GOLD STANDARDS AND LIFE DECISIONS.

GOLD STANDARDS AND LIFE DECISIONS. This article by Gina Kolata about Avastin in the New York Times interested me for a couple reasons. First, the choice that is being questioned (to use specific numbers) seems to be whether it is worth while to spend $50,000 to extend a cancer patient’s life an average of four months. It seems to me the answer is obviously yes, especially given the value attached to an additional year of life in other contexts (such as suits for wrongful death). The article seems to reflect the greater weight that is usually given to preventing deaths from side effects than to extending lives. Second, the article illustrates the contrast between the “gold standard” model of statistical inquiry, and the approach that is taken when the inquiry is view in terms of decisions involving real people. The “gold standard” model is focused on the slow accumulation of knowledge. Usually, although the assumption is unstated, the “gold standard” model assumes that all tumors are the same, and information is sought on the average performance of a drug in treating an average tumor. However, it is generally thought that “every tumor is different.” How can information be developed about which cancer drugs are effective against certain kinds of tumors, but not all tumors? And what happens to patients while “gold standard” tests are being done?