To me, 99% of the million pages could be eliminated and there would still be enough meaningful information to approve the vaccine, but the regulators insist on doing things as they always have whether the pages are actually printed, killing all those trees, or handled electronically.
I can give you examples of where companies submitted based on trial results that looked good until the FDA used the raw data to recreate the analysis and saw serious issues.
One of the best examples was DCTH's Chemostat years ago. The trial had a SPA (an agreement with the FDA that if the trial "works" it should be approvable). The trial easily met the predefined primary endpoint. The results were published i peer reviewed papers.
The FDA noted that the stats made clear some devices (it was a device/drug combo) were killing patients.
There is a reason why the look over the raw data instead of believing what sponsors say.,
I do agree with you though that we will probably never see this type data for -L. That would only happen if there is either an ADCOM or an actual approval.
Market Data 
Markets 
AI
