CONTROL SELECTION METHODS FOR CHILDHOOD CANCER STUDIES

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): The investigators propose to determine the feasibility and validity of using birth certificate controls in studies of childhood cancer. Most U.S. studies of childhood cancer have used ROD controls, but concerns about bias have increased as response rates have declined. A major drawback of ROD controls is that possible selection bias cannot be assessed, as no information is available on non-respondents. Controls identified through birth certificates are an appealing alternative on theoretical and practical grounds; they come from the same base population as the cases and a substantial amount of information is available on non-participants. However, the feasibility and validity have not been evaluated comprehensively or nationwide. Birth certificates will be randomly selected in three age groups and four race/ethnicity groups, and will be distributed among five geographic regions according to the proportion of the US population of the race/ethnicity group residing in that region. The investigators will locate, contact, and interview 450 mothers in each race/ethnicity group, evenly divided among the age groups. The nationwide sample of 1,800 will consist of 600 in each age group and 450 in each race/ethnicity group. The investigators will evaluate response rates overall and by age group, and race/ethnicity. Validity will be assessed by comparing the interviewed controls to all U.S. births using data from birth certificates. The proposal is timed to coincide with the continuing development of the Childhood Cancer Research Network (CCRN), which will provide large, nearly population-based series of cases. The results of the proposed study will provide much needed data to inform the choice of control group. With the CCRN and the results of the proposed study, epidemiologists will be able to design and execute state-of-the-art studies of childhood cancer. The results of the proposed research will also apply to studies of autism, asthma, and other pediatric conditions. Relevance: In a case-control study of childhood cancer, cases, who are children with a specific type of cancer, are compared to controls, children without cancer. The proposed study will evaluate whether controls identified through birth certificates are a feasible and valid control group. A feasible and valid control group is needed to ensure that studies of childhood cancer are scientifically valid and therefore, will provide results that can eventually lead to prevention of these cancers.