The diagnostic distinction between epilepsy and psychogenic nonepileptic seizures (PNES) can be challenging. Previous studies have demonstrated that experts in conversation analysis can identify linguistic and interactional features in transcripts and recordings of interviews with patients that reliably distinguish between epilepsy and PNES. In this study, ten senior neurology trainees took part in a one-day intervention workshop about linguistic and interactional differences in the conversation behavior of patients with epilepsy and those with PNES. Participants were familiarized with a 12-item questionnaire designed to capture their conversational observations immediately after talking to a patient with seizures. After the intervention, 55 initial outpatient visits of patients referred to seizure clinics were video and audio recorded. All medical diagnoses were confirmed two years after initial presentation on the basis of a chart review (including MRI and EEG findings) by a fully trained epilepsy expert. Postvisit questionnaires relating to patients confirmed to have epilepsy (n=20) or PNES (n=13) were analyzed. Doctors' mean responses to 6 of the 12 questions about linguistic and interactional observations differed significantly between the groups with epilepsy and PNES. Receiver operating curve analysis showed that a summation scale based on items demonstrating significant between-group differences correctly classified 81.8% of patients as having epilepsy or PNES. This study shows that a brief Conversation Analytic teaching intervention can enable neurologists to identify linguistic and interactional features supporting the differentiation of epilepsy and PNES as they take their patients' history in routine seizure clinic consultations, potentially improving diagnostic accuracy.
Keywords: Conversation analysis; Doctor–patient communication; Epilepsy; Improving diagnostic accuracy; Intervention; Nonepileptic seizures.
Copyright © 2016 Elsevier Inc. All rights reserved.