Family psychiatric screening instruments for epidemiologic studies: pilot testing and validation

Family history, a risk factor for psychiatric disorders, is infrequently assessed in epidemiologic studies due to time and cost constraints. We designed a brief computer-scorable instrument, the Family History Screen for Epidemiologic Studies (FHE), which collects a pedigree and screens for 15 DSM-III diagnoses in an informant and in his family members. The FHE was administered to one informant in 77 families in which we had collected pedigrees, interviewed 77 informants and 239 relatives using the Lifetime Anxiety version of the schedule for Affective Disorders and Schizophrenia or the Epidemiologic version of the Schedule for Affective Disorders and Schizophrenia for School-Aged Children, and performed best-estimate diagnoses. We evaluated the accuracy with which the FHE predicted best-estimate diagnoses. For adults reporting on themselves, the FHE demonstrated high levels of sensitivity and specificity for depression (67.4, 75.0) and panic (92.5, 89.2), and low sensitivity and high specificity for substance abuse (33.3, 93.6). For informants reporting on adult relatives, sensitivity was low and specificity was high for depression (35.2, 84.9), panic (20.0, 91.7), and substance abuse (42.1, 93.4). For informants reporting on children, perhaps due to lower prevalence, sensitivity and specificity were poor. The FHE is a good screen for psychiatric disorders in adult informants, but it is not useful for family history. It may be useful in primary care medical settings as a screen for psychiatric history.