The personality assessment inventory as a tool for diagnosing psychogenic nonepileptic seizures

Using 184 subjects with valid personality assessment interview (PAI) profiles and video-electroencephalography (VEEG)-confirmed diagnoses of epileptic seizures (ES; n = 109) or psychogenic nonepileptic seizures (PNES; n = 75), we present the diagnostic test performance of the PAI PNES Indicator and other PAI scales when used to differentiate PNES from ES. Subjects with PNES reported significantly higher somatic, conversion, depressed, anxious, and suicidal symptoms. As a diagnostic tool, the PNES Indicator does not add additional accuracy beyond the conversion subscale (SOM-C). The somatization (SOM-S) and physiological depression (DEP-P) subscales perform as well as the SOM-C subscale. The SOM-C scale (cut point ≥70) was 58.7% sensitive and 83.5% specific at diagnosing PNES. Assuming a 30% prevalence of PNES, the SOM-C scale has a positive predictive value (PPV) of 60.4% and negative predictive value (NPV) of 82.5%. Overall, the PAI SOM-C subscale does not appear more accurate than other psychometric tests used to differentiate PNES from ES.