AI
Thursday, August 12, 2021 4:38:02 AM
Ah ah ah this is everything but rigorous scientifically.
I see on the internet many claims about "science" from authors who very obviously know nothing about real science. So let me synthesize a few key points.
The scientific method requires rigorously controlled trials, double/quadruple blinds with placebo arms to control for the placebo effect, a sufficient population to control for bad luck and predefined endpoints. This is the gold standard and the only one worldwide to deem a treatment effective. These EINDs & anecdotes were everything but that. Placebo effect, small samples that had uncontrolled mITT corrections, no predefined endpoint.
When this rigor was partially introduced - through the CD10 and CD12 COVID trials - magically Leronlimab missed all its endpoints. Despite supporters and hardcore fans predicting a few weeks before that "without a doubt" the results would be stellar. But but but ... the miracle vanished into the air.
Instead we got botched statistics that don't resist a 30 seconds examination by a professional statistician. One example: Nader acknowledged in a conf call that they DID NOT PERFORM P-VALUE CORRECTIONS TO ACCOUNT FOR THE DATA MINING. That, alone, KILLS FURTHER ANY P-VALUE NUMBER THEY DATA MINED. If you perform these corrections, their p-values (even filtering in/out patients by age, condition or number of days since onset) become ridiculously high, just pure noise.
Sorry, that's REALITY. Investors should consider exiting this mess that is only a stock promotion whose goal is to scam them and enrich insiders with their capital.
I see on the internet many claims about "science" from authors who very obviously know nothing about real science. So let me synthesize a few key points.
The scientific method requires rigorously controlled trials, double/quadruple blinds with placebo arms to control for the placebo effect, a sufficient population to control for bad luck and predefined endpoints. This is the gold standard and the only one worldwide to deem a treatment effective. These EINDs & anecdotes were everything but that. Placebo effect, small samples that had uncontrolled mITT corrections, no predefined endpoint.
When this rigor was partially introduced - through the CD10 and CD12 COVID trials - magically Leronlimab missed all its endpoints. Despite supporters and hardcore fans predicting a few weeks before that "without a doubt" the results would be stellar. But but but ... the miracle vanished into the air.
Instead we got botched statistics that don't resist a 30 seconds examination by a professional statistician. One example: Nader acknowledged in a conf call that they DID NOT PERFORM P-VALUE CORRECTIONS TO ACCOUNT FOR THE DATA MINING. That, alone, KILLS FURTHER ANY P-VALUE NUMBER THEY DATA MINED. If you perform these corrections, their p-values (even filtering in/out patients by age, condition or number of days since onset) become ridiculously high, just pure noise.
Sorry, that's REALITY. Investors should consider exiting this mess that is only a stock promotion whose goal is to scam them and enrich insiders with their capital.
