I have indeed read the paper, but I have to admit not with the devils advocate perspective as my statistical knowledge is not good enough even though I have some of it in my Economics degree.If you read the paper, that is not actually how the patients were assigned. The patients were assigned by the ward that they were located in and it appears that the patients in the wards that did not get treatment had a significantly lower level of Vitamin D before the treatment even started. That is a red flag as it indicates that the wards are not randomly populated. It is broadly accepted that Vitamin D levels are associated with worse outcomes. The question is whether supplementation helps. Another red flag is that they did not do their statistics correctly based on how they designed the study. The correct statistical test suggests that it is possible that there is no benefit at all.
They couldn't even successfully balance by gender. There is a significantly higher proportion of men in the non-treatment arm. And we know that is also associated with worse outcomes.
I said randomly because the study says:
Participants (n=551) were randomly assigned to calcifediol treatment (532 ug on day one and 266 ug on day 3, 7, 15, and 30) at the time of hospital admission or as controls (n=379).
Aren't you exaggerating any problems with the study now?
It is true that they study was not balanced perfectly in Vitamine D levels or by gender and a bunch of other factors..
but the regression analysis adjusts for a host of these issues, including age, sex, vitamine D levels and comorbidities.
The chance of this being coincidental is according to the paper found to be p=0.001 for mortality and p<0.001 for ICU treatment.
But as I said, the paper is awaiting peer-review in the Lancet, so I am sure any issues will be adressed.