Tests for Superiority, Equivalence, and Non-inferiority

Question 1

Superiority Trial

Answer 1

the goal is to show that the response to the
investigational product is superior to a comparative
agent (active or placebo control)

Question 2

What is the null in a Superiority Trial?

Answer 2

H0: There is no difference between the means of 2 arms (or the HR=1)

If the Confidence interval of 95% includes 0, then we can’t reject the null

If it does not include 0 (cross 0) we reject the null -> statistical significant

Question 3

What is the goal of an Equivalence Test?

Answer 3

to see if a new intervention (drug) is neither worse nor better than an established intervention

Question 4

Equivalence Test

Answer 4

An investigator would determine a lower and upper margin (margins are depicted by (-) delta and (+) delta) that would indicate that if the findings of the new drug would be within that margin, it would be equivalent =

Equivalence Margin

Question 5

The null hypothesis of Equivalence studies

Answer 5

H0: The new drug differs from standard drug by the value of the margins: (+) and (-) margin

It is the opposite of what we usually do in H0

-when rejecting the NULL (proving the 2 drugs are equivalent) the p-value will be less than 0.05
-> Even though the CI includes 0 -> bc we want to show that there is no difference

Question 6

What is the goal of a Non-inferior study?

Answer 6

To show that the drug is not worse than the drug it is compared to

-Select a margin on the inferior site

The H0 would be: The drug is worse than the comparison group by at least a margin
Mean 1 - Mean 2 > (or equal) to the margin

Question 7

When is a treatment considered inferior?

Answer 7

When the entire 95% CI of the treatment is outside of the non-inferior margin

Question 8

Can the findings of the new drug be superior?

Answer 8

Yes it can cross at the superior side and will still be considered non-inferior unless the CI doesn’t cross the non-inferior margin