on a separate set of naive patients from both centers.
Of 211 enrolled subjects from the training center, 181 met the study criteria for either PNES (n = 48, 27%), epilepsy (n = 116, 64%), or coexisting PNES and epilepsy (n = 17, 9%). Variable reduction procedures identified 53 questionnaire items that were necessary to accurately predict PNES diagnosis. The hybrid classifier predicted PNES diagnosis with 94% sensitivity and 83% specificity at the training center, and 85% sensitivity and 85% specificity at the second center (n = 46; 17 PNES, 26 epilepsy, 3 with coexisting PNES and epilepsy).
We developed and prospectively validated a self-administered psychogenic nonepileptic seizure screening questionnaire that could hasten referral for video electroencephalography and reduce the health and economic burdens from delayed diagnosis or misdiagnosis.