Abstracts

Valid screening methods to identify people with a history of seizures or epilepsy are crucial for epidemiologic studies. They are also essential to identify affected family members in genetic studies. Screening relies on careful history-taking, but it is not clear what questions should be asked, or how accurately they identify affected individuals., We designed an 8-question screening interview to identify people with a history of seizures. The first question asks, [quot]Have you ever had, or has anyone ever told you you had, epilepsy or a seizure disorder?[quot] If this question is answered [quot]no,[quot] it is followed by a second question that asks, [quot]Have you ever had a seizure, convulsion, fit or spell under any circumstances?[quot] This is followed by six additional questions that screen for symptoms possibly related to seizures (e.g., uncontrolled body movements, unexplained change in mental state, etc.). Subjects who screen positive are given a detailed diagnostic interview to confirm seizure occurrence and obtain additional clinical details.
The interview was administered by telephone to 148 subjects with medical record documented incidence of unprovoked seizures from 1935-1994 while residing in Rochester, Minnesota, and 168 age- and sex-matched controls. Interviewers were blinded to case-control status., Sensitivity was defined as the proportion who screened positive ([quot]yes[quot] or [quot]possible[quot] to any screening question) among those with either epilepsy (N=124) or an isolated unprovoked seizure (N=24); specificity was defined as the proportion who screened negative among the controls. Sensitivity was 86% for epilepsy and 75% for isolated unprovoked seizures. Specificity was 92%. Sensitivity was significantly higher for subjects with incidence in 1955 or later than for those with earlier incidence (epilepsy: 89% vs. 71%, isolated unprovoked seizure: 82% vs 0% [0/2]). Among subjects with epilepsy who screened positive, 87% answered [quot]yes[quot] or [quot]possible[quot] to the first question (asking about epilepsy or a seizure disorder), 9% to the second question (asking about seizures under any circumstances), and 4% to another question. Among subjects with isolated unprovoked seizures who screened positive, 56% answered [quot]yes[quot] or [quot]possible[quot] to the first question and 44% to the second question., This screening instrument is highly effective (89% sensitivity) for identifying people with onset of epilepsy up to 50 years before interview. Subjects are more likely to be missed if they had onset of epilepsy even earlier, or had an isolated unprovoked seizure rather than epilepsy. Further analyses should address the relation of sensitivity to clinical epilepsy features, so that sensitivity can be improved by targeting new questions to particular subgroups., (Supported by NIH R01 NS043472.)