The Epidemiologic Necropsy

To the Editor.—  The suggestion in the article by McFarlane et al¹ that the increased occurrence of lung cancer in the United States during the past 30 years is due to improved diagnosis of disease is not supported by the data presented and is based on faulty logic. This article confirms what numerous other published studies have shown previously: many cancers go undetected. While the existence of an "undetected reservoir" leaves open the possibility that changes in the occurrence rates of a given cancer site could be due to improved detection methods, this is not likely to be the case for lung cancer for the following reasons. Technological advances in the detection of primary lung cancer have not been as extensive as purported in the article, and none have been widely applied in screening the population. Certainly, advances in detection methods for cancers such as those of the colon