Download and explore the data featured in this webinar:
- Non-Inferiority t-test for Two Means.nqt
- Inferiority Tests for the Difference of Two Proportions.nqt
- Non-Inferiority t-test for Two Means with Unequal n's.nqt
- Two One-Sided Equivalence Tests for Two Group Design.nqt
Power for Complex Hypotheses
Sample Size for Non-Inferiority, Superiority by a Margin and Equivalence Testing
In this tutorial, we have explored complex hypotheses as an alternative to inequality testing. Inequality testing is used to determine if a clinically meaningful difference between a new treatment and a placebo. If however, we want to compare a new treatment to a standard therapy or existing drug we need to test a different hypothesis.
You will learn about:
-
Power for Complex Hypotheses
-
Important design considerations for complex hypotheses
-
Sample size determination for complex hypotheses
-
Non-inferiority design for means
-
Non-inferiority design for proportions
-
Superiority by a margin design for means
-
Equivalence design for means
What is power for complex hypotheses?
Complex hypotheses can be used to test if a new treatment is non-inferior, superior, or equivalent to an existing one. These type of tests are common in the areas of generic drug development, medical devices, and design and sample size for vaccine trials. For example, if a proposed device or treatment were less invasive than the standard treatment then testing for non-inferiority would be an appropriate route to improve patients’ treatment choices.
A common thread across all of these hypothesis tests is the requirement of a margin, in order to establish non-inferiority/superiority/equivalence. In each of these tests the margin will have a different definition, but it is always crucial to select an appropriate one. Margin selection is a much discussed topic by researchers and regulatory agencies, and is a crucial part of any trial design.
These complex hypotheses can be constructed for a wide range of endpoints, for example, continuous, binomial, or survival. Each of these endpoints will have various statistical choices and design choices.
In this tutorial, we have reviewed these considerations and demonstrated worked examples of sample size determination for a selection of these endpoints.
nQuery helps make your clinical trials faster, less costly and more successful with tools for sample size calculation - refining Frequentist, Bayesian & Adaptive designs.
Start your unlimited 14-day trial to help design efficient, informative and ethical adaptive trials.
