multiplicity

Question 1

what is multiplicity?

Answer 1

multiple testing in a trial. includes:

• Multiple primary endpoints (or endpoints of
  composite endpoints)
• Multiple treatment groups
• Multiple analyses (e.g. repeated measures)
• Subgroup analysis
• Study design (e.g. adaptive design)

Question 2

why do we need to adjust for multiple testing in a trial?

Answer 2

because it increases risk of type 1 error

Question 3

ways to control for multiple comparisons

Answer 3

> contorlling for familywise error rate. this ist he overall probability of at least one false positive conclusion anywhere within a defined set of trial hypothesis

> adjusting p values
- single stem methods: bonferroni, false discovery rate
- stepwise methods: holm, hochberg

> Hierarchical test procedures
- Two or more primary variables ranked according to
clinical relevance. No formal adjustment is necessary.
However, no confirmatory claims can be based on
variables that have a rank lower than or equal to that
variable whose null hypothesis was the first that could
not be rejected. EMEA

> closed test procedure