Rating Scale for Psychogenic Non-epileptic Seizures: Scale Development and Clinimetric Testing

Our aim was to develop a clinimetric scale evaluating motor phenomena, associated features, and severity of psychogenic nonepileptic seizures (PNES). Sixty video/EEG-recorded PNES induced by suggestion maneuvers were evaluated. We examined the relationship between results from this scale and results from the Clinical Global Impression (CGI) scale to validate this technique. Interrater reliabilities of the PNES scale for three raters were analyzed using the AC1 statistic, Kendall's coefficient of concordance (KCC), and intraclass correlation coefficients (ICCs). The relationship between the CGI and PNES scales was evaluated with Spearman correlations. The AC1 statistic demonstrated good interrater reliability for each phenomenon analyzed (tremor/oscillation, tonic; clonic/jerking, hypermotor/agitation, atonic/akinetic, automatisms, associated features). KCC and the ICC showed moderate interrater agreement for phenomenology, associated phenomena, and total PNES scores. Spearman's correlation of mean CGI score with mean total PNES score was 0.69 (P<0.001). The scale described here accurately evaluates the phenomenology of PNES and could be used to assess and compare subgroups of patients with PNES.

Rationale: Psychogenic non-epileptic seizures (PNES) are a diagnosticand therapeutic challenge for the clinician, since they can mimic the fullspectrum of epileptic seizures. To date, no scales are available to accurately assess the PNES. A validated scale assessing the psychogenicmotor disorders has recently been reported (1). We adapted this scale todevelop and test a clinimetric scale for PNES Methods: Patients were 46 woman and 14 men, ranging in age 17 to 85 years (38.9 € 18.7). Eight out of 60 patients had both epileptic seizuresand PNES. Eight PNES occurred spontaneously, 50 were induced by different suggestion manoeuvres. Patients were recorded according to astandardized video-EEG protocol. Three epileptologists independentlyrated these videotapes. The PNES scale rated the motor phenomenainvolving different body regions, the associated features (sphyntericincontinence, morsus, drooling, eye closure, hyperventilation, lament orcrying) and the responsiveness. The relation between the PNES scale andthe Clinical Global Impression Scale [CGI] (2) was examined. Interraterreliabilities of both PNES and CGI scales were then assessed (Kendall'scoefficient of concordance, Spearman correlation, and intraclass correlationcoefficients).Results: Interrater agreements for presence or absence of each phenomenon ranged from 58.3 to 100%. Kendall's concordance coefficients were 0.704 and 0.666 for total phenomenology and PNES scores, respectively. ICCs values for total phenomenology (0.536) and PNES scores(0.452) showed a significant interrater reliability (p<0.001). Spearman correlations between total PNES scores and CGI ratings for the three raters ranged from 0.311 to 0.606 (p<0.02)Conclusions: This new scale is proposed as a user-friendly and clinically sound instrument for the quantitative assessment of PNES.