Records of 317 patients at least 48 years old who were diagnosed as having endometrial c...

Records of 317 patients at least 48 years old who were diagnosed as having endometrial carcinoma were obtained from two hospitals (Smith et al. 1975). Matched controls for each case were obtained from the two institutions; the controls had cervical cancer, ovarian cancer, or carcinoma of the vulva. Each control was matched by age at diagnosis (within four years) and year of diagnosis (within two years) to a corresponding case of endometrial carcinoma. This sort of design, called a retrospective case-control study, is frequently used in medical investigations where a randomized experiment is not possible. The following table gives the numbers of cases and controls who had taken estrogen for at least 6 mo prior to the diagnosis of cancer. Is there a significant relationship between estrogen use and endometrial cancer? Do you see any possible weak points in a retrospective case-control design?