Validation of a Standardized Diagnostic Interview for Epilepsy

DESCRIPTION (provided by principal investigator): Standardized tools for clinical diagnosis and classification are essential for consistency in epidemiologic studies, and are especially important for epilepsy because diagnosis and classification are largely based on patient history. Despite the importance of this problem, however, few tools have been developed for standardized collection of clinical information in the epilepsies. In the late 1980's we created a telephone-administered semistructured Epilepsy Diagnostic Interview for clinical classification. The original instrument had good inter-rater reliability and validity for distinguishing between generalized and partial seizures. However, since these early validation studies, we have made major revisions to the instrument to refine our ability to characterize seizure symptoms and thereby improve classification. In the current study we propose to validate this revised instrument and make it available to the research community in a web-based computerized telephone interview format. The data for validation will be from a population-based investigation using Rochester Epidemiology Project resources, the Genetic Epidemiology of Seizure Disorders in Rochester (GESDR) study. GESDR included all Rochester residents born in 1920 or later with incidence of unprovoked seizures from 1935-1994. For each of these subjects, two expert epileptologists carried out a comprehensive review of all medical record information from onset of seizures to last residency in the Rochester area (including seizure descriptions, EEG, neuroimaging, treatment history, etc.) and rendered a final epilepsy diagnosis and classification based on current ILAE criteria. Independently of the medical record review, we administered our current diagnostic interview to each surviving subject who consented, and the same epileptologists reviewed the interview data and rendered a diagnosis and classification based on interview data alone. In the validation study we will assess agreement between these two diagnoses, using the medical record diagnosis as the gold standard. PUBLIC HEALTH RELEVANCE: This project will evaluate the accuracy of clinical classification of epilepsy based on information collected in a standardized telephone interview. It will also produce a web-based computer-assisted telephone interview version of the validated interview. This should facilitate the conduct of epidemiologic research on the causes and consequences of epilepsy at a significant savings of time and money.