Family History of Cancer in Patients With Glioma: a Validation Study of Accuracy

Extract

Family studies of cancer rely on accurate reporting of cancer by relatives. The family history data often form the basis for inferences about cancer risk factors and patterns of susceptibility. In this study, we evaluated the accuracy of cancer reports by comparing information from interviews with data contained in medical records and death certificates.

The participants were first-degree and selected second-degree relatives of patients with glioma who were registered at The University of Texas M. D. Anderson Cancer Center between June 1992 and June 1995. All probands were younger than 65 years at diagnosis and resided in the United States or Canada. We used a sequential sampling scheme where second-degree relatives were included only if a first-degree relative had cancer. We contacted relatives of probands directly when the information on the cancer site, date of diagnosis, or hospital where the relative was treated was uncertain. We asked each living relative with invasive cancer (excluding nonmelanoma skin), or if deceased, their closest next of kin, to sign a medical record release. On average, besides the initial informant, we contacted three relatives of each proband to obtain complete information to document the reported cancer. We attempted to document all cancers reported in relatives diagnosed under age 75 years who were treated at a hospital in the United States or Canada. We requested death certificates for deceased relatives reported to have cancer for whom medical records were not available.