Abstract
The reported incidence of primary malignant brain tumors in children increased by 35% during the period from 1973 through 1994 (1). In this issue of the Journal, Smith et al. (2) analyze the incidence data to determine if this change represents a true increase in the disease frequency or an artifact of increased detection by magnetic resonance imaging (MRI). Similar controversies have arisen previously with regard to primary malignant brain tumors in the elderly and computed tomography (CT) (3), other neoplasms, including ductal carcinoma in situ of the breast, and mammography (4), and non-neoplastic diseases, including abdominal aortic aneurysms, and ultrasonography (5). In light of these past and present controversies, let us consider how a new diagnostic test can affect the reported incidence of disease when its real incidence is stable. Next, we can see how well this general phenomenon can explain the particular case of MRI and primary malignant brain tumors in children.