Correcting a Bias?

Abstract

Horwitz and Feinstein, in an article in this issue of the Journal, suggest that a selection bias explains the strong association between exogenous estrogens and endometrial cancer reported by many investigators. To correct this bias, they propose a method that was abandoned by others precisely because it introduced a large selection bias into a research design in which selection bias was not inherently an important feature.Horwitz and Feinstein believe that much of the strong association observed between estrogen use and endometrial cancer results from increased medical attention and therefore greater detection of the cancer among estrogen users, in . . .