The data on Fluvoxamine is quite weak. The primary outcome was a composite outcome of ED stay length >6hrs and hospitalizations. With a difference of 10.4% (Fluvoxamine) vs. 14.7% (placebo). As anyone who has ever been in the ED, length of stay can be quite long and can be quite arbitrary too based on hospital capacity, etc. More importantly, when you look closer at the data, hospitalization is 10% (fluvoxamine) vs 12.6% (placebo) with a p-value of 0.12, not statistically significant. I highly recommend the blog: https://rebelem.com/the-together-trial-covid-19-and-fluvoxamine-take-two/ from a very good ER researcher. Matt, please don't make fluvoxamine the liberal ivermectin. Both drugs have some promising data but are by no means strong enough to recommend.
Yes! And mortality and time-to-death are not changed at all with Fluvoxamine. That's right there in the text of the TOGETHER study. Matt, Vox, Astral Codex, etc. are a big echo chamber with none of them reading the primary sources beyond the headline. To throw shade on FDA based on such shallow "research" is really shameful. The official TOGETHER report is here: https://www.thelancet.com/journals/langlo/article/PIIS2214-109X(21)00448-4/fulltext . To opine on FDA's decisions on Fluvoxamine without reading the whole thing is just disgusting and dishonest.
If we're talking about what's ACTUALLY optimal -- absent supply shortages -- isn't two three-week-spaced shots, followed ~six months later by a booster the optimal thing over the entire timeframe? Which is, basically what everyone in the US got? I'm not sure what the "screwup" is that's being argued here.
Giving those two initial shots conferred significantly better protection than just one (especially against Delta) for those six months before booster-time came around.
Even in a supply-constrained situation where you only ever get two shots, there's a reasonable argument to be made that having six months of suboptimal single-shot protection isn't worth the tradeoff of better post-six-months protection. Especially when new treatments and therapeutics are coming online, and we can pretty reasonably expect that dangers six months from now will be less than they are now.
When I saw MY's stash of testing kits on twitter I felt a little chagrined that I only bought one kit when they had a big stack at my Walgreens. But I guess I was thinking it was smart to wait for the reimbursement rule to kick in before buying a bunch -- they are not exactly expensive, but each set of two is a decent bottle of wine cost and the wine is more enjoyable.
The challenge for the intelligent generalist pundit, like Matt, is to be read up enough on a subject and fast enough to be able to say something interesting and meaningful. The problem is that you don't know as much as the people on the inside and often don't really have the professional background to discern what is a screwup and what is driven by concrete obstacles.
For all I know, the FDA and the CDC have been and continue to screw up in unforgiveable ways (Scott Gottlieb's new book, "Uncontrolled Spread" is very good on this, though I note that he's a lot more unforgiving to the CDC than the FDA, which he had headed). But what I also know from professional experience is that it's often the things you don't know that are really meaningful and can refute what you think is the truth and the path forward. Life is complicated.
An example of this is congestion at the LA/Long Beach port. Bad problem, and people who know very little instantly opined on how to fix it. But as you dig deeper into the details, you realize that we're facing a series of bottlenecks, and each one needs to be resolved sequentially, after doing due diligence. It's usually never "one weird trick."
So Matt may be way off base here -- though none of us is expert or informed enough to know. That said, we need people like Matt out there asking interesting questions and trying to inform the public, even when he doesn't know enough to give us the full truth.
If the people on the inside are consistently incapable of explaining things to intelligent people like Matt then the fault is clearly with them.
The people on the inside have also been saying things that turned out to be wrong/lies all throughout the pandemic, making things even moreso their fault
Well, organizations should always do a better job explaining themselves, but I don't really want them to sit down with every Substacker with 13,000 subscribers to explain in detail what they're doing.
Also on masking. LA Public health is still handing out CLOTH masks at vaccine centers, with public health logos on them. This suggests that LA Public Health thinks cloth is just as good as N95 (which is widely available in Los Angeles), which is clearly not the case.
"But is it actually true that the Moderna vaccine is “more effective,” or are they just giving you triple the dose? An ibuprofen gel tab is no more effective than a normal ibuprofen pill, but if you take 3x the dosage of the gel tab that will have a stronger effect."
This is a terrible analogy! It's like observing that pressing down on the gas peddle makes your car go faster, so logically, pouring gasoline all over the engine and lighting on fire will make it go even faster.
Ibuprofen circulates in the bloodstream. It binds receptors (on cyclo-oxygenase) antagonistically, meaning it is inhibitory. The higher the concentration of ibuprofen, the further the equilibrium shifts towards bound ligand-protein, i.e., the greater the effect. Fluvoxamine is an agonist towards a receptor that leads to the up-regulation of a an anti-inflammatory. In both cases, higher doses lead to higher concentrations in the bloodstream, which lead to greater (side)effects.
The Pfizer and Moderna vaccines are suspensions of lipid nanoparticles containing messenger RNA. They are injected locally in to muscle tissue specifically so that they do not enter the blood stream. Their efficacy is unrelated to binding constants; once a strand of m-RNA makes it into the cytoplasm, it is actively transported into the ribosome where its sequence is transcribed into a protein. That protein is then released, but since it is exogenous, the immune system identifies it as such, generates antibodies to attack it and then retains a record of the antibodies for later use. Higher doses may elicit stronger immune responses by producing proteins in more cells (with the upper-bounds being full-blown COVID-19 symptoms), but they might also cause necrosis.
Moreover, the Moderna and Pfizer vaccines encode different sequences, so they are not even producing the same proteins. It could just as easily be the case that higher Pfizer doses would lead to worse immune responses because they interact with antibodies differently from the Moderna-instantiated proteins.
It's easy to armchair quarterback vaccine development and testing with bad analogies, easier still to assume a profit motive behind Pfizer's vaccine schedule. But the reality is that these things are not arbitrary and people far more knowledgeable about mRNA vaccines, epidemiology and clinical trials than any of us were intimately involved in planning and executing the vaccine trials.
I'm not arguing against more testing, but I see stupid analogies like this everywhere, usually in the context of trying to make the CDC or FDA or scientists look lazy or stupid. Worse, this is the same faulty reasoning that is used to attack epidemiology; it is common sense that higher doses work better than lower doses, so clearly there is a conspiracy theory and/or these hacks with their fancy degrees do not know what they are doing.
I don't think Matt is denying that there could be a relevant difference between Moderna and Pfizer/BioNTech beyond just the dose size. But his point is that there have been precisely 0 studies discussed in any public arena investigating this. If Moderna has studies showing that 50 micrograms is less effective than 100, they should show it! Or even better if they have studies showing that it's equally effective, so that we get twice as many doses! If Pfizer or BioNTech have studies showing that their vaccine won't do any better at 100 micrograms, let them show it!
It's just one person's opinion, so I don't want to put much stock in the question of "right" or "wrong", but it at least shows a fraction of the many key decisions that had to be made with limited support. (as an aside: I was looking to see if anyone's done proteomics to check these questions but didn't find anything yet.)
I know your point is mainly about Matt's use of the analogy, but just to continue your metaphor, if not armchair QB, it's not so crazy to Monday-morning QB a bit and at least ask if there's still more room to optimize some of the decisions made previously, like dosing and spacing.
Hey, I am all for optimization and improvement. But let me play devil's advocate for a moment.
Matt has been a vocal critic of CDC policy that tries to engineer behavior, like saying not to wear masks to head off a run on surgical masks needed by medical professionals. Although it's pretty clear that was a dumb decision, if the CDC starts changing the spacing and/or size of doses, I could see two bad things happening.
First, it would take time for the policy to trickle down to pop-up vaccine clinics and such, so you'd have people on social media saying "hey, look over here, they're spacing my doses differently than yours", which will lead to a deluge of speculation and all the social media bullshit that entails. There is no way that does anything but further polarize people around the "trust the science" issue. Second, and worse, bad-faith opportunists would hold it up as proof that the CDC doesn't know what it is doing because the facile perception that "they would have gotten it right the first time if they actually knew what they were doing" is pervasive.
Now, let's say that the benefits of spacing the Pfizer doses at eight weeks instead of three is a 99% chance in the decrease in the rate of hospitalization; i.e., the effect is very significant. But let's say that the decrease is only 1%; i.e., the effect-size is rather small. At the global population level, that will make a huge difference, but it is trivial at the personal/local level.
If I were a policy-maker at the CDC, would I want to expend resources and risk credibility for a tiny effect on the rate of hospitalization? Or would I rather those resources go towards getting more people vaccinated? The Oxford study Matt linked to is a pre-print, meaning it hasn't even been peer-reviewed. And the Reuters article only discusses antibody levels, which is probably not a great metric for the CDC; they should be concerned about real-world outcomes, not updating their guidelines every time Reuters picks up a story about a preprint. (BTW that is the main reason the CDC trusts their own data more than anything else; they are measuring very specific outcomes in the American population/system.)
Fair points! But a lot hinges on where the improvement falls between 99% and 1%. And I'm not entirely convinced that their credibility is best protected by never deviating from any decision.
I fully agree. The CDC and FDA need some structural reforms to enable them to be more nimble during an emergency, particularly with respect to how they injest and react to emerging data from external sources. But I do think they have to be extremely careful not to seem capricious. It is a sad truth that one of the two major political parties has decided the CDC, FDA and science, in general, are political enemies. There is an entire media ecosystem that will pounce on and amplify any perceived inconsistency as proof of a secret, malicious agenda.
There are probably already snippets of Matt's post, taken out of context, circulating on right-wing social media circles as proof that even the libs think the CDC is incompetent and mendacious. My dad occasionally forwards that stuff to me to debunk when he thinks there is merit to the claims. They do a great job at cherry-picking from legitimate sources and using technical terms (incorrectly) to sound authoritative.
The pfizer vaccine did in fact show a dose-response between 10 and 20 ug doses, possibly leveling out at 30 ug. They did test 10, 20, and 30 ug doses in a phase I study. Fine, maybe the ibuprofen analogy oversimplifies a bit, but it’s basically true that the vaccines show a dose-response and it’s not unthinkable that the higher dose used in the Moderna vaccine could contribute to its higher efficacy. Also, the regulatory authorities are wrong about the dose-schedule and lots of people pointed this out in real time, so it’s not hindsight.
Yes, but that is what I am saying; you need to study a vaccine to establish a dose-reponse curve, not assume that one exists. Likewise, you cannot base assumptions on "vaccines" because there are many different types with many different methods of action. You cannot base the dose of an mRNA vaccine on that of an attenuated virus. That is why the ibuprofen analogy is just plain wrong. Post-hoc rationalizations based on bad analogies is not how science is done.
"Post-hoc rationalizations based on bad analogies is not how science is done."
No - but its often how science is explained to laypeople. How many classes in school did you take to learn the basics and then get to the higher level class where they tell you that while the previous explanation gave an overall idea of how things work, its actually all wrong in the details.
That is the nut of my beef—simple analogies to help students understand difficult concepts are one thing, but in the specific context of vaccines, they are proliferating on the Internet with high salience.
We are all familiar with the bad-faith version of this phenomenon. But the good-faith efforts to understand and to question—like this post—also contribute to anti-elitism and the denigration of institutions like the CDC that, I think, is ultimately not constructive.
Maybe. Here's the challenge I see: There are numerous things that generalists like Matt and others here/elsewhere think that the FDA/CDC/etc. could be doing better. Unfortunately, I see very little good faith response to those criticisms about why they are wrong, or why the above institutions are focused on other more important priorities. Instead, more often than not there is a retreat behind the curtain of expertise/credentialism. I would hope that most people are open to being informed by experts, but that actually requires experts explaining their actions (and possibly using their very own crummy analogies to do so).
I fully agree that they could have done better, could do better and should improve. But if we are going to criticize them, we need to do so thoughtfully. The reason that analogy irked me is that it is inaccurate and gives a sense that these are easy problems with common sense solutions. That thinking has taken root amongst bad-faith actors and, increasingly, good-faith skeptics and contrarians. The underlying issues are much more complex than “more ibuprofen works better than less” and using logically faulty analogies does nothing to help people understand that complexity. It also feeds the narrative that experts are full of it and any random idiot could do better… that was a successful presidential campaign and I fear that the same mentality is infecting the public perception of science.
Is there any information on how long between infection and booster confers the biggest benefit? I am recovering from mild COVID now, which I got before I was able to get a booster. I’m on the fence about whether to get a booster, but think I probably will. But if I do I’d like to time it for optimum protection. OTOH maybe I should just wait until there’s an omicron booster.
I don't think that there's great information out there. Probably the most relevant thing you can try to find is whether *recent* infection gives decent protection against omicron. We know that just "prior infection" is weak protection against omicron, but is that like "omicron is better at handling waning immunity from infection 6 months ago," or is it "omicron blows through protection from prior strains regardless of how recent"?
If the latter, a near-term boost might be valuable for you, but my guess is that waiting 2-3 months is probably better. This is based on lots of like not-very-direct data, so it may be badly wrong, and I am not a medical professional. Good luck, sorry you were sick, and hope you don't get sick again.
PS My family all tested rapid negative before seeing my elderly parents for t-giving. Our symptoms showed up the day after we left their house, and of course we passed to my parents (both boosted). This whole episode is upending my thinking on the possibilities of containing the virus.
If you read the SEC filings, earnings conference calls, etc. from Quidel, maker of the QuickVue rapid test and a smaller competitor to Abbott, it is absolutely clear that the limiting factor in Quidel rolling out more tests is production capacity. And on top of that, the US government itself has tied up a lot of Quidel's output with a large DoD contract to buy QuickVue tests. But nowhere is the government subsidizing increases to the production capacity.
Meanwhile, a company like Quidel is beholden to shareholders and cannot reasonably justify building out crisis-level capacity when it may not be needed in 2 or 5 years. At the same time, such capacity - if created with government help - could be an element of a government-led quick-response program for future pandemics.
The contracts specify a one-year period to supply the tests so Quidel needs to have at least that level of annual capacity and could obtain financing for that based on the DoD contract. That capacity would remain once the contract ends. The DoD contract is government help in creating capacity. Whether that capacity is shut down after the contract ends is a different matter
That is fair but it seems short-sighted. Because then the only way to expand the capacity is to tie up that capacity for the first year which is exactly when it's needed most.
I suspect that DoD will not be hoarding its tests to a large degree. if they don't need them for their own personnel, they will distribute them outside DoD, much like PPE and ventilators
Chickens always come home to roost. As an old guy I well remember when governments had a hard time finding manufacturers of vaccines at all. The margins were low, even the best produced rare serious and occasionally lethal side effects so the absolutely insane American tort laws made this a very risky business to do. The intelligent thing was done and a kind of insurance pool was created to assist people with the consequences of rare adverse side effects and legal recourse was limited. I note that this was universally excoriated on the left, as was all product liability tort reform. They wanted vaccines but also the power to sue the vaccine manufacturers out of existence.
I will add, as an interesting side point, that before Covid exploded into our world, the largest manufacturer of N95 masks in the United States, 3M, was successfully sued by exactly two coal miners for about thirty million dollars. The theory here being that defective masks resulted in their getting black lung. Do I believe this? Nope. But a jury did.
With regard to test kits for Covid one has to ask, why would any sane manufacturer rush to invest in making tests of dubious reliability (thereby creating a low margin high risk business model) when they are virtually certain to face class action suits related to false positives or false negatives. Is someone planning on addressing this issue? I can't find anyone doing so. And I am willing to bet that various members of the ABA are salivating at the prospects.
Great comment. Also a great example of how America's ancient, creaking constitutional order holds the country back. The US is the only developed country that still does civil trials with a jury deciding the verdict- every other country uses a judge. And why are we the only ones? The 7th Amendment:
"In Suits at common law, where the value in controversy shall exceed twenty dollars, the right of trial by jury shall be preserved, and no fact tried by a jury, shall be otherwise re-examined in any Court of the United States, than according to the rules of the common law."
And we also have the single most difficult-to-alter constitution of almost any democracy, regardless of wealth. Good luck fixing that one!
"So everyone is following procedure and doing things by the book."
Of course they are. Because if the school does one thing that hasn't been government approved and one child gets sick, they are going to get sued. Until the longer dosing regime is approved by the government, no school or other entity is even going to entertain the thought of doing it differently.
Even if they really, really wanted to go with the best evidence and change the dose timing, their lawyers would advise them not to and get signed statements saying this was our advise and they are acting against our advise.
The government healthcare community picked sides in the culture wars. Once they did that conservatives shut them out. The scientist need to say “I don’t do culture wars, I just do science. I follow facts wherever they lead”. Fauci should step down and Biden should appoint someone who takes this approach. That would be better for the country. Fair or not fair, Fauci is viewed as a partisan.
I think that also created a feedback loop. The CDC and Fauci started endorsing policies that weren’t supported by science (closing schools for over a year) but were the opposite of what conservatives were pushing for. It has not been good for science
I don't know if it's fair to Fauci, but that's a good point. I'm not on conservative media but I can easily imagine that "Everybody social distance unless you're protesting for social justice" was credibility killing for many on the right
The CDC seems to have a weird inverted pyramid of evidence where “any study conducted by the CDC” is at the top, “double blind study conducted by a fancy drug company” appears somewhere in the middle, and “anything out of Israel or Great Britain” is at the bottom. That’s how we end up with essentially all guidance determined by a couple observational studies of Provincetown Bear Week and some incident in a Marin county school while the high quality studies conducted by nationalised health care systems go seemingly ignored.
There really is something very concerning about how the CDC ranks the quality of evidence. Sometimes I wonder if they are now a marketing agency rather than a scientific organization
It is apparent that large factions of the CDC view themselves as belonging to a disease research organization, not an operational public health organization. There's nothing wrong with having a former, but we have witnessed the consequences of lacking the latter.
Only with concordant findings. If you have symptoms and you're positive - you can stop. If you are asymptomatic and test negative, you can stop. If you're asympomatic and positive you need a PCR. If you're symptomatic and negative, you need a PCR.
Note that a 91.7% accuracy rate (assuming same rate false positive/false negative) can still be misleading. If 1/100 people have Covid, and 100 people get tested, you'll have 7x as many false positives as real positives.
Still pretty good - and you can go get the PCR then, but better to use if you have some reason to believe your base rate is > 1/100 (you're feeling sick, or were exposed to someone with Covid)
If I tested positive I'd go get a PCR to confirm(I'd stay away from people in the meantime - but the meantime isn't very long in that case)
Accuracy rate is pretty meaningless. If a test is totally incapable of detecting COVID and you test 1000 people without COVID and 917 test negative while 83 have false positives, that's 91.7% accuracy. Or if 50/1000 have Covid and four of the 50 test positive and 913 of the other 950 test negative, you are 91.7% accurate but you only increase the positive predictive value from 5% to less than 8%
True - I was showing how even if it meant what you thought it meant (same test accuracy whether testing 1000 people without or 1000 people with) it could still be misleading, but as you say, it can be worse.
According on Nebraska's website:
Accuracy: The BinaxNOW test correctly gave a positive result 84.6% of the time compared to PCR. In the same study, the test correctly gave a negative result 98.5% of the time.
So 91.7 looks roughly like the (UNWEIGHTED) average of the two (almost exactly halfway between).
One point many people have emphasized in discussing false negatives is that usually the test isn't totally random in which positives come out as false negatives. There's some number of hours after you are infected before you have enough virus copies to test positive on PCR. There's a larger number of hours after you are infected before you have enough virus copies to test positive on the antigen test. It sounds like there's several more hours after that before you have enough virus copies to be highly infectious. So although a false negative is still a false negative, if you have people do the antigen test within a few hours before the event they are going to, even a false negative is likely to be non-infectious, or at least not very infectious.
For a lot of use cases, it's only false negatives when you're already at high infectiousness that matter.
Yes. the underlying assumption is a 50% pretest probability of infection which is arbitrary. Most people seeking testing will have a pretest probability that is much different than 50%
There's a really fascinating how-the-other-side-lives angle to this article, somewhat orthogonal to its main thrust.
Away from the Coastal Elites (TM), in my lefty, 85%-of-adults-vaccinated MSA, I've gotten one covid test ever (got violently sick from some bad chicken--practice safe refrigeration, folks!). Regular testing as policy, beyond definite exposure to someone who is confirmed positive, isn't a culture-war flashpoint--it's unheard-of, not discussed as even a theory. Aside from pretending masks matter when we're standing in the restaurant and not when we're sitting, the 'rona is over, and our death rate last month was about 1/100,000.
¯\_(ツ)_/¯
I'm not yet bought into the idea that there's a segment of the population that adopted the pandemic as a part of their identity and refuses to let it go, but I'm no longer able to casually dismiss that idea. The local paper sure has figured out what types of coronavirus get clicks, though. The vaccines "may" offer "somewhat less" protection against a new variant, and other technically-correct weasel language of that nature.
I took exactly one test voluntarily in extremely Blue Illinois after getting a runny nose the day after seeing a friend is a trainer (would have had huge impact on her business).
I have been tested a bunch of other times though since the pandemic is over for me and I've been traveling a fair amount and various governments are requiring tests to fly.
My wife is anxious by nature and doesn't feel right going through her day without something to be afraid of. The count of times she has become convinced she has COVID is well into double digits now. At-home tests have been great for us, they've reduced the consequences of psychosomatic COVID from "potential daylong freakout" to "10 minutes and some grumbling about test accuracy."
Having kids is where tests-in-every-home would help. Even wearing masks in school, they occasionally catch a cold and we've been going to get them tested. Seeing MY mention that antigen tests are easily obtained on Amazon, I picked some up (but most folks won't do that since they're $10/test).
I have free saliva-based tests at work any time I want one. I have taken precisely two since getting vaccinated in April (both negative). I hosted a large Thanksgiving gathering, which included my older parents, and no one got tested beforehand (everyone is vaccinated plus boosted). I agree that broader availability of cheaper (or even free) tests would have been *very* desirable a year ago, but at this point in the pandemic, I agree with Psaki that it's probably a waste of money.
The point Matt is making though is that the market price of tests would be super cheap if more were authorized, so it wouldn't cost the government anything at all.
Even among the Coastal Elites (I live within eyesight of a DC Metro station) I've been tested exactly once, and that was just because I was headed to a Thanksgiving event with a lot of old people.
At this point Covid affects my life in that I wear a mask on the Metro and in movie theaters, and that's about it.
We've used about 5 at home test kits at this point (might have used more but price and availability is definitely a factor)
3/5 cases:
Kid has sniffles. Almost positive it's not Covid, test anyway. Negative. When one of us then came down with the same thing we didn't bother to test - assumed still not Covid.
1/5 cases:
I came back from a 1500 person indoor convention where everyone signed an affidavit saying they were vaxxed/tested negative and I (other than some dining) always wore a mask and stayed around other masked people - but I was about to visit my family for thanksgiving so I tested myself after I got back. Not perfect timing - if I had caught it at the end of the convention I might not test positive until already at thanksgiving, but I was working with what I had (and everyone at thanksgiving who should be boosted was). Had I tested positive I was going to confirm with the more accurate PCR.
1/5 cases:
My husband felt a bit sick and tested himself as well - similar to kid sniffles.
The tests are tolerable but unpleasant.
If the tests magically dropped on my doorstep every week that's probably how we'd use them - testing it after exposure, or when I did something _particularly_ risky (fly, go to a convention) but testing weekly? Only if everyone else were doing it.
I curious where you live that the death rate is an order of magnitude lower than for the rest of the United States. According to Our World in Data, there were 34,240 COVID deaths Nov. 7 to Dec. 6. Dividing by the US population of 330 million, that's 10.4 deaths per 100,000.
The data on Fluvoxamine is quite weak. The primary outcome was a composite outcome of ED stay length >6hrs and hospitalizations. With a difference of 10.4% (Fluvoxamine) vs. 14.7% (placebo). As anyone who has ever been in the ED, length of stay can be quite long and can be quite arbitrary too based on hospital capacity, etc. More importantly, when you look closer at the data, hospitalization is 10% (fluvoxamine) vs 12.6% (placebo) with a p-value of 0.12, not statistically significant. I highly recommend the blog: https://rebelem.com/the-together-trial-covid-19-and-fluvoxamine-take-two/ from a very good ER researcher. Matt, please don't make fluvoxamine the liberal ivermectin. Both drugs have some promising data but are by no means strong enough to recommend.
Yes! And mortality and time-to-death are not changed at all with Fluvoxamine. That's right there in the text of the TOGETHER study. Matt, Vox, Astral Codex, etc. are a big echo chamber with none of them reading the primary sources beyond the headline. To throw shade on FDA based on such shallow "research" is really shameful. The official TOGETHER report is here: https://www.thelancet.com/journals/langlo/article/PIIS2214-109X(21)00448-4/fulltext . To opine on FDA's decisions on Fluvoxamine without reading the whole thing is just disgusting and dishonest.
"... 30-micrograms of Pfizer work but 100-micrograms of Moderna work better."
Better -- until one factors in the higher side-effect profile associated with the higher dosage.
If we're talking about what's ACTUALLY optimal -- absent supply shortages -- isn't two three-week-spaced shots, followed ~six months later by a booster the optimal thing over the entire timeframe? Which is, basically what everyone in the US got? I'm not sure what the "screwup" is that's being argued here.
Giving those two initial shots conferred significantly better protection than just one (especially against Delta) for those six months before booster-time came around.
Even in a supply-constrained situation where you only ever get two shots, there's a reasonable argument to be made that having six months of suboptimal single-shot protection isn't worth the tradeoff of better post-six-months protection. Especially when new treatments and therapeutics are coming online, and we can pretty reasonably expect that dangers six months from now will be less than they are now.
I learned a new word today! "Alacrity". Is this a common word I have just missed or is Matt flexing?
When I saw MY's stash of testing kits on twitter I felt a little chagrined that I only bought one kit when they had a big stack at my Walgreens. But I guess I was thinking it was smart to wait for the reimbursement rule to kick in before buying a bunch -- they are not exactly expensive, but each set of two is a decent bottle of wine cost and the wine is more enjoyable.
The challenge for the intelligent generalist pundit, like Matt, is to be read up enough on a subject and fast enough to be able to say something interesting and meaningful. The problem is that you don't know as much as the people on the inside and often don't really have the professional background to discern what is a screwup and what is driven by concrete obstacles.
For all I know, the FDA and the CDC have been and continue to screw up in unforgiveable ways (Scott Gottlieb's new book, "Uncontrolled Spread" is very good on this, though I note that he's a lot more unforgiving to the CDC than the FDA, which he had headed). But what I also know from professional experience is that it's often the things you don't know that are really meaningful and can refute what you think is the truth and the path forward. Life is complicated.
An example of this is congestion at the LA/Long Beach port. Bad problem, and people who know very little instantly opined on how to fix it. But as you dig deeper into the details, you realize that we're facing a series of bottlenecks, and each one needs to be resolved sequentially, after doing due diligence. It's usually never "one weird trick."
So Matt may be way off base here -- though none of us is expert or informed enough to know. That said, we need people like Matt out there asking interesting questions and trying to inform the public, even when he doesn't know enough to give us the full truth.
If the people on the inside are consistently incapable of explaining things to intelligent people like Matt then the fault is clearly with them.
The people on the inside have also been saying things that turned out to be wrong/lies all throughout the pandemic, making things even moreso their fault
Well, organizations should always do a better job explaining themselves, but I don't really want them to sit down with every Substacker with 13,000 subscribers to explain in detail what they're doing.
No offense to Matt, of course.
Also on masking. LA Public health is still handing out CLOTH masks at vaccine centers, with public health logos on them. This suggests that LA Public Health thinks cloth is just as good as N95 (which is widely available in Los Angeles), which is clearly not the case.
"But is it actually true that the Moderna vaccine is “more effective,” or are they just giving you triple the dose? An ibuprofen gel tab is no more effective than a normal ibuprofen pill, but if you take 3x the dosage of the gel tab that will have a stronger effect."
This is a terrible analogy! It's like observing that pressing down on the gas peddle makes your car go faster, so logically, pouring gasoline all over the engine and lighting on fire will make it go even faster.
Ibuprofen circulates in the bloodstream. It binds receptors (on cyclo-oxygenase) antagonistically, meaning it is inhibitory. The higher the concentration of ibuprofen, the further the equilibrium shifts towards bound ligand-protein, i.e., the greater the effect. Fluvoxamine is an agonist towards a receptor that leads to the up-regulation of a an anti-inflammatory. In both cases, higher doses lead to higher concentrations in the bloodstream, which lead to greater (side)effects.
The Pfizer and Moderna vaccines are suspensions of lipid nanoparticles containing messenger RNA. They are injected locally in to muscle tissue specifically so that they do not enter the blood stream. Their efficacy is unrelated to binding constants; once a strand of m-RNA makes it into the cytoplasm, it is actively transported into the ribosome where its sequence is transcribed into a protein. That protein is then released, but since it is exogenous, the immune system identifies it as such, generates antibodies to attack it and then retains a record of the antibodies for later use. Higher doses may elicit stronger immune responses by producing proteins in more cells (with the upper-bounds being full-blown COVID-19 symptoms), but they might also cause necrosis.
Moreover, the Moderna and Pfizer vaccines encode different sequences, so they are not even producing the same proteins. It could just as easily be the case that higher Pfizer doses would lead to worse immune responses because they interact with antibodies differently from the Moderna-instantiated proteins.
It's easy to armchair quarterback vaccine development and testing with bad analogies, easier still to assume a profit motive behind Pfizer's vaccine schedule. But the reality is that these things are not arbitrary and people far more knowledgeable about mRNA vaccines, epidemiology and clinical trials than any of us were intimately involved in planning and executing the vaccine trials.
I'm not arguing against more testing, but I see stupid analogies like this everywhere, usually in the context of trying to make the CDC or FDA or scientists look lazy or stupid. Worse, this is the same faulty reasoning that is used to attack epidemiology; it is common sense that higher doses work better than lower doses, so clearly there is a conspiracy theory and/or these hacks with their fancy degrees do not know what they are doing.
I don't think Matt is denying that there could be a relevant difference between Moderna and Pfizer/BioNTech beyond just the dose size. But his point is that there have been precisely 0 studies discussed in any public arena investigating this. If Moderna has studies showing that 50 micrograms is less effective than 100, they should show it! Or even better if they have studies showing that it's equally effective, so that we get twice as many doses! If Pfizer or BioNTech have studies showing that their vaccine won't do any better at 100 micrograms, let them show it!
I think it was Ken in MIA who posted this interesting paper last week (apologies if I misremembered): https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8310186/
It's just one person's opinion, so I don't want to put much stock in the question of "right" or "wrong", but it at least shows a fraction of the many key decisions that had to be made with limited support. (as an aside: I was looking to see if anyone's done proteomics to check these questions but didn't find anything yet.)
I know your point is mainly about Matt's use of the analogy, but just to continue your metaphor, if not armchair QB, it's not so crazy to Monday-morning QB a bit and at least ask if there's still more room to optimize some of the decisions made previously, like dosing and spacing.
Hey, I am all for optimization and improvement. But let me play devil's advocate for a moment.
Matt has been a vocal critic of CDC policy that tries to engineer behavior, like saying not to wear masks to head off a run on surgical masks needed by medical professionals. Although it's pretty clear that was a dumb decision, if the CDC starts changing the spacing and/or size of doses, I could see two bad things happening.
First, it would take time for the policy to trickle down to pop-up vaccine clinics and such, so you'd have people on social media saying "hey, look over here, they're spacing my doses differently than yours", which will lead to a deluge of speculation and all the social media bullshit that entails. There is no way that does anything but further polarize people around the "trust the science" issue. Second, and worse, bad-faith opportunists would hold it up as proof that the CDC doesn't know what it is doing because the facile perception that "they would have gotten it right the first time if they actually knew what they were doing" is pervasive.
Now, let's say that the benefits of spacing the Pfizer doses at eight weeks instead of three is a 99% chance in the decrease in the rate of hospitalization; i.e., the effect is very significant. But let's say that the decrease is only 1%; i.e., the effect-size is rather small. At the global population level, that will make a huge difference, but it is trivial at the personal/local level.
If I were a policy-maker at the CDC, would I want to expend resources and risk credibility for a tiny effect on the rate of hospitalization? Or would I rather those resources go towards getting more people vaccinated? The Oxford study Matt linked to is a pre-print, meaning it hasn't even been peer-reviewed. And the Reuters article only discusses antibody levels, which is probably not a great metric for the CDC; they should be concerned about real-world outcomes, not updating their guidelines every time Reuters picks up a story about a preprint. (BTW that is the main reason the CDC trusts their own data more than anything else; they are measuring very specific outcomes in the American population/system.)
Fair points! But a lot hinges on where the improvement falls between 99% and 1%. And I'm not entirely convinced that their credibility is best protected by never deviating from any decision.
I fully agree. The CDC and FDA need some structural reforms to enable them to be more nimble during an emergency, particularly with respect to how they injest and react to emerging data from external sources. But I do think they have to be extremely careful not to seem capricious. It is a sad truth that one of the two major political parties has decided the CDC, FDA and science, in general, are political enemies. There is an entire media ecosystem that will pounce on and amplify any perceived inconsistency as proof of a secret, malicious agenda.
There are probably already snippets of Matt's post, taken out of context, circulating on right-wing social media circles as proof that even the libs think the CDC is incompetent and mendacious. My dad occasionally forwards that stuff to me to debunk when he thinks there is merit to the claims. They do a great job at cherry-picking from legitimate sources and using technical terms (incorrectly) to sound authoritative.
The pfizer vaccine did in fact show a dose-response between 10 and 20 ug doses, possibly leveling out at 30 ug. They did test 10, 20, and 30 ug doses in a phase I study. Fine, maybe the ibuprofen analogy oversimplifies a bit, but it’s basically true that the vaccines show a dose-response and it’s not unthinkable that the higher dose used in the Moderna vaccine could contribute to its higher efficacy. Also, the regulatory authorities are wrong about the dose-schedule and lots of people pointed this out in real time, so it’s not hindsight.
Reference:
https://www.medrxiv.org/content/10.1101/2020.08.17.20176651v2.full-text
Yes, but that is what I am saying; you need to study a vaccine to establish a dose-reponse curve, not assume that one exists. Likewise, you cannot base assumptions on "vaccines" because there are many different types with many different methods of action. You cannot base the dose of an mRNA vaccine on that of an attenuated virus. That is why the ibuprofen analogy is just plain wrong. Post-hoc rationalizations based on bad analogies is not how science is done.
"Post-hoc rationalizations based on bad analogies is not how science is done."
No - but its often how science is explained to laypeople. How many classes in school did you take to learn the basics and then get to the higher level class where they tell you that while the previous explanation gave an overall idea of how things work, its actually all wrong in the details.
That is the nut of my beef—simple analogies to help students understand difficult concepts are one thing, but in the specific context of vaccines, they are proliferating on the Internet with high salience.
We are all familiar with the bad-faith version of this phenomenon. But the good-faith efforts to understand and to question—like this post—also contribute to anti-elitism and the denigration of institutions like the CDC that, I think, is ultimately not constructive.
Maybe. Here's the challenge I see: There are numerous things that generalists like Matt and others here/elsewhere think that the FDA/CDC/etc. could be doing better. Unfortunately, I see very little good faith response to those criticisms about why they are wrong, or why the above institutions are focused on other more important priorities. Instead, more often than not there is a retreat behind the curtain of expertise/credentialism. I would hope that most people are open to being informed by experts, but that actually requires experts explaining their actions (and possibly using their very own crummy analogies to do so).
I fully agree that they could have done better, could do better and should improve. But if we are going to criticize them, we need to do so thoughtfully. The reason that analogy irked me is that it is inaccurate and gives a sense that these are easy problems with common sense solutions. That thinking has taken root amongst bad-faith actors and, increasingly, good-faith skeptics and contrarians. The underlying issues are much more complex than “more ibuprofen works better than less” and using logically faulty analogies does nothing to help people understand that complexity. It also feeds the narrative that experts are full of it and any random idiot could do better… that was a successful presidential campaign and I fear that the same mentality is infecting the public perception of science.
Is there any information on how long between infection and booster confers the biggest benefit? I am recovering from mild COVID now, which I got before I was able to get a booster. I’m on the fence about whether to get a booster, but think I probably will. But if I do I’d like to time it for optimum protection. OTOH maybe I should just wait until there’s an omicron booster.
Also even mild COVID sucks. I don’t recommend it.
Eight weeks is probably a good guess if there is no new booster on the horizon at that point
I don't think that there's great information out there. Probably the most relevant thing you can try to find is whether *recent* infection gives decent protection against omicron. We know that just "prior infection" is weak protection against omicron, but is that like "omicron is better at handling waning immunity from infection 6 months ago," or is it "omicron blows through protection from prior strains regardless of how recent"?
If the latter, a near-term boost might be valuable for you, but my guess is that waiting 2-3 months is probably better. This is based on lots of like not-very-direct data, so it may be badly wrong, and I am not a medical professional. Good luck, sorry you were sick, and hope you don't get sick again.
PS My family all tested rapid negative before seeing my elderly parents for t-giving. Our symptoms showed up the day after we left their house, and of course we passed to my parents (both boosted). This whole episode is upending my thinking on the possibilities of containing the virus.
If you read the SEC filings, earnings conference calls, etc. from Quidel, maker of the QuickVue rapid test and a smaller competitor to Abbott, it is absolutely clear that the limiting factor in Quidel rolling out more tests is production capacity. And on top of that, the US government itself has tied up a lot of Quidel's output with a large DoD contract to buy QuickVue tests. But nowhere is the government subsidizing increases to the production capacity.
Meanwhile, a company like Quidel is beholden to shareholders and cannot reasonably justify building out crisis-level capacity when it may not be needed in 2 or 5 years. At the same time, such capacity - if created with government help - could be an element of a government-led quick-response program for future pandemics.
Here's some detail on the DoD contracts: https://www.nsmedicaldevices.com/news/four-firms-win-647m-contracts-from-us-dod-to-supply-covid-19-test-kits/
The contracts specify a one-year period to supply the tests so Quidel needs to have at least that level of annual capacity and could obtain financing for that based on the DoD contract. That capacity would remain once the contract ends. The DoD contract is government help in creating capacity. Whether that capacity is shut down after the contract ends is a different matter
That is fair but it seems short-sighted. Because then the only way to expand the capacity is to tie up that capacity for the first year which is exactly when it's needed most.
I suspect that DoD will not be hoarding its tests to a large degree. if they don't need them for their own personnel, they will distribute them outside DoD, much like PPE and ventilators
It's SOP for the military to hoard stuff
Chickens always come home to roost. As an old guy I well remember when governments had a hard time finding manufacturers of vaccines at all. The margins were low, even the best produced rare serious and occasionally lethal side effects so the absolutely insane American tort laws made this a very risky business to do. The intelligent thing was done and a kind of insurance pool was created to assist people with the consequences of rare adverse side effects and legal recourse was limited. I note that this was universally excoriated on the left, as was all product liability tort reform. They wanted vaccines but also the power to sue the vaccine manufacturers out of existence.
I will add, as an interesting side point, that before Covid exploded into our world, the largest manufacturer of N95 masks in the United States, 3M, was successfully sued by exactly two coal miners for about thirty million dollars. The theory here being that defective masks resulted in their getting black lung. Do I believe this? Nope. But a jury did.
With regard to test kits for Covid one has to ask, why would any sane manufacturer rush to invest in making tests of dubious reliability (thereby creating a low margin high risk business model) when they are virtually certain to face class action suits related to false positives or false negatives. Is someone planning on addressing this issue? I can't find anyone doing so. And I am willing to bet that various members of the ABA are salivating at the prospects.
Great comment. Also a great example of how America's ancient, creaking constitutional order holds the country back. The US is the only developed country that still does civil trials with a jury deciding the verdict- every other country uses a judge. And why are we the only ones? The 7th Amendment:
"In Suits at common law, where the value in controversy shall exceed twenty dollars, the right of trial by jury shall be preserved, and no fact tried by a jury, shall be otherwise re-examined in any Court of the United States, than according to the rules of the common law."
And we also have the single most difficult-to-alter constitution of almost any democracy, regardless of wealth. Good luck fixing that one!
"So everyone is following procedure and doing things by the book."
Of course they are. Because if the school does one thing that hasn't been government approved and one child gets sick, they are going to get sued. Until the longer dosing regime is approved by the government, no school or other entity is even going to entertain the thought of doing it differently.
Even if they really, really wanted to go with the best evidence and change the dose timing, their lawyers would advise them not to and get signed statements saying this was our advise and they are acting against our advise.
It is a giant CYA.
The government healthcare community picked sides in the culture wars. Once they did that conservatives shut them out. The scientist need to say “I don’t do culture wars, I just do science. I follow facts wherever they lead”. Fauci should step down and Biden should appoint someone who takes this approach. That would be better for the country. Fair or not fair, Fauci is viewed as a partisan.
I think that also created a feedback loop. The CDC and Fauci started endorsing policies that weren’t supported by science (closing schools for over a year) but were the opposite of what conservatives were pushing for. It has not been good for science
I don't know if it's fair to Fauci, but that's a good point. I'm not on conservative media but I can easily imagine that "Everybody social distance unless you're protesting for social justice" was credibility killing for many on the right
The CDC seems to have a weird inverted pyramid of evidence where “any study conducted by the CDC” is at the top, “double blind study conducted by a fancy drug company” appears somewhere in the middle, and “anything out of Israel or Great Britain” is at the bottom. That’s how we end up with essentially all guidance determined by a couple observational studies of Provincetown Bear Week and some incident in a Marin county school while the high quality studies conducted by nationalised health care systems go seemingly ignored.
There really is something very concerning about how the CDC ranks the quality of evidence. Sometimes I wonder if they are now a marketing agency rather than a scientific organization
It is apparent that large factions of the CDC view themselves as belonging to a disease research organization, not an operational public health organization. There's nothing wrong with having a former, but we have witnessed the consequences of lacking the latter.
Are the rapid (antigen) tests accurate enough to rely on for at-home testing?
Only with concordant findings. If you have symptoms and you're positive - you can stop. If you are asymptomatic and test negative, you can stop. If you're asympomatic and positive you need a PCR. If you're symptomatic and negative, you need a PCR.
The Binax ones say they're 91.7% accurate, which sounds more than good enough to me
Note that a 91.7% accuracy rate (assuming same rate false positive/false negative) can still be misleading. If 1/100 people have Covid, and 100 people get tested, you'll have 7x as many false positives as real positives.
Still pretty good - and you can go get the PCR then, but better to use if you have some reason to believe your base rate is > 1/100 (you're feeling sick, or were exposed to someone with Covid)
If I tested positive I'd go get a PCR to confirm(I'd stay away from people in the meantime - but the meantime isn't very long in that case)
Accuracy rate is pretty meaningless. If a test is totally incapable of detecting COVID and you test 1000 people without COVID and 917 test negative while 83 have false positives, that's 91.7% accuracy. Or if 50/1000 have Covid and four of the 50 test positive and 913 of the other 950 test negative, you are 91.7% accurate but you only increase the positive predictive value from 5% to less than 8%
True - I was showing how even if it meant what you thought it meant (same test accuracy whether testing 1000 people without or 1000 people with) it could still be misleading, but as you say, it can be worse.
According on Nebraska's website:
Accuracy: The BinaxNOW test correctly gave a positive result 84.6% of the time compared to PCR. In the same study, the test correctly gave a negative result 98.5% of the time.
So 91.7 looks roughly like the (UNWEIGHTED) average of the two (almost exactly halfway between).
One point many people have emphasized in discussing false negatives is that usually the test isn't totally random in which positives come out as false negatives. There's some number of hours after you are infected before you have enough virus copies to test positive on PCR. There's a larger number of hours after you are infected before you have enough virus copies to test positive on the antigen test. It sounds like there's several more hours after that before you have enough virus copies to be highly infectious. So although a false negative is still a false negative, if you have people do the antigen test within a few hours before the event they are going to, even a false negative is likely to be non-infectious, or at least not very infectious.
For a lot of use cases, it's only false negatives when you're already at high infectiousness that matter.
Yes. the underlying assumption is a 50% pretest probability of infection which is arbitrary. Most people seeking testing will have a pretest probability that is much different than 50%
Presumably you mean 4 of the 50 positives test negative? 4 of the 50 positives testing positive seems substantially worse than 91.7% accurate.
No. the accuracy is the total number of correct tests out of 1000. 913+4=917/1000=91.7%
There's a really fascinating how-the-other-side-lives angle to this article, somewhat orthogonal to its main thrust.
Away from the Coastal Elites (TM), in my lefty, 85%-of-adults-vaccinated MSA, I've gotten one covid test ever (got violently sick from some bad chicken--practice safe refrigeration, folks!). Regular testing as policy, beyond definite exposure to someone who is confirmed positive, isn't a culture-war flashpoint--it's unheard-of, not discussed as even a theory. Aside from pretending masks matter when we're standing in the restaurant and not when we're sitting, the 'rona is over, and our death rate last month was about 1/100,000.
¯\_(ツ)_/¯
I'm not yet bought into the idea that there's a segment of the population that adopted the pandemic as a part of their identity and refuses to let it go, but I'm no longer able to casually dismiss that idea. The local paper sure has figured out what types of coronavirus get clicks, though. The vaccines "may" offer "somewhat less" protection against a new variant, and other technically-correct weasel language of that nature.
I took exactly one test voluntarily in extremely Blue Illinois after getting a runny nose the day after seeing a friend is a trainer (would have had huge impact on her business).
I have been tested a bunch of other times though since the pandemic is over for me and I've been traveling a fair amount and various governments are requiring tests to fly.
Ooops - make that voluntarily twice. I took the Abbott test that time I caught COVID attending a wedding after being double vaxxed.
My wife is anxious by nature and doesn't feel right going through her day without something to be afraid of. The count of times she has become convinced she has COVID is well into double digits now. At-home tests have been great for us, they've reduced the consequences of psychosomatic COVID from "potential daylong freakout" to "10 minutes and some grumbling about test accuracy."
Having kids is where tests-in-every-home would help. Even wearing masks in school, they occasionally catch a cold and we've been going to get them tested. Seeing MY mention that antigen tests are easily obtained on Amazon, I picked some up (but most folks won't do that since they're $10/test).
I have free saliva-based tests at work any time I want one. I have taken precisely two since getting vaccinated in April (both negative). I hosted a large Thanksgiving gathering, which included my older parents, and no one got tested beforehand (everyone is vaccinated plus boosted). I agree that broader availability of cheaper (or even free) tests would have been *very* desirable a year ago, but at this point in the pandemic, I agree with Psaki that it's probably a waste of money.
The point Matt is making though is that the market price of tests would be super cheap if more were authorized, so it wouldn't cost the government anything at all.
Even among the Coastal Elites (I live within eyesight of a DC Metro station) I've been tested exactly once, and that was just because I was headed to a Thanksgiving event with a lot of old people.
At this point Covid affects my life in that I wear a mask on the Metro and in movie theaters, and that's about it.
Also, to be clear, there's approximately zero enforcement regarding masks on the Metro and in movie theaters.
We've used about 5 at home test kits at this point (might have used more but price and availability is definitely a factor)
3/5 cases:
Kid has sniffles. Almost positive it's not Covid, test anyway. Negative. When one of us then came down with the same thing we didn't bother to test - assumed still not Covid.
1/5 cases:
I came back from a 1500 person indoor convention where everyone signed an affidavit saying they were vaxxed/tested negative and I (other than some dining) always wore a mask and stayed around other masked people - but I was about to visit my family for thanksgiving so I tested myself after I got back. Not perfect timing - if I had caught it at the end of the convention I might not test positive until already at thanksgiving, but I was working with what I had (and everyone at thanksgiving who should be boosted was). Had I tested positive I was going to confirm with the more accurate PCR.
1/5 cases:
My husband felt a bit sick and tested himself as well - similar to kid sniffles.
The tests are tolerable but unpleasant.
If the tests magically dropped on my doorstep every week that's probably how we'd use them - testing it after exposure, or when I did something _particularly_ risky (fly, go to a convention) but testing weekly? Only if everyone else were doing it.
I curious where you live that the death rate is an order of magnitude lower than for the rest of the United States. According to Our World in Data, there were 34,240 COVID deaths Nov. 7 to Dec. 6. Dividing by the US population of 330 million, that's 10.4 deaths per 100,000.
I live in the land of not-enough-coffee-to-comment-yet.
With reporting delays (related: WTF?), Dane County, WI is up to 11 November deaths on a population about 560,000. Last I'd looked, I think it was 7.
https://publichealthmdc.com/coronavirus/dashboard
We have to divide the US into two countries: one vaccinated (Dane County being a rough approximation) and one unvaccinated.