Rationale:
Psychogenic non-epileptic seizures (PNES) are often misclassified as epileptic seizures (ES). The aim of this study was to create an easy to use but comprehensive screening tool that can guide further evaluation of patients presenting with this diagnostic dilemma.
Method:
Demographic, clinical and diagnostic data were collected on patients investigated with video-EEG monitoring for clarification of their diagnosis. Patients were classified as having PNES vs ES. Univariate associations were performed between the two cohorts and those parameters that were proven to be statistically significant were fitted in a multivariate logistic regression model to identify predictors of PNES. Results43 patients with PNES and 165 patients with ES were recruited. In the univariate analysis, predisposing factors for PNES, semiology suggestive of PNES, lack of antiepileptic polypharmacy, increased number of comorbidities, increased seizure frequency, shorter disease duration, normal electroencephalogram (EEG) and normal brain imaging were identified as potential predictors of PNES. In the multivariate analysis, presence of predisposing factors for PNES (β=1.68, p=0.05), semiology suggestive of PNES (β=3.26, p=0.0002), increased seizure frequency (β=0.06, p=0.005), shorter disease duration (β=-0.07, p=0.03) and normal EEG (β=1.94, p=0.04) sustained as important predictors of PNES. The performance statistics of the model were excellent (area under the receiver operating characteristic curve=0.96).
Conclusion:
We propose the D-PNES mnemonic (Duration short, Predisposing factors for PNES, Numerous seizures, EEG normal, Semiology suggestive of PNES) as a promising screening tool to Differentiate PNES from ES. Future validation studies are needed to evaluate its utility in clinical practice.
Funding:
:No funding
FIGURES
Figure 1