ABSTRACT
It is important to evaluate from the evidence in the data whether an observed nding is real and can be reproduced in any similar or different relevant setting. In evaluating the strength of the evidence in the data, the KefauverHarris amendment of the Food and Drug Cosmetic Act of 1962 denes “substantial evidence” as “evidence consisting of adequate and well-controlled investigations.”1 According to Huque,2 the U.S. Food and Drug Administration (FDA) has interpreted this as “the need to conduct at least two adequate and well-controlled studies, each convincing on its own, as evidence of ef- cacy of a new treatment of a given disease.” There are conditions where data from a single adequate and well-controlled trial can be considered to constitute substantial evidence.3 An interpretation of this is data from a large, adequate, and well-controlled multicenter study with a sufciently small p-value that is internally consistent and clinically meaningful.2
