To the Editor: In their editorial1 criticizing the Horwitz and Feinstein paper on endometrial cancer,2 Hutchison and Rothman claim that a sample of patients with endometrial cancer selected from a gynecologic-cancer treatment center will be unbiased for estrogen usage. Their argument is based on the premise that cancer is a progressive disease that inevitably leads to diagnosis. Thus, although estrogen-induced bleeding may cause earlier treatment, the patient with cancer who took estrogen will be no more likely to be selected as a case than the one who did not. The premise of inevitable diagnosis requires, however, that the rate of.
ASJC Scopus subject areas