Clinical events in psychogenic non-epileptic seizures based on semiological seizure classification.

OBJECTIVES: None of the classifications of psychogenic non-epileptic seizures (PNES) have been widely accepted and used by physicians so far. In this study we aimed at classifying PNES on the basis of a modified version of semiological seizure classification (SSC). We also sought to assess the interrater reliability (IRR) of the PNES diagnosis based on SSC.

METHODS: We classified PNES into four types on the basis of our modification of SSC: pseudoaura, dialeptic, motor, and special (atonic, astatic, hypotonic) spells. Pseudoauras were not included in the statistical analysis. Ninety-one PNES attacks were observed during the 55 video-EEG sessions recorded for all patients. The interrater agreement was assessed by the kappa coefficient.

RESULTS: Twenty-nine women (78·3%) and eight men (21·6%) were surveyed, with a mean age of 28·4 ± 9·6 (range 16-54). The final diagnosis of PNES was established after a mean of 4·5 ± 2·3 years following the onset of PNES attacks in the patients. The mean seizure duration in the PNES was 241 seconds and 40·5% of our patients had PNES longer than 300 seconds. Motor and special PNES were the most common types observed by all the raters. The kappa values for each pair were as follows: Observers I-II 0·51 (p = 0·000), Observers I-III 0·47 (p = 0·000), and Observers II-III 0·73 (p = 0·000).

CONCLUSIONS: Interobserver agreement was moderate and substantial for three observers who classified PNES according to our modified SSC. The modified version of SSC could be used without difficulty in classifying PNES. Using SSC for PNES both shortens the period before diagnosis and eliminates the need to learn another new and acceptable classification for PNES.