The impact of cancer biology, lead time bias, and length bias in the debate about cancer screening tests.

Tumor marker screening by insurers is one of the most significant and controversial developments in insurance testing. The debate is ongoing. Key dimensions of the issue have been recently described.1,2 We do not wish to reiterate any of those dimensions, nor do we wish to comment favorably or unfavorably on tumor marker screening by insurers. Rather, we believe there are some important and interesting theoretical considerations for the evaluation of cancer screening tests in clinical or insurance medicine. This article illustrates the biologic variability of two malignancies which may not be widely appreciated. New work has shown dichotomous populations within groups of patients known to have monoclonal gammopathy of unknown significance (MGUS) and childhood neuroblastoma. In addition, this article elucidates the role of lead time bias and length bias in the interpretation of screening tests. This has significance for all screening tests, not just cancer screening tests or tumor marker screening tests.