Data collection strategies in genetic epidemiology: The Epilepsy Family Study of Columbia University.

A large-scale study of genetic influences on seizure disorders is described here as a primer of tested methods for collection of family history data. 1957 adult probands with epilepsy were ascertained from voluntary organizations. Personal and family history data were obtained from probands in semistructured telephone interviews. To increase sensitivity, an independent family history was obtained from a second family informant in a similar interview. To increase specificity and diagnostic detail, family members reported to be affected were interviewed, and medical records of probands and affected relatives were collected. Participation rates for probands were 84-90%. Interviews were completed with second informants in 67% of families, and with 51% of eligible affected relatives. The main reasons for non-interview were lack of permission from probands and difficulties in locating relatives. Although 90% of probands gave verbal permission for medical record review, only 75% of these signed and returned consent forms for this purpose. Physicians returned 87% of the records requested. The resulting proportion of probands with medical records was 59%. These findings illustrate the complexity involved in assembling useful databases in genetic epidemiology.