Instead of testing a null hypothesis of no adverse effect, the presence of adverse effect can also be assessed with confidence intervals, and confirmed if a priori defined boundaries meet those confidence intervals. Boundaries commonly set are those of.
This chapter assesses how testing works and why and when it is particularly relevant.
If, in a controlled trial a new treatment is compared to a standard treatment instead of placebo, there is a big risk of finding small differences, and testing against prior boundaries is particularly helpful.
Null hypothesis Adverse effect Confidence intervals Priori defined boundaries Equivalencce Inferiority Superiority Comparison versus placebo
This is a preview of subscription content, log in to check access.