Well, we're on twitter so I'm a bit informal If you want a more pedantic answer, the statistical analyses appear unjustified, are under-reported (R^2?) , and thus the results may be entirely spurious. The confounding, which is briefly discussed, is not just a limitation
-
-
But is an intrinsic flaw in the study design. Pre/post studies, by design, are challenging to infer causality from, and while there are observational designs that could have supported the statements made by the authors, this is not one of them. It is probably not appropriate
1 reply 0 retweets 1 like -
To even use the word "association" given that no effort was made to quantify the strength of that association in the paper. Moreover, given the obvious outliers, it appears likely that moving the analysis slightly would entirely change the results, which raises the possibility
2 replies 0 retweets 1 like -
Replying to @GidMK @auscandoc and
I agree it was a weak study with multiple limitations. Must be interpreted with the existing literature. Do you think medical masks worn by HCW in health care setting prevent HCW getting respiratory illnesses?
1 reply 0 retweets 1 like -
Replying to @TheSGEM @auscandoc and
I'm not against masks at all! What is frustrating me - and in part drove my glib response - is that really poor research studies are gaining enormous traction simply because they support mask wearing. This study has an altmetric of 2,400!
1 reply 0 retweets 2 likes -
Replying to @GidMK @auscandoc and
I think we are probably more on the same page. My concern has been the over-interpretation of the weak literature. I was pleased to see this study was not making claims of causation. Here are my thoughts on universal mandatory public masking from a debatehttp://thesgem.com/2020/05/sgem-xtra-mask4all-debate/ …
1 reply 0 retweets 2 likes -
Replying to @TheSGEM @auscandoc and
I think the problem is, at this fraught point in time, that any studies like this are bound to be interpreted causally by the public regardless of the statements made in the conclusions. Given that, I think it's more appropriate to just not publish such flawed research
1 reply 0 retweets 2 likes -
I can vouch that you both are very much on the same page. The page of, dare I say it, “robust” scientific research. Why I’m very glad I tagged you both _and_ you shared such useful critique..../2
1 reply 0 retweets 3 likes -
Replying to @auscandoc @GidMK and
2. But the discussion makes me ponder another vexed question: what if the research had been published but with the summary “masks _not_ associated with decreasing infections in HCW?” How would that have been misused (I’m already guessing Ken’s reply
). ../31 reply 0 retweets 3 likes -
Replying to @auscandoc @GidMK and
3. Of course the other option is, as you have said Gideon, self censoring and not publishing at all. But, then is this healthy for science, even if the quality is not great? I genuinely don’t know.
1 reply 0 retweets 2 likes
My two cents - there is no ethical imperative to publish research from which basically no conclusions can be drawn. On the other hand, there's a strong reason NOT to publish these studies
-
-
Replying to @GidMK @auscandoc and
There is a balance that’s hard to obtain. How to figure out the best signal to noise ratio. Literature already suffers from publication bias. It’s complicated...
1 reply 0 retweets 3 likes -
Based on this discussion maybe a different title. “Masks may (or may not) decrease SARSCOV2 infections in HCW. An important question that our study is not precise enough to answer”
0 replies 0 retweets 1 like
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.