Abstract
There is much believed, but nothing sacred, about a p-value of 0.05. This number, 0.05, is potentially the most dangerous in all of drug development. For example, the finding of a p-value <0.05 in Phase II may directly result in a costly Phase III effort for a given drug. This p-value, however, is also a measure of the probability that the decision to proceed is wrong. Moreover, when a portfolio is considered, p-values measure the relative risk of misallocating resources among projects. If the costs of Phase III differ significantly between projects, it may be irrational to rely upon the typical threshold p-value of 0.05. As a result, required p-values should vary inversely with projected future costs. The principles that underlie this assertion lend themselves to a general model for rational decision making in drug development, particularly where costs dominate and risk is adjustable.
Get full access to this article
View all access options for this article.