This is just crazy. Mainstream experts have been trying to get through with almost no success. But take an out-there position and you get access. Of course take extra precautions for the most vulnerable. But don't relax everything else before evidence these precautions work.
The argument is incoherent if you don't do low-cost low-inconvenience things like universal masking. Surely any rational strategy uses low-downside strategies to reduce transmission in the whole population while shielding the vulnerable.
Two of these scientists, @SunetraGupta and @MartinKulldorff, have long been my friends. But I think they are dead wrong without a demonstrated plan for how such shielding would work. There is no good example in a dense western country.
Most of us in the #IDepi world support a belt and suspenders approach - suppress transmission and try to shield the vulnerable. At present the "shield the vulnerable" alone strategy is taking scissors to the suspenders before there is any evidence of a viable belt.
Worse than that, this administration has systematically dodged its responsibility to provide a belt or suspenders or any other approach -- falling down on border control, testing, PPE, ICU capacity, and more, and encouraging irresponsible behavior.
Even if a shield-the-vulnerable strategy could work in principle, advocates of it should be pushing first for better testing, PPE, sick leave for staff, mental health protections, and others for nursing homes, and THEN for letting it rip, not the other way around.
Yes there have been proposals for how to protect the vulnerable as some have noted. But proposals and mathematical models of how they might work are not solutions; they are hypotheses (I say that as someone who creates these sometimes). No evidence that they work at scale.
And as @akcayerol put it in more colorful language, this is not an abstract discussion. It is a political discussion that has been made such by the White House. It is the height of hypocrisy to say "protect the vulnerable" and fail to use basic protections for high-risk @POTUS
• • •
Missing some Tweet in this thread? You can try to
force a refresh
New chart reveals military’s vast involvement in Operation Warp Speed. Fascinating article. statnews.com/2020/09/28/ope…
Don’t quite understand how source thinks “foot on the neck to make them go go go” is ok to say in 2020 or even, insensitivity aside, makes sense.
Also don’t know how @HHSGov Spox Mango can say with certainty what the vax trial results will be when they are still not unblinded. Every time someone corporate or govt says something like that they should be asked how they know.
usually agree with both @StevenSalzberg1 and @nataliexdean. In this case agree with the cautious view. Even if safety were known (which I don’t think it is for this) RCT r really important for efficacy. Alternatives, which both @nataliexdean & I work on, are full of pitfalls.
The article by @StevenSalzberg1 just gets some things wrong. Published test of eg the Oxford vaccine had 126 and 253 vaccinated, enough to detect adverse events if occur in 2.3% and 1.4% respectively.
Just learned from a good news article in @naturenature.com/articles/d4158… about this site metrics.covid19-analysis.org. To be crystal clear it has nothing to do with @CCDD_HSPH. It is not reasonable in my opinion to mske such estimates with any confidence for large parts of the world
The uncertainty stated on this site is purely statistical uncertainty assuming data and model are accurate. This _vastly_ understates uncertainty. In many places, case confirmation is delayed dramatically (weeks) & variably, but this assumes 5 days from infection to confirm'n.
Changing testing practices mean changing proportions of cases ascertained and thus changing estimates of cases and R separate for reality. No correction or acknowledgment of uncertainty.
As one colleague emailed me facetiously "Can you please forward this to John Snow?"
Notwithstanding that economists historically favor different techniques for causal inference from observational data, making such inference is the goal of much observational epidemiology (not all -- sometimes we aim for description or prediction) amstat.tandfonline.com/doi/full/10.10…
@DiseaseEcology@US_FDA@rebeccajk13@Steve_Bellan I may not be explaining well, and it's a subtle thing I'm trying to say. I think it is important to know effect on serologic infection. My point is that there may well be a vaccine that 1) makes disease less likely/severe, 2) makes shedding much less and 3) permits seroconversion
@DiseaseEcology@US_FDA@rebeccajk13@Steve_Bellan Such a vax would look good on the disease endpoint (esp if they use our approach or similar to correct for missed infections), but null on the infection endpoint. This vaccine would be very good for herd immunity (and direct protection) but the analysis might miss that fact.