Table 5.

Information required to estimate a sample size for equivalence, superiority and non-inferiority trials.

Outcome variable	Study design	Arguments	References
Continuous	Equivalence trial	The expected mean of the outcome variable in the treatment and control groups, the expected population standard deviation of the outcome variable, the equivalence limit (expressed as a proportion), the required study power, the ratio of the number of exposed subjects to the number of unexposed subjects.	(15–17).
Binary	Equivalence trial	The expected proportion of successes in the treatment and control groups, the equivalence limit (expressed as a proportion), the required study power, the ratio of the number of exposed subjects to the number of unexposed subjects.	(15–17).
Continuous	Superiority trial	The expected mean of the outcome variable in the treatment and control groups, the expected population standard deviation of the outcome variable, the equivalence limit (expressed as a proportion), the required study power, the ratio of the number of exposed subjects to the number of unexposed subjects.	(15).
Binary	Superiority trial	The expected proportion of successes in the treatment and control groups, the equivalence limit (expressed as a proportion), the required study power, the ratio of the number of exposed subjects to the number of unexposed subjects.	(15)
Continuous	Non-inferiority trial	The expected mean of the outcome variable in the treatment and control groups, the expected population standard deviation of the outcome variable, the equivalence limit (expressed as a proportion), the required study power, the ratio of the number of exposed subjects to the number of unexposed subjects.	(15–17)
Binary	Non-inferiority trial	The expected proportion of successes in the treatment and control groups, the equivalence limit (expressed as a proportion), the required study power, the ratio of the number of exposed subjects to the number of unexposed subjects.	(16, 17)