|
Abstract:
|
Equivalence design is often used to demonstrate the similarity of a biosimilar to some marketed product. At design stage, it is often challenging to correctly estimate the range of the nuisance parameters which affect the sample size calculation of the study. If the estimate is wrong, the study could face the risk of being underpowered. Blinded sample size re-estimation can be used to mitigate such risk. We will show how to perform blinded sample size re-estimation in equivalence design through a real case study. We will also highlight the pros and cons of such method comparing to other strategies such as group sequential design and unblinded sample size re-estimation design.
|
