Differentiating epilepsy from psychogenic nonepileptic seizures using neuropsychological test data

Brad T Tyson; Shannon Baker; Matthew Greenacre; Katrina J Kent; Jonathan D Lichtenstein; Alana Sabelli; Laszlo A Erdodi

doi:10.1016/j.yebeh.2018.08.010

Differentiating epilepsy from psychogenic nonepileptic seizures using neuropsychological test data

Epilepsy Behav. 2018 Oct:87:39-45. doi: 10.1016/j.yebeh.2018.08.010. Epub 2018 Aug 29.

Authors

Brad T Tyson¹, Shannon Baker², Matthew Greenacre², Katrina J Kent², Jonathan D Lichtenstein³, Alana Sabelli⁴, Laszlo A Erdodi⁴

Affiliations

¹ Neuroscience Institute, EvergreenHealth Medical Center, Kirkland, WA, USA. Electronic address: bttyson@evergreenhealthcare.org.
² Schulich School of Medicine, Western University, London, ON, Canada.
³ Department of Psychiatry, Dartmouth-Hitchcock Medical Center, Lebanon, NH, USA.
⁴ Department of Psychology, University of Windsor, Windsor, ON, Canada.

PMID: 30172082
DOI: 10.1016/j.yebeh.2018.08.010

Abstract

Objective: Differentiating epileptic seizures (ES) from psychogenic nonepileptic seizures (PNES) represents a challenging differential diagnosis with important treatment implications. This study was designed to explore the utility of neuropsychological test scores in differentiating ES from PNES.

Method: Psychometric data from 72 patients with ES and 33 patients with PNES were compared on various tests of cognitive ability and performance validity. Individual measures that best discriminated the diagnoses were then entered as predictors in a logistic regression equation with group membership (ES vs. PNES) as the criterion.

Results: On most tests of cognitive ability, the PNES sample outperformed the ES sample (medium-large effect) and was less likely to fail the Reliable Digit Span. However, patients with PNES failed two embedded validity indicators at significantly higher rates (risk ratios (RR): 2.45-4.16). There were no group differences on the Test of Memory Malingering (TOMM). A logistic regression equation based on seven neuropsychological tests correctly classified 85.1% of patients. The cutoff with perfect specificity was associated with 0.47 sensitivity.

Conclusions: Consistent with previous research, the utility of psychometric methods of differential diagnosis is limited by the complex neurocognitive profiles associated with ES and PNES. Although individual measures might help differentiate ES from PNES, multivariate assessment models have superior discriminant power. The strongest psychometric evidence for PNES appears to be a consistent lack of impairment on tests sensitive to diffuse neurocognitive deficits such as processing speed, working memory, and verbal fluency. While video-electroencephalogram (EEG) monitoring is the gold standard of differential diagnosis, psychometric testing has the potential to enhance clinical decision-making, particularly in complex or unclear cases such as patients with nondiagnostic video-EEGs. Adopting a standardized, fixed neuropsychological battery at epilepsy centers would advance research on the differential diagnostic power of psychometric testing.

Keywords: Epilepsy; Multivariate assessment models; Neurocognitive profiles; Neuropsychology; Psychogenic nonepileptic seizures.

MeSH terms

Adult
Diagnosis, Differential
Epilepsy / diagnosis*
Female
Humans
Male
Middle Aged
Neuropsychological Tests / standards*
Reproducibility of Results
Retrospective Studies
Seizures / diagnosis*
Somatoform Disorders / diagnosis*