Concurrent administration of the MMPI-2 and PAI in a sample of patients with epileptic or non-epileptic seizures: Implications for an inpatient epilepsy monitoring unit

The Minnesota Multiphasic Personality Inventory-second edition (MMPI-2) and the Personality Assessment Inventory (PAI) are commonly used in the epilepsy monitoring unit (EMU) to evaluate personality characteristics and mood-related symptoms in those individuals being evaluated for epileptic seizures (ES) or psychogenic non‐epileptic seizures (PNES). A direct comparison of these measures through concurrent administration to the same group has not been carried out. Both measures were administered to 40 patients (17 ES and 23 PNES). Logistic regression suggested the optimal predictive model for EMU discharge diagnosis included subscales from each measure, which outperformed either measure separately. Combining the conversion (SOM-C) and health concerns (SOM-H) subscales from the PAI and the hysteria subscale (Hy) from the MMPI-2 resulted in 85% overall classification accuracy, 86.7% sensitivity, and 82.4% specificity. Variability in the literature regarding the predictive utility of these measures may stem from the possibility that they measure different aspects of PNES.