"Big RCT Fundamentalism" is an interesting term. Reductionist positions are not uncommon.
A test of 20 people would very probably not have uncovered the unusual clotting problem in AZ, which is 1:100,000 order occurance. Yet, it has shaped public policy.
Tests of the order 1,000 would confirm AZ worked. Tests of the order 10,000 confirmed Pfizer and Moderna worked "better"
The PEG problem in mRNA, like the clotting problem, needs tests of the order 1m to become apparent.
If Ivermectin and HCQ had benefits, they'd be clearer if better RCT at scale were done. The benefit of steroids (for instance) were so clear, they were able to decide the protocols from thousands of tests (from what I read) which is maybe what "big RCT fundamentalism" tells us: These trials of 10 or 20 people, observational, are too strongly distorted by one and two person effects to say what the underlying trend is.
In primary school we were warned away from small sample sizes in basic stats. I'm not seeing anything here to make me think this was wrong. 100 people tell you more than 10 do, and 1000 confirms the accuracy. 1 person variance in 100 is not altering the fundamentals the way 1 person variance in 10 is.
* 0 out of 788 (0%) in the treatment group got COVID.
* 237 out of 408 (58.2%) in the control group got COVID.
The treatment group were volunteers who agreed to take it, so it wasn't an RCT.
"Big RCT Fundamentalism" discards that study, because it wasn't a big RCT, even though the signal to noise is obviously clear, and it has been replicated study after study, including smaller RCTs (when using a weekly dose of at least 0.2mg/kg).
There are no big RCTs that show the same effect, because they can cost millions and there's no incentive for anyone to fund it.
You're using a very specific complaint about science formalism not about numbers but about recruitment and randomisation. My comment went to numbers, not recruitment. If your problem is outsider science, please be more specific.
You could (for instance) cite PLOS one reports on Ivermectin instead of random domains with an axe to grind:
The ivmmeta website is a collection of public information, you can always refer to the original study/journal if you don't find it trustworthy.
It's not a lack of data. Ivermectin already has many times more data (including RCTs) compared to other WHO approved treatments (Ivermectin vs. Remdesivir, Ivermectin for COVID-19 vs Ivermectin for Scabies).
That author appears to have an axe to grind around HCQ.
He mentions - writing in August 2020 - favorably the argument that an RCT shouldn’t be performed of HCQ because it is already known to be effective. ‘ RCT fundamentalists called their [hydroxychloroquine] study “flawed” and “sloppy,” implying it had a weak methodology.’
Well, now that it’s not august 2020 and we actually have the results of well conducted HCQ RCTs, we know that HCQ turned out to be no better than snake oil. All the claimed observational improvements were just anecdotes and luck. https://www.nih.gov/news-events/news-releases/hydroxychloroq...
The mistaken conclusion that he came to around HCQ should be factored into evaluating the validity of the arguments he makes - which led him to downplay the need to study and verify whether HCQ was actually working.
There was a disinformation campaign against Hydroxychloroquine. In the failed JAMA study, they administered a nearly toxic dose at late stage patients (when there's no viral replication). This is now being investigated.
https://www.tabletmag.com/sections/science/articles/randomiz...
https://youtu.be/LkIKCbwBcLU?t=1940