Frankly I am tired of your schtick. I am not reaching, I read the study and see they failed to include data to assess disease severity. You read DL's blog.
If you can't see the flaws in that paper then you are blind. Patients spanning back to last December when they did not even know what they were treating or how to treat. Up until 6 weeks ago we were killing patients with an incorrect ventilation approach. 48 hrs after hospitalization is roughly 15 days of disease course which is no different then almost every other retrospective study that shows little benefit. Also, hospitalization by definition means the course was severe. Mild are those who are either symptomless or nearly so. Moderate is I can tough it at home. Hospitalization is the 3% who can't breath and are in serious trouble.
They used qSOFA to determine severity LOLOLOLOL.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7167215/
We anticipate that the qSOFA is not appropriate to identify Covid-19 patient to have poor outcomes typical of sepsis. Also, where is the min, max, mean of SPO? If you have two groups of 1000 and both have 50% with patients below 94%. Are they equivalent populations if one has min, max, mean of 50, 96 and 71 and the other 88, 98, 91?
That is how you manipulate data to say what you want it too. Please remember what I have been doing for 20 years. I see it all the time.
However, since all you can do is read DL blog, I will help you understand the data a little better with a free insight:
The ICU stay was twice as long for treated patients vs the control and that was extremely significant. The treated patients needed mechanical ventilation at a rate of 3X the amount in the control. Again an EXTREMELY significant difference between treated and control. Conclusion, the treated patients were worse off. Which is what has happened in every retrospective study so far. Logically, it makes sense. If someone's IL-6 and CRP indicate extreme situation naturally doctors start trying hail mary's. I would bet you any sum of money the IL-6, CRP, etc of the treated group was SIGNIFICANTLY worse than the control.
The Indian Govt did a study on several thousand workers and made decisions that impact many thousands of lives directly based on the results. Just because they have not shared that data does not mean anything other than they are not sharing it. The Lancet paper is retrospective with no data on disease severity and itself admits its limitations "Nevertheless, a cause-and-effect relationship between drug therapy and survival should not be inferred."