[FRIAM] FW: Covid-Lancet-PART-2 (002).doc

[David Eric Smith]

Pieter, hi and thank you,

Yes, very good.

Let me respond to your root post here, in parallel with whatever other branches may have grown from it.  I want to note that the points in Glen’s first reply to the same post speak well for the things that animate me, and are better worded than mine.  So I won’t recap them, but want to acknowledge and share endorsement of them.

> On May 7, 2021, at 9:35 PM, Pieter Steenekamp <pieters at randcontrols.co.za> wrote:
> 
> So please help me, I don't understand how you get from, I quote from the transcript  (my underlining) :
> "Safe and effective COVID-19 vaccines were produced far faster than any expert expected. Yet almost all of the time that it took to bring the vaccines to market was due to safety testing and other governmental mandates that could have been sped up without endangering anyone. By January 13, 2020—only two days after the Chinese researchers shared the genetic sequence of the COVID-19 virus and before most Americans had heard of the disease—the biotech company Moderna had devised the formula for its vaccine. BioNTech launched its COVID-19 vaccine program in January and had partnered with Pfizer to manufacture it by mid-March of last year. The first volunteer was injected with Moderna's vaccine on March 16, 2020, yet it was only approved by the FDA last December 17th, a week after Pfizer's vaccine met the agency's approval. Had the agency been faster off the mark and used human-challenge trials and other innovative testing techniques, the vaccines could have been brought to market months earlier with no compromise in safety. That would have conceivably saved hundreds of thousands of lives globally."
> 
> to your comment? I quote from your email (my underlining):
> But of course the article puts up the mRNA vaccines as evidence of how, because the agencies got out of the way (is implied), BioNTech and Moderna had vaccines in a few days.  That is deliberate BS, and I doubt the writer is such an idiot that he doesn’t know it. 

So very good.  I took hours thisAM writing a bunch of awful stuff, feeling stupider and stupider for having written a screed instead of a good argument yesterday, and realized I have to re-examine what I am responding to.  

I think my objection is to a salience structure of the article that I think grossly miss-apportions credit and fault, in a way that is not only unfair but that supports an anti-public and anti-institutional point of view that is harmful in many cases where it succeeds.  So let me list for you what I experience as the salience structure, and what I would replace it with that I think would be more fair and also a better foundation for decisions.  I am not complaining about any of Gillespie’s facts.  Those he quotes that I know are consistent with what I understand from other sources, and I don’t doubt the veracity of those I didn’t know in detail.

So here’s what I read in the paragraph you quote above:

*. Natural timescales for production, accomplished by Companies, were very fast.
**. What Government provided was a vastly longer and unnecessary gap, which didn’t have a good reason to exist.
(Then, to adduce evidence)
1. Companies invented formula for vaccines (only the mRNA ones mentioned): pertinent timescale was 2 days.
2. Company partnered with other Company for production, Company produced a first human-injectable product; pertinent timescale was a bit under 2 months.
3. Government, Regulator, with Mandates, stepped in and 9 months were spent needlessly
4. Because of those 9 lost months (or some unspecified subset of them), some hundreds of thousands of people died needlessly.

Here’s how I would have written a salience analysis for the first part — I will come back a bit later to the last two points, both their substance and how Gillespie renders them.

-2. Government funds mRNA uptake and expression research, but just barely and precariously.  Academic labs keep getting shut down, and their directors go to Companies, where that work is not in the portfolio.  Somehow, remarkably, it does get pushed through to a working system; the outcome could easily have been quite different; pertinent timescale: 30 years, during which much of the cost and all of the long-horizon risk is carried in the public sector.  (Could be, however, that Company grants do contribute at places, and perhaps fill in key steps that happen to fall on the timelines they can support.  If so, note it.)
-1. Government (Obama admin) seeds startup in thermostable lipid delivery systems for mRNA; pertinent timescale is 5-10 years.  (I don’t know BioNTech’s history or who funded it, if it was Company funded as a venture, that matters and should be noted.  If public, then that.)  There are no products in either country, because the disease targets turn out to have either low enough severity and sales potential, or small enough coverage that it requires too-large clinical trials to get statistics, for any Company to be willing to foot that bill for an unproven technology.  
0. Technology sits around and skowly gets more mature, but has no clinical trial testing history beyond basic safety of the delivery system.
1. mRNA vaccine Companies are finally able to pull the trigger on the gun that has been sitting loaded for years, and provide a formula in a couple of days.  EXCELLENT — CREDIT TO ALL INVOLVED.  Standard technologies, like adenovirus technologies (I think not mentioned in this article), were already ready and are fairly quick, though less so because of the inherent limits of the technology.  STILL EXCELLENT.  
2. Companies partnered with other Companies for production (GREAT — JUST WHAT THEY ARE SET UP TO DO).  Some important Government regulatory protocols that normally are serialized get parallelized because of the singular circumstances.  Potential testing timeline is shortened from typical 3-5 years to potentially less than 1 year if we get lucky.  (We should come back to why they are normally serial, and whether even in light of what we have learned, they would be reset that way anyway because it costs less and is acceptable for many routine procedures.  If not, and we want to go parallel, then that’s a great prod to innovation prompted by this pandemic.). In any case, first human clinical trial; pertinent timescale: a bit less than 2 months.
...
etc.

So, the way I responded yesterday, which you quote, is both needlessly rude and also unclear; thank you for pushing back.  Let me be careful in the next, and reply to your exact words: 

> Do I miss something? I don't read in the transcript that they said or implied that BioNTech and Moderna had vaccines in a few days? 

I wasn’t attributing that as a Gillespie claim; I was taking that as a factually roughly-correct description of what really happened.  It is probably right that the “design of the formula” isn’t the same as “having the first vaccines”.  However, RNA manufacturing is an off-the-shelf technology, which both these companies almost-certainly have in-house and don’t even need to go across town to get.  Since the lipid delivery platforms are their product, the transition from sequence design to first injectable droplets was probably very fast; an assembly of components already on hand in their own shop.  If they do get their RNA from a foundry, for a high-priority project they could probably get it made in a day, across town or via airmail.  It probably took a bit longer to get enough production to run a clinical trial — that could have been weeks or even a month or more — but that is production scale-up and consistency checking, all exacting but established Pharma technical stuff.

> Maybe I'm such an idiot that I don't see it?

What I was saying here was the I don’t think Gillespie is at all an idiot.  Therefore I expect he understands well that the rapid delivery of the first mRNA vaccines (specifically) was putting the cherry on top of a sundae that had been a long time in the making, with the biggest foundation being public, and essential but timescale- and risk-limited roles for companies in partnership.  My objection was the salience one above, putting up short timescales for that turnaround as if they were the natural timescales of the process, mentioning them only in association with Companies, which then only Big Regulatory Government with its Mandates came in and broke down.  

If he had talked _only_ about the adenovirus vaccines, already established in company production lines, rolled out unusually quickly (to their credit, in just the way he says generally), and with a history of clinical trial evaluation of essentially the same technology, then his discussion of regulatory roles would have indeed focused on the only remaining discretionary variable in affecting the time of the public availability.  Everything else was more or less constrained.  Only mentioning the adeno vaccines would have been a lot less compelling rhetorically, though, since J&J is a minor player in the US, Astra not yet at all (in the US, unless it has been approved and I haven’t kept up), both are slightly less-flashy by the efficacy numbers, may have more limited scope for use in boosters if there is immune response against the viral vectors that disables the delivery system in second doses before it can deposits its payload, and also have rare inflammatory side-effects that have been enough to give them some negative PR (though do not change the fact that they are excellent vaccines).  

Yet I think the only companies he mentions are the mRNA ones, for which the turnaround timescales are shortest and most impressive, and seem by comparison to make the Government Regulatory timescale look most blameworthy.  That is what I felt is disingenuous.  If you are going to use those as the standard for government responsiveness, why not mention that the most impressive timescale is that they exist at all, and which actor ultimately made the difference between now and never?  


But let me let that part go.  You may still find my objection unjust as well as inept.


I do want to respond to the points 3 and 4 above, though, because they are part of a tone that undoubtedly was what I was ranting against.  The specific sentences are:
"Had the agency been faster off the mark and used human-challenge trials and other innovative testing techniques, the vaccines could have been brought to market months earlier with no compromise in safety. That would have conceivably saved hundreds of thousands of lives globally”

I want to flag where I think this is specifically manipulative, and suggest what I think would be a more correct and productive framing.  

Those two sentences are an assertion about constraint and causation, meaning that removal of a constraint would have caused a better outcome. Of course, they are not literally an “assertion", since everything is framed in conjectures or conditionals.  That’s how lawyers and debaters talk.  But with a clear intent to have the reader conclude what they only express as conditional.

Let me give an ad absurdum for why I say it is a bogus assertion of causation.  Not that the ad absurdum is a great literal model of the Gillespie language, but so I can be unambiguous about the structure of the argument I want to flag.  Here:
1. This pig doesn’t have feathers.
2. Feathers have all these important functional roles in flight.
3. If this pig had features, he could fly.
Why is that a bad argument?  After all, each of the first two points is incontestably true.  It is a bad argument because there is a complicated multi-factor relation behind any “true” concept of cause.  The above not only leaves out things, it does so in a way that is designed to distort.  There are probably many reasons this pig (and others) don’t fly.  For instance, maybe they just don’t want to.  A careful man might worry that, even if the pig had feathers, he still wouldn’t want to, which would keep us from concluding he “could fly”.

So, if the FDA approvers had been faster off the mark “could” the vaccines have been brought to market faster?  (Btw, great language; keeps our minds in the right frame: they are not “made available to people”; they are “brought to market”.)  “Would” they then have saved hundreds of thousand of lives _worldwide_?

So let’s go through some context-seeking questions.
1. [dumb and annoying: worldwide?]. FDA is a US agency.  How much do we think that US FDA approval of vaccines made in the US could have saved lives worldwide in 2020?  (Is Moderna even used outside the US?  Have any vaccines originally bought by the US government from either Pfizer or Moderna been distributed outside the US?)  Does Astra's use in any country besides the US depend on US approval?  So the “worldwide” looks to me like a red herring, put there to allow bigger numbers who “would conceivably have” been saved, but not real.
2. So let’s be a bit more conservative.  It could/would have saved how many lives in the US then?  Hundreds of thousands?  That is a reasonable number to have been saved.  Do we think vaccine approval on any achievable timeframe is the major, or even an eligible, factor in their not having been saved?

In my email yesterday, I used the snarky language 
“There was nothing else going on at the time?  Hmm, can’t recall.  Or since?  Or still, even worse?”
Glen, in his reply late-night US, I think was reminded of BLM and other social upheavals, perhaps by that comment.  Those are true and also relevant, but had not been in my thought at the time of writing.  I was referring, rather, to two HUGE things going on at the time:

A. Almost every federal agency had had its chief offices staffed by people who were either out of their depth or incompetent, or put there as deliberate saboteurs.  I would class Redfield at CDC as out of his depth and incompetent, but not a saboteur.  Azar was a kind of flatly-unqualified lackey, though probably less than a pure saboteur.  Giroir may have been tolerably competent, and in any case did some things okay.  I forget now exactly what I thought of people like Stephen Hahn heading FDA at the time, when news was more immediate.  We do have names of various others who were harassing, interfering in, and distorting the output of, multiple officers within CDC.  Again, I forget which people and which news articles, but all that can be dug up.

Anyway, to get to what matters:  I think it is fair to say that FDA and CDC committed a couple of important errors early on (CDC in not allowing tests to be used just bypassing a superficial third screen that wasn’t essential, or accepting early German designs), FDA (maybe, though I want to revisit their reasoning) in blocking universities or others from participating on various things (was it testing?)

Some of that could have been creeping incompetence within the agencies; it’s fair to ask.  Maybe some of the choices were actually hard, given standing rules and available information.  But I think it is sure that all of them were operating crippled, internally disrupted, I believe on cut funding and cut staffing in many cases, and certainly not by any means in their best form.  We saw the same people, under good leadership, do much better, for instance under Tom Frieden on the ebola response, so I think we know they can.  

To criticize agencies while they are under attack by a government whose goal is to make them fail so it can eliminate them — not mentioning that any of that is going on — while suggesting that if they could be moved aside either the same government would save lots of lives, or that the lives would somehow get saved by other means not relying on the government (a question on which we have good data from the current administration’s role), was one thing I was calling a BS move.  


B. In parallel with the undermining of institutional function officially within the administration, there was an active and full-time campaign to amplify conflict and dysfunction in the public.  It was in place at a kind of baseline from the US right-wing media before the last presidency started, added a whole new circus show starting with trump for the first six months of 2020, and then propagated down through the whole republican apparatus as the year went on, becoming even more fanatical and destructive now that he is out of office.  We now have the vaccines and 30% of the country won’t use them; I am comfortable suggesting that those  people were the same in 2020 as they are in 2021.  Once masks were being advocated (after the early-year mess about how to handle that, as we discussed), during the long period when public health measures were _all_ we had under the best conceivable vaccine outcome, and when every country in the Eastern Hemisphere was using them to good effect, we still had half the country refusing to wear them, and all but a few right-wing politicians pouring fuel on that fire as fast as they could carry it.  So it is not idle of me to say that the political effort had a strong causal role in what could or could not have been done, or what lives would have been saved by vaccines “because" there weren’t other ways to save them before. and “because” they would have been used afterward.

Not only not admitting, but indeed choosing not to mention, the constant and committed role of the political right in sowing confusion, mistrust, belligerence, and sociopathic behavior as an important factor in what could/would happen in the US, is what I call a second deliberate BS move in the Gillespie presentation.

(Note: I do understand that not every bad behavior in the US came from right-wing political forces.  No country in the West did well, compared to many in the East, presumably because the East had experienced SARS and MERS and took these things seriously.  Likewise, we learned since that the trump government didn’t have much of a plan for distribution in place beyond shipments to states;  that, together with states limited by budget crises, also affected availability post-approval.  So, many factors go into how much difference in mortality really turned on some number of months’ difference in FDA approval time.)


Next, let’s come back to the wording “ with no compromise in safety.”.  That was my point in the Hippocratic oath.  I believe that is bogus several ways.  Here again is the ad absurdum:  I ran across the highway and didn’t get hit by a car, so clearly there wasn’t any risk in running across the highway.   Nobody thinks that’s how risk assessment works, so insinuations that because some particular thing is found ex post not to have gone wrong, that shows that there would have been no compromise in not checking for it ex ante, looks to me like deliberate distortion.  Please recall that when the mRNA vaccines were first found to have 90-95% efficacy, people were floored.  They were hoping to get above 50% for symptomatic disease.  If the vaccines had been in that expected performance range, and human challenge clinical trials had been done, how many people would have been expected to get sick, possibly severely, possibly with long-term problems, or possibly dying?  Some.  Not none.  If a human-challenge trial had been designed to get comparable statistics to those delivered by the 66,000 people in the non-challenge clinical trials for Pfizer and Moderna taken together, maybe quite a few.  Medical risk assessment is hard.  There are lots of occasional but very consequential troubles that arise with such mundane things as flu vaccines — cytokine storms and crippling encephalitises, etc.  (I happen to know a woman, since deceased, who lived for decades with brain damage from a flu vaccine that just happened to hit her the wrong way.)  This is why approvers are careful.

Yes, lots of people died in the months before a vaccine was available.  How do we assess that fact, when other public health measures that could (as we know with certainty) have saved some of them were not used?

Glen did mention another thing, about trust in institutions, that is dead center to this question.  If I wanted to do the unpleasant thing that debaters do, I would trot out Media Trope #15 that is currently going around every outlet:  Black Americans say they don’t want to get vaccinated because they don’t trust the government because of the Tuskegee syphilis trials.   Good reason, but of course it is embedded in a fabric of institutional distrust that is a tilted playing field for _everything_ a government agency has to do, with every constituency for one or another reason.  And that is the center of the Hippocratic oath.  The reason doctors won’t do things, even if the patients are desperate and would approve them, is partly because doctors are trying to preserve medicine as a practice that can be trusted.  One thing that did or didn’t go wrong in one vaccine roll-out — called after the fact — is not the relevant context for a decision structure.  One could go on and on about threads in that fabric.  The awful Harvard psych experiments that get trotted out because they were used on Ted Kaczynski, the LSD experiments on US servicemen, all of anatomy that was written using data from Nazi doctor dissections, for which reason it was withheld from use in medical schools for decades even though ti was the most complete data available.  I don’t even know how much else across countries and wars, because I am not informed on that topic.  All these together make up the decision context these agencies work under, and I think to be fair one must grapple with the difficulties that result.

That brings us back to whether the FDA approval timeline, which would have been the major available expediting variable for the adeno vaccines Gillespie doesn’t mention, really had the same relation to the mRNA vaccines he does mention.  I don’t think it does.  Precisely because nothing had been through full clinical trial, and because mRNA is a rather different insertion than viral DNA, I think the standard for caution with the mRNA vaccines had to be much higher.  We appear to have got very lucky (on the shoulders of much good design), and they seem likely be even safer with fewer side effects than viral vectors, as well as better for boosters, faster roll-out, greater mass-production scale, etc.  But before 100M people got vaccinated, that wasn’t known.  


Anyway, I’ll stop.  I have probably driven you mad with tedium if you have read this far. 

It’s shocking and appalling how much better a writer Gillespie is than I am, isn’t it.


Eric



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20210508/20159cf3/attachment.html>