Thread by @alexandrosM

OK, what's going on with the retracted FLCCC paper? I dug in and here's what I found: 🧵

https://twitter.com/PierreKory/status/1458892074510651405

First things first, the paper. Twitter doesn't like scihub links, so you'll have to put the following three strings one after the other in your browser. I don't make the rules 🤷

https://sci-hubtw

.hkvisa.net/

10.1177/0885066620973585

Secondly, the retraction notice:
journals.sagepub.com/doi/10.1177/08…

So, to understand what is happening we need to understand what the MATH+ protocol is. At the heart of it is a cutesy acronym:

(M)ethylprednisolone
(A)scorbic acid (aka. Vitamin C)
(T)hiamine
(H)eparin (specifically Enoxaparin)
(+) a number of co-interventions

The hospital saying "Only 73/191 patients (38.2%) received at least 1 of the 4 MATH+ therapies, & their mortality rate was 24.7%. Only 25/191 patients (13.1%) received all 4 MATH+ therapies, & mortality rate was 28%"...

doesn't deny 191 patients got treated with the "+" of MATH+

If they wanted to say "only 73 patients were treated with the MATH+ protocol", they could have written just that.

Instead, they're parsing how many of the M/A/T/H therapies patients were treated with. The only reading that makes sense to me is that they're being disingenuous.

As you can see in this table, all hospitalized patients receive Atorvastatin, Melatonin, Zinc, Famotidine. What does it matter if they got none of the headline MATH therapies, and just certain ones of the "+" co-interventions?

AFAICT, the conflict is about the last line of the table here, and the only number mentioned is the 6.1% one, which the hospital challenges. Why go into the other details, if not to imply that either the protocol wasn't used, or that where it was used, mortality was far higher?

So what's left of the dispute? The difference between 6.1% and 10.5%. According to the hospital, an additional 8 people died after the end of the study, in addition to the 12(?) that died during the study.

So what's the proper way to calculate deaths? I looked at the first paper on the list, the Docherty study. bmj.com/content/369/bm…

The key word is "continued to receive care at the reporting date". So in the first compared study, 40%+ was still not discharged, and I assume many of them died.

The mortality reported in the FLCCC paper table is 26%, same as reported in the Docherty paper itself, excluding the 41% still in hospital. So why would the hospital demand FLCCC revise its numbers up if that's the same method followed in the other papers they're comparing with?

Overall, no idea what the hell went on, and I understand there's a legal case either filed or in the works, but so far as I can see, what Kory said checks out, and what the hospital is saying looks optimized to mislead. I may have missed something, in which case let me know.

Correction, Arorvastatin was not given to all hospitalized patients, only Melatonin, Zinc, Famotidine.

https://twitter.com/DrAlexofTruth1/status/1459271251030065152?t=krrn4PK87oh24XmVdJHT6g&s=19

Please note the below tweet. I'm still digging in, but if this is right then not receiving one of the M/A/T/H treatments would be the same as "not being treated with MATH+". Still trying to understand if this is a correct reading of the table.

https://twitter.com/DrAlexofTruth1/status/1459272200394006528?t=pShiqmjkueRS7rFJJPlkxg&s=19

OK. This is odd. Because @DrAlexofTruth1 is correct above, then what the hospital is saying is that:

18/73 treated with at least 1 MATH therapy died, in line with other hospitals, but only 2/118 of the remaining patients died, which is wild for hospitalized covid patients.

There is a conflict between what I read in the table, and what I understand the hospital is claiming. Can anyone help explain the discrepancy?

The only explanation that makes everything fit together is that possibly there are "hospitalized patients" that are neither "hospital ward patients" nor "ICU patients", but are instead in some milder class? If anyone can help explain this it will help greatly.

OK. Figured it out! At Sentara Norfolk the protocol was administered upon admission to the ICU only. Meaning that the 118 patients never got to the ICU and never got the protocol but they were recorded the same way so as to make the results comparable to the other studies.

This was clearly explained in the paper, so there's no foul play left to speak of, as far as I can tell.

Good question, here's the answer: Sentara is reported in Table 2. As mentioned, the 1st paper on the list, Docherty, counts patients from hospital admission to discharge. For the results to be comparable, studies must track patients in that same interval.

https://twitter.com/pclafferty/status/1459483073662406656?t=RaX4cIWEwIw0nH_aAv666w&s=19

It's worth mentioning that *not* treating patients before they go to ICU makes their results worse, therefore this inclusion is not some sort of sleight of hand. In fact you can see that their numbers in the other hospital, using MATH+ on admission are even better. (4.4% vs 6.1%)

As a conclusion, I'm back to what I wrote originally, with somewhat more confidence. Looks like the hospital was looking for any reason to trigger a retraction, and the one they came up with was not very good.

https://twitter.com/alexandrosM/status/1459274284195467265?t=RaX4cIWEwIw0nH_aAv666w&s=19

PS. The fog is clearing: this seems to be part of broader situation, the hospital has banned doctors from using MATH+ in its ICU, so they stand to benefit their (upcoming legal) case by having the original MATH+ paper retracted.

https://twitter.com/marc_g_wathelet/status/1459488650086002698?t=qPy1WaxT_bLVSVbNtn9YMQ&s=19