Health Sciences Study Design: Non-Inferiority Testing

[Amy Ly]

Hello,
I am designing a health sciences study where I will compare the accuracy of 3 different diagnostic tests to subtype a tumor.
A cohort of tumors will be subjected to all 3 diagnostic tests. Each diagnostic test returns a categorical result (the tumor is categorized as type A, B, or C). All 3 diagnostic tests will be compared to a Gold Standard. I hypothesize that all 3 diagnostic tests will be roughly equivalent to each other and may be used interchangeably. Hence, I would consider this study as one of Non-Inferiority/Equivalence.

Would I use McNemar for comparing the accuracy of each diagnostic test (paired with Gold Standard)?
How would I calculate required Sample Size to demonstrate Non-Inferiority/Equivalence of the diagnostic tests with each other? I have looked at the Power/Sample Size analysis calculator in STATA; it asks for info on discordance. Since I hypothesize the tests will be equivalent, I would enter zero for discordance, but that returns an error.

Appreciate any guidance. Apologies in advance for simplistic questions.
Amy

[Amy Ly]

For a Sample Size calculation, would this make sense:
sampsi .8 .75, power(.8) onesided

Parameters: Test #1 will correctly identify 80% of type A tumors. Test #2 will be considered non-inferior if it identifies 75% of type A tumors correctly.