To the Editor.—In their article entitled "An Ecogenetic Hypothesis for Lung Cancer in Women,"1 Horwitz et al conclude that a "strong ecogenetic (defined as environmental genetic interactive) relationship is present in women among family history, cigarette smoking, and the risk of lung cancer."
Their data are not sufficient to support this claim. First, their attempts to identify and minimize biases in this chart review study do not eliminate the fact that the accuracy of family data from charts is inherently poor. The authors attempted to circumvent this problem by using control charts that had family history information. However, it is not clear that the control charts selected contained complete histories. Both doctors and patients tend to differentially report family information relevant to the patient's diagnosis. Thus, control patients with heart disease might have been asked about family members with heart problems, diabetes, and hypertension, but might not have