3 ms·

> And yet we still can't show the drug works. There has been just one trial done properly to my knowledge, and that was published yesterday. And even that is j

> And yet we still can't show the drug works. There has been just one trial done properly to my knowledge, and that was published yesterday. And even that is just applied in a narrow case (post-exposure prophylaxis, modulo some important study limitations).

kevin_b_er 6y ago
That one's being attacked as we speak for certain poor controls on testing. This could lead to a compromised dataset and invalidating the whole thing.
- lbeltrame 6y ago
  Yes, I've read the limitations of the paper (after the Lancet debacle, I too learned my lesson and read a little more thoroughly). I believe it was done the best they could, given that UMN had to use very little money for it (~100K USD, as far as I understand). That's why IMO the limitations need to be fully read and understood. But I think it was done in good faith (unlike the Lancet study): the PI tried to his best to avoid spilling the beans until it was published to avoid sensationalism, and at the same time, he signed the open letter against the Lancet paper (after he sent the paper for review, so he knew the results already). Larger studies will likely be needed. It's clear that the effect size (if any exists) is quite small, like with remedisivir.
tjenkinsqs 6y ago
Five studies. Key difference: Used with zinc and used with outpatients. https://academic.oup.com/aje/advance-article/doi/10.1093/aje/kwaa093/5847586?fbclid=IwAR2x3eEIuV9EbOmtAN7U35eZsmMNsohWXrh-RIwxTIei5KhHa5iKsk2YE2o https://academic.oup.com/aje/advance-article/doi/10.1093/aje...
- abbracadabbra 6y ago
  Since this is being repeated multiple times, quoting benmaraschino: One of these "studies" is a 2-page Google Doc with no real data. Another is the discredited Raoult study, which is under investigation by the journal [1]. Another study cited is also by Raoult's team. The Brazil study (ref. 29) doesn't seem to have passed peer review (yet?) and has severe methodological flaws. Like the Raoult studies, the Brazil study was not randomized, but makes no attempt to control for confounding by indication or self-selection, such as with propensity score matching or similar methods commonly used in observational studies. The last study cited isn't a study, but an article from a local news station.

kevin_b_er 6y ago

That one's being attacked as we speak for certain poor controls on testing. This could lead to a compromised dataset and invalidating the whole thing.

lbeltrame 6y ago
Yes, I've read the limitations of the paper (after the Lancet debacle, I too learned my lesson and read a little more thoroughly). I believe it was done the best they could, given that UMN had to use very little money for it (~100K USD, as far as I understand). That's why IMO the limitations need to be fully read and understood. But I think it was done in good faith (unlike the Lancet study): the PI tried to his best to avoid spilling the beans until it was published to avoid sensationalism, and at the same time, he signed the open letter against the Lancet paper (after he sent the paper for review, so he knew the results already). Larger studies will likely be needed. It's clear that the effect size (if any exists) is quite small, like with remedisivir.

tjenkinsqs 6y ago

Five studies. Key difference: Used with zinc and used with outpatients. https://academic.oup.com/aje/advance-article/doi/10.1093/aje/kwaa093/5847586?fbclid=IwAR2x3eEIuV9EbOmtAN7U35eZsmMNsohWXrh-RIwxTIei5KhHa5iKsk2YE2o https://academic.oup.com/aje/advance-article/doi/10.1093/aje...

abbracadabbra 6y ago
Since this is being repeated multiple times, quoting benmaraschino: One of these "studies" is a 2-page Google Doc with no real data. Another is the discredited Raoult study, which is under investigation by the journal [1]. Another study cited is also by Raoult's team. The Brazil study (ref. 29) doesn't seem to have passed peer review (yet?) and has severe methodological flaws. Like the Raoult studies, the Brazil study was not randomized, but makes no attempt to control for confounding by indication or self-selection, such as with propensity score matching or similar methods commonly used in observational studies. The last study cited isn't a study, but an article from a local news station.