) The selection of an appropriate control group for case-control studies is a concern in designing epidemiologic studies. Studies in which the cases are obtained from specialized cancer hospitals pose a particular problem, since the source population for these cases is poorly defined. In recent years, commercial services have developed databases that combine data from a number of sources to allow companies and charitable organizations to describe their customer or donor bases. One feature of these databases is the identification of """"""""lifestyle"""""""" clusters based on census and marketing data. These clusters are homogeneous with respect to residential and behavioral characteristics not readily ascertainable from other sources. We will study the feasibility of using one of these databases to identify a sampling frame from which to select controls. This study will be conducted in the context of an ongoing case-control study of ovarian cancer at Memorial Sloan-Kettering Cancer Center. Using the zip plus 4 codes of cases with ovarian cancer, TRW, an owner of a commercial database, will identify a sample of households in the New York metropolitan region with similar characteristics. We will contact women in these households by letter and telephone to recruit them into the control group. We will evaluate the response rate in comparison to other published studies that use community controls and the cost and timing involved in assembling this control group. The design will include evaluation of two approaches that might lead to an increased response rate. First, the sampling frame will be divided into two groups based on whether or not households have responded to appeals from health-related charities to see if those who have responded are more likely to participate. Second, half of each of these groups will receive, along with the introductory letter inviting them to participate, a promotional flyer designed to persuade them to participate. Results of this study will indicate whether commercial databases offer a means of identifying potential controls that is a useful alternative to using other community controls or hospital controls, and whether either of the two approaches is successful in increasing the response rate.
| Olson, S H; Carlson, M D A; Ostrer, H et al. (2004) Genetic variants in SOD2, MPO, and NQO1, and risk of ovarian cancer. Gynecol Oncol 93:615-20 |
| Harlap, Susan; Olson, Sara H; Curtin, John P et al. (2002) Epithelial ovarian carcinoma and fertility of parents. Epidemiology 13:59-65 |
| Harlap, Susan; Olson, Sara H; Barakat, Richard R et al. (2002) Diagnostic x-rays and risk of epithelial ovarian carcinoma in Jews. Ann Epidemiol 12:426-34 |
| Olson, S H; Mignone, L; Nakraseive, C et al. (2001) Symptoms of ovarian cancer. Obstet Gynecol 98:212-7 |
| Harlap, S; Olson, S; Akhmedkhanov, A et al. (2001) Epithelial ovarian carcinoma and European birthplace of grandparents. Gynecol Oncol 81:25-32 |
| Olson, S H; Mignone, L; Harlap, S (2000) Selection of control groups by using a commercial database and random digit dialing. Am J Epidemiol 152:585-92 |
