In our study, we find a lack of sufficient evidence for weather-mediated seasonality of #COVID19. 1/X
I’ll let @MauSantillana take the lead on fielding questions about this work, but I wanted to point out two things.
First, I use the term “weather-mediated seasonality” conscientiously. There are a lot of other phenomena that *also* mediate seasonality, like behavior change. 2/X
For example, I don’t doubt that congregating indoors – which will likely happen more often as winter emerges in the U.S. – may very well increase transmission risk in temperate climates.
These issues are extremely important, but they aren't what we’re studying in this paper. 3/X
Instead, we examine the spatial variability of #SARSCoV2 transmission across China & show that weather-related (or environmental) factors alone can't explain said variability. From there, we draw conclusions that are now unsurprising, given summertime upticks in #COVID19. 4/X
Which brings me to the second point I want to make today.
This paper was first posted as a preprint in March 2020 (papers.ssrn.com/sol3/papers.cf…). At the time, summer hadn’t yet arrived in the Northern Hemisphere (which made us the bearers of bad news, considering our findings). 5/X
…But if we hadn’t posted our work as a preprint, this study wouldn’t have been released to the public until today – well-past summertime in places like the U.S. Thanks to preprinting, we were able to share what we’d found early enough to (potentially) make a difference. 6/X
Whether or not we actually *made* a difference is a separate story... But to me, this is the beauty of preprinting. Though preprinting has serious weaknesses (e.g., it can prompt the proliferation of studies that are largely unvetted), it also allows us keep science current. 7/X
And with that, I'll wrap up with a thank you to @MauSantillana for letting me be part of this important study. That said, our work (like all science) has limitations (many of which are noted in the manuscript) – so if there are any curiosities, please direct them his way! 8/8
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Recently, I've had a lot of folks ask me whether it'll be possible to safely visit loved ones during the 2020 holiday season. In the US, #COVID19 will likely still be with us then – but "merging bubbles" may be an option (if done responsibly).
Let's discuss what that means. 1/X
The concept of "merging bubbles" involves bringing two or more households together such that they can interact with each other (indoors and in person). In this thread, I'll be offering my own personal views on how to minimize #coronavirus transmission risk while doing this. 2/X
Let me preempt this by saying that the advice I offer today is based off of the best available knowledge at present and is subject to change between now and the holiday season. If it does, I'll be adding an addendum to this thread – so be sure to bookmark and check back then! 3/X
We use machine learning techniques to map the existing #coronavirus literature & identify research needs for #SARSCoV2 (compared to #MERS- & #SARS-CoV here).
1/N
(This study was led by @AnhvinhDoanvo in collaboration with @Money_qxl, @qramjee, @EvolvingEpitope, @angeldesaimd, & myself. It started off as one of our research hackathon projects in late March, & it's been such a delight to oversee its development over the last 11 weeks!)
2/N
Perhaps unsurprisingly to those familiar with scientific funding mechanisms, we find via PCA & LDA that #COVID19 studies (both preprints & peer-reviewed) have primarily been dominated by clinical, modeling, & field-based – as opposed to laboratory-based – research to date.
OK, folks. I know things seem really scary right now. The landscape of our response to #COVID19 is changing by the hour, much like our knowledge about this disease (from a scientific lens) has changed by the hour over the last two months.
Please prioritize self-care, y’all. 1/N
For most of us, the best we can do is practice disease prevention (wash our hands, cover our coughs & sneezes, etc.) and (if we are in the position to do so) limit face-to-face time with others.
I’ve talked about some of these costs here on Twitter, wrt the consequences such measures — when institutionally mandated — may have on our most marginalized communities. That doesn’t mean we shouldn’t implement them; it means we need to do what we can to limit the harms. 3/N
New analysis from me & @mandl suggests that preprints might have driven global discourse about #nCoV2019 (#COVID19) transmissibility prior to the publication of relevant peer-reviewed studies. Find our preprint here [ssrn.com/abstract=35366…] as well as an explainer thread below!
@mandl Assuming representativeness, we first collected Google search trend interest & MediaCloud news volume data on #nCoV2019 (#COVID19) transmissibility. We then curated relevant studies from Google Scholar & four popular preprint servers. (Discovery specs are noted in our preprint.)
@mandl After plotting search interest & news media volume, we overlaid when each of the relevant studies were published. We found that both search interest & news media RE: the transmissibility potential of #nCoV2019 (#COVID19) peaked before publication of the first peer-reviewed study.
I've seen a few tweets recently about how R_0 is the mean of a distribution (via @nntaleb) and how its dispersion is important to understand (via @DFisman & @C_Althaus).
This is very true (for #nCoV2019 & otherwise), and it's why I posted this graphic last week. [THREAD] 1/x
@nntaleb@DFisman@C_Althaus As y'all may recall, R_0 is the *average* number of people a new case will infect in a fully susceptible population... But in a given population, the number of ways R_0 can be 2 (as in the above visualization) is essentially countless because *each person is different*. 2/x
@nntaleb@DFisman@C_Althaus We have different biology and different behaviors that impact our individual likelihood of passing on the infection in question. This is why some people may infect lots of people and some people may infect no one at all, yet yield an R_0 of 2 (see Scenario 2 in the viz). 3/x
We've updated our transmissibility assessment for #nCoV2019! R_0 estimates (based off of publicly reported confirmed cases through 1/26/20 & subject to change) remain ~stable, now ranging from 2.0 to 3.1.
On January 24, this study [thelancet.com/pb-assets/Lanc…] provided onset data for the first 41 cases & back-dated the start of the outbreak to December 1 (instead of December 8, as we'd thought before due to the WHO [who.int/csr/don/12-jan…]). Our estimates now include these data.
Because of the ~1 day delay in reporting, we were able to incorporate these new data starting with our estimates on January 23 onward. There was an initial dip (though still within initial bounds) in our R_0 range, likely due to back-dating, on January 23. It's since stabilized.