Comparison of self-reported and database-linked family history of cancer data in a case-control study

Abstract

The authors linked interview data drawn from Utah participants in the Diet, Activity, and Reproduction in Colon Cancer (DARCC) Study (1992-1995) to genealogic and cancer information contained in the Utah Population Database (UPDB). They evaluated the sensitivity of subjects' reports of familial cancers and measured the overall agreement between reported and database records with the kappa (kappa) statistic. They calculated odds ratios from logistic regression to compare the relative risk estimates that would result from use of either data set (or both data sets). Overall, 37.6% (331 of 881) of the Utah DARCC subjects were linked to the UPDB genealogy. High sensitivities were observed for subjects' reports of breast (83%), colorectal (73%), and prostate (70%) cancers, while ovarian (60%) and uterine (30%) cancers were not reported as well. Results for kappa were similar, with values of 0.63 for breast cancer and 0.56 for colorectal cancer. Although the observed kappa s of 0.36 and 0.25 for ovarian and uterine cancers, respectively, exceeded chance expectations, the agreement between subjects' reports and database records was unimpressive. No consistent difference was observed between cases and controls in the accuracy of self-reports. In general, higher sensitivities were observed among younger subjects than older subjects; females reported family histories of cancer only slightly better than males. A college education was not consistently associated with more accurate reporting of family history of cancer. These results indicate that subjects in a case-control study are able to report accurately family histories of several common kinds of cancer and that they can do so without observable recall bias. The accuracy of self-reports may not be adequate for reproductive tract cancers and cancers such as rectal cancer that are frequently confused with cancers of similar organs.