—Mr Mantel is concerned about the misclassification issue. It has been argued that some women with lung cancer who were or have been smokers may have intentionally denied (because of guilt or social pressures) this fact. It is difficult to imagine that such a bias could operate in the context of an autopsy study. Furthermore, in this study, the outcome variable is semiquantitative and odds ratios cannot be interpreted as estimates of RR for lung cancer. Mantel does not point out any bias, nor does he consider any issue that we did not ourselves discuss in the "Comment" section of our article; he simply indicates that he disagrees with our conclusions.Coggins and Steichen worry about "uncontrolled confounding in such a small sample," but confounding does not depend on study size. With strange logic, they consider the fact that the mean EPPL score in four subjects with lung
Trichopoulos D, Katsouyanni K, Tomatis L, et al. Active and Passive Smoking and Pathological Indicators of Lung Cancer—A Report of Limited Value?-Reply. JAMA. 1993;270(14):1691. doi:10.1001/jama.1993.03510140049020
Customize your JAMA Network experience by selecting one or more topics from the list below.