Résumé
Objective To find out whether conversation analysis helps to differentiate psychogenic nonepileptic seizure (PNES) from epileptic seizure in Chinese patients.Methods Twelve unselected patients from Peking Union Medical College Hospital during 2014 to 2016 with diagnostic uncertainty were included.Interactions following standard protocol were carried out.A linguist blinded to all medical data and a neurologist studied videos and transcripts of the interactions.Using a diagnostic scoring aid which includes 17 conversation features summarized from previous researches, they attempted to predict the medical diagnosis of those patients independently.Results Accurate diagnosis was predicted in 10/12 patients by both raters.Average scores of patients with epileptic seizures were 8.00 (linguist) and 6.75 (neurologist), while average scores of paitents with PNES were-5.75 (linguist) and-7.88 (neurologist).Both raters agreed on most individual items (81.86%, 167/204).To demonstrate different features between these two groups, a case comparison was made between one patient with frontal lobe epilepsy and one patient with PNES.Conclusion In Chinese patients, conversation analysis can help differentiate between epileptic seizure and PNES.