Objective: Seizure semiology is important for classifying patients' epilepsy. Physicians usually get most of the seizure information from observers though there have been few reports on the reliability of the observers' description. This study aims at determining the reliability of observers' description of the semiology.
Methods: We included 92 patients who had their habitual seizures recorded during video-EEG monitoring. We compared the semiology described by the observers with that recorded on the videotape, and reviewed which characteristics of the observers affected the reliability of their reported data.
Results: The classification of seizures and the individual components of the semiology based only on the observer-description was somewhat discordant compared with the findings from the videotape (correct classification, 85%). The descriptions of some ictal behaviors such as oroalimentary automatism, tonic/dystonic limb posturing, and head versions were relatively accurate, but those of motionless staring and hand automatism were less accurate. The specified directions by the observers were relatively correct. The accuracy of the description was related to the educational level of the observers.
Conclusions: Much of the information described by well-educated observers is reliable. However, every physician should keep in mind the limitations of this information and use this information cautiously.