In regards to power I’ve always wondered this:

In regards to power I’ve always wondered this:

Why would the FDA authorize a trial design that at full enrollment isn’t considered “powerful” enough at a resulting p-value of .05 (industry standard for p-value “success”) when both the company and the FDA are involved from the get go? I’d get falling short of power at a 50% or 75% interim analysis and requiring a stronger p-value result to consider it a success worth stopping early, but what’s the point of carefully designing a trial that does exactly what it’s supposed to do and “fails” to meet power?

(I’m guessing the answer lies in shift of standard of care, or something similar, but if results/repeatability are cool at 95% but not 94% aren’t statisticians and the FDA just being nitpicking dicks at that point?)

((Now I’m guessing you’ll say that’s exactly why power is important and where an EUA off of clinical significance rather than statistical significance comes in to play))

(((This is where I begin to think I answered my own question but because I’m not sure I still wanted to ask)))

((((Thank you in advance))))

(((((#DaytimeRespert)))))