Abstracts

Rationale: The analysis of seizure semiology is well established in presurgical epilepsy monitoring. Classifying seizure semiology from patient history may be more difficult, because the recollection by the patient and next of kin may be inconsistent and flawed by subjectivity. Possibly, though, seizure semiology could be an important piece of information to triage patients whether to undergo epilepsy monitoring or not, especially in less affluent societies. We therefore studied whether history taking of seizure semiology can be improved by structured interviews and by demonstrating seizure semiology through video examples. We grouped semiological signs into signs for focal epilepsy and a group of lateralizing and localizing signs. Methods: In a first step, we compared the seizure semiology of 208 consecutive patients as described in the patients' history and compared it to the seizure semiology recorded in the epilepsy monitoring unit of the University of Munich. In a second step we compared the yield of seizure semiology information of 72 patients, when taken from patients' records, after structured interviews and after showing seizure videos to patients and their next of kin at the Epilepsy Center of the University Del Desarollo in Santiago de Chile. Results: Of the 208 patients of the first part of the study, 138 (66.3%) could correctly report signs for focal epilepsy as compared to 54 patients (26.0%) that did not report any signs of focal epilepsy but had them recorded in the epilepsy monitoring unit. Six patients had generalized epilepsies and five of them had incorrectly reported focal semiology. Lateralizing signs could correctly be reported by 15.3% of patients. 7.4% reported lateralizing signs, but we could not reproduce them in the epilepsy monitoring unit and 44.1% of patients did not report lateralizing signs but we recorded them in epilepsy monitoring. Lastly, 26.2% of patients did not have lateralizing signs in either history or recording. In the second part of the study, only 6 patients out of 56 (10.7%) with focal epilepsy had lateralizing signs on record (4 with localizing signs). After a structured interview, these numbers increased to 31 (55.4%) for lateralizing signs and 28 (50.0%) for localizing signs. After showing exemplary videos to the patients, these numbers increased even more, to 49 patients (87.5) for lateralizing signs and 33 patients (58.9%) for localizing signs. Conclusions: Our study demonstrates that the analysis of seizure semiology from patient history may be incomplete, but that it may be improved by structured interviews and even more so by demonstrating video examples of typical semiology to patients. Taken together with other information, this may improve the triage of patients towards presurgical epilepsy monitoring.