Ann: COVID ARDS Trial Topline 60-Day Results, page-485

After digging a bit deeper re the multiplicity stuff mentioned earlier in the thread, and reading through the FDA guidance link provided:

https://www.fda.gov/media/102657/download

My understanding is that an adjustment is required when there’s multiple endpoints. Otherwise an abundance of endpoints will ensure at least some are reached by pure chance.

This particular section is troubling


With MSB's non-adjusted descriptive p-values (note 5) I can see why the stock has been steadily been sold off day after day despite the stunning results. The p-values while statistically rigorous in isolation, are not in aggregate after the 40~ endpoints

GLTAH but im not holding my breath for Novartis anymore. Will reassess on release of full data dump or peer reviewed results.

(disc: filthy non-holder as of this arvo)