Consistency of self-reported first-degree family history of cancer in a population-based study.

The aim of this study was to assess the prevalence and consistency of self-reported family history of cancer among first-degree relatives (FDR) in a population-based study. Women at primary care units (PCU) were submitted to a questionnaire about cancer family history. Consistency of the report was determined by comparing self-reported history at the PCU to data from subsequent genetic evaluations and/or cancer confirmatory documents. Consistency in relation to degree of education, reported tumor type and reported age at cancer diagnosis in FDR was assessed. In 8,881 women interviewed, the prevalence of cancer in an FDR was 25.14% (CI 95%: 24.14; 25.94). Mean age was 40.29 years and most (70.26%) had < or = 8 years of education. There was a good agreement of self-reported cancer history at the PCU and in subsequent genetic evaluations [Kappa coefficient = 0.76 (P < 0.05)]. Inconsistencies were not related to low literacy (chi (2) = 2.027; P = 0.363). Consistency of the reported information for cancer status, cancer type and age of onset was 92.59%, 85.33% and 92.64%, respectively. The prevalence of cancer history in an FDR was similar to previous reports in other populations. Consistency and reliability of the self-reported information was high, regardless of educational level.