I generally like this article but I think "But current vaccines do not stop you from transmitting Delta to others" is problematic enough that you should IMO change it.
1) The study in Eric's tweet is about AZ which is not available in the US and there seems to be good evidence that it's less effective against Delta than the mRNA vaccines.
2) You probably already know this and it's mostly hedging against anti-vaxxers (and potentially compatible with your sentence) but the vaccines reduce your chance of being infected in the first place so that reduces transmission.
3) There's decent emerging evidence that infected vaccinated people have less infectious virus (even though viral loads may be high) and that the period during which they are infectious is shorter.
I know you don't want to hear it but you should have left this question to experts. You and I absolutely don't know whether it would be medically useful or not. I suppose it depends on the minutiae of how differently the variants behave.
I am skeptical as I have not seen evidence of severe disease among the vaxed except among severely immunocompromised. If boosters are not medically useful and are merely a political project to make the vaxed feel better during an epidemic among the vaxed, that would be a waste and distraction from the real problem of getting more people vaxed.
But I'd pay attention to an expert making this recommendation. There just isn't any value in you weighing in here.
Very much, we have seen that the drop in any severe COVID in vaxxinated persons in UK. Only 1% of fully vaccinated people in UK make up COVID Delta deaths. Most were infected before second jab came into effect. Only 0.5% (256 people) were those who had full immunity. It shows that of those deaths, 76% were among people already deemed clinically vulnerable. So 25% of 0.5% is frankly, error margin under any other circumstance. At that point we're talking a grand total of 64 people. A sample size not good enough for even a pilot study.
Before we worry about boosters, let's prevent the spread of new variants by vaccinating the poorer nations.
This assumes we only care about deaths among the vaccinated. If we also care about deaths among the unvaccinated (which I grant you, is becoming increasingly difficult to do with every passing day), then this isn’t the right metric. And if we care about people not getting sick, it’s also not the right metric. I know quite a few people who had breakthrough cases. They were all completely fine in a few days thanks to the vaccine, but if I can get a booster to help me avoid what they went through, why not?
Also Noah specifically gives reasons to reject the idea that vaccinating the poorer nations needs to conflict at all with booster shots. You don’t have to disagree with him, but it’s a little strange to repeat the point he already addressed in the article without acknowledging you’re doing that.
How medically useful boosters would be is a separate issue. I think it's a fair summary of expert consensus that boosters will probably be useful soon for the most vulnerable segment of the population, and it's likely they'll be useful for increasingly broad segments as time goes on.
This article is arguing we need to develop variant-specific boosters. That's a different question, and one that I'm saying only virologists and especially those studying Covid variants could usefully comment on. And honestly my guess is that most of them would say they don't know yet.
If you don’t actually know what you are talking about, then maybe you should stay in your damn lane.
It is really easy to make random guesses that the consensus opinion is wrong, and occasionally by chance you are going to be right. Early in the pandemic everyone and their dog at some point wondered if masks would in fact help. But it wasn’t the endless tweets by armchair epidemiologists that discovered the problem, it was a bunch of actual scientists that did the work.
Delta specific covid shots seems like a good idea. It probably seems like a good idea to the army of people who have probably been working on the problem since delta emerged. You even link to a report that says that Pfizer is working on the problem.
How hard is it to understand that building a new vaccine, testing it, and making sure that it is safe and effective is an insanely difficult problem that takes a staggering amount of time, resources, work, and money? Do you actually think we can just tweak the sequence, and get the new dose into arms by the end of the week?
Are you actually implying that the constraints on vaccine production are due to artificial political scarcity, and not because we are running up against real physical limits?
What I find infuriating is your assumption that the roadblocks are political, rather than practical, and therefore the CDC et al must be incompetent or corrupt or playing larger political games. You have no evidence of any of that.
You might want one round of that for some basic trials first to double-check that nothing weird happened, before making the new version the main version
“How hard is it to understand that building a new vaccine, testing it, and making sure that it is safe and effective is an insanely difficult problem that takes a staggering amount of time, resources, work, and money?”
How is this relevant? We’ve already done all of that. We’re talking about tweaking it slightly to match up with the current variant. The whole POINT of mRNA vaccines is that it is very easy to do this.
And yes, it was “a bunch of actual scientists that did the work” on masks…after a bunch of actual scientists did the work of lying to people by claiming masks aren’t important because they were scared we’d take masks away from medical professionals. This is the worst possible example to choose to prove your point.
There’s science, and there’s policy. People like Noah shouldn’t be in the lab working on the vaccines. But they absolutely can and should be involved in the conversation about what our policies and priorities should be. We all should.
If there are technical issues with what Noah proposes, than that’s a great topic for conversation. If the only issue is “I don’t like who it is who is saying it,” that’s not useful or productive.
Moderna's vaccine was created and delivered to NIH for testing between Jan 11th and Feb 24th, which is what - 44 days? And they're saying future mRNA vaccines will be even faster. It's actually totally feasible that they could sequence a variant and get shots into arms of the first trial group within a week.
People don't complain about Nate Silver, Alex Taborek, and others because they 'don't stay in their lane', but because they can be insufferable arrogant at times. They have posted half baked critiques, created straw men, and not acknowledged errors. They also claim as their own brilliant insights, facts and ideas that have been well known by other experts.
Does anyone really believe that vaccine developers aren't attempting to improve the vaccine to tackle variants, as has been the standard practice for flu vaccines for years, as Alex T implies?
Has Alex T considered that the problem is that the challenges are greater than what he is aware of, instead of the researchers and public health people having only half his intellect?
“People don't complain about Nate Silver, Alex Taborek, and others because they 'don't stay in their lane', but because they can be insufferable arrogant at times. They have posted half baked critiques, created straw men, and not acknowledged errors. They also claim as their own brilliant insights, facts and ideas that have been well known by other experts.”
I just want to double emphasize this because it’s exactly right. The problem with Silver isn’t that he isn’t an expert; the problem with Silver is that he’s an asshole who is more interested in scoring cheap Twitter points than in making the world a better place.
These are basically guys who gained a following being really good at creating and understanding models. Eventually they decided that they were so super smart that they didn't need to create models. Eventually people just get lazy.
If had bothered to look at his implied risk model, then he would have spotted the flaw in his reasoning. For many people the personal cost of wearing a mask in public is smaller than the expected cost of getting covid and/or spreading it to others - even if they are vaccinated.
During the pandemic it was rational for them to wear a mask. Even after getting vaccinated, the expected risk is still higher than the cost of wearing a mask, so it is still rational for them to wear a mask.
Tyler is aware of how insanely difficult it is to fit economic models to the real world, and how hard it can be to tease out subtle effects. For some reason, however, he seems to think that evaluating the safety and effectiveness of vaccines can be done in Excel in an afternoon.
I don't read Tabarrok enough to follow all his idiocy, but a random sample is:
He doesn't understand why "people got fixated on the specific trial design of two-doses, three-to-four weeks apart". I don't understand why he doesn't understand that you can't just change the conditions of a trial, and assume you are going to get completely predicatable results. To change the dosage and scheduling, they would have needed to run another phase three trial, and prove that the new dosage had better results than the original dosage.
Re your first link, did Silver say anywhere that he was talking about masks? I think he would agree that people who aren't annoyed by masks could be rational to continue to use them, and their behavior changes should be made elsewhere by changing the behaviors that *are* costly to them.
Re Tabarrok, I think you and he just disagree about which of the FDA's procedures are grounded in scientific fact as opposed to bureaucratic inertia. Other countries like Canada have done things like altering dosage and scheduling, without bad results, because their take on the science was less legalistic than the FDA's.
I am reading Nate as saying that because the vaccine makes you safer, then you are irrational for not engaging in riskier behavior.
I choose to drive carefully when driving an old rustbucket, because I don’t want to get into an accident. If you do get into an accident, it is much safer when you are driving a big suv with airbags and crumple zones than the old ruskbucket.
It is perfectly valid for me to choose to drive carefully in the suv, even though my risk of death is now lower. You can’t say it is irrational to continue to drive carefully.
With Tabarrok, he is indeed assuming a priori that the FDA is acting slower than necessary. I think his base assumption is that all government is, by definition, always overly bureaucratic. He provides no evidence of this.
But he is going further than just saying that the FDA is overly bureaucratic. He is actively asking why we don’t just skip phase 3.
I don't think we should skip phase three. The government simply shouldn't *ban* anyone from buying it before phase three is completed. Second, the government shouldn't ban people from volunteering to be exposed to the disease, so that months can be shaved off the time needed for a trial.
Not only does figuring out if a half dose works as well as a full dose take a phase three trial, it takes a much much larger phase three trial than he original phase three trial.
I have zero problems with the FDA trying to stop people from injecting themselves with unproved treatments that the learned about on the internet.
I understand that there are a lot of different opinions about challenge trials, and this is probably something we need to work out sooner rather than later. I think it is pretty likely there will be other pandemics this decade, and this all might just be a trial run for what comes next.
But in this particular case, it is not clear that challenge trials would have actually saved us any time. A big percentage of covid cases are asymptomatic and unreported, so we still don’t have a solid idea of the true infection rate. That makes it really hard to use challenge trial data to infer efficacy rates, so we would need to do phase three trials anyway.
That said, I would love to have been able to get my greedy little paws on challenge trial data. Knowing exactly what percentage of people infected are asymptomatic, symptomatic, and are hospitalized would make modelling so much easier. I just don’t think the data it is worth the human cost.
It is certainly *valid* to choose to drive carefully in the SUV. But, given that a little bit of extra precaution provides more benefit in the rustbucket than in the SUV, it seems unlikely that someone who is both interested in safety and interested in reaching destinations would drive *precisely* as carefully in the SUV as in the rustbucket. Different balances of interest in safety and in reaching destinations will yield different behaviors in the same vehicle, and if you are driving more cautiously than me in the same vehicle, it likely just reflects our different values. But if we switch to a vehicle where caution has less of an effect on safety, and you still exercise just as much caution, then it seems that safety mustn't have been playing a role in your decision.
I don't know whether Tabarrok believes that all government is always overly bureaucratic. I think the question here is whether the American CDC reached the right level of caution, or whether the Canadian agencies did (who were willing to deviate from the protocol that was tested, because the evidence suggested that these deviations would likely save more lives).
I completely agree that most people would change driving habits in the SUV. But it would be silly to call someone “irrational” if they chose to drive with the same amount of care. Why would we want to insult people for insisting on driving safely?
I haven’t actually been following what Health Canada has been doing (bad Canadian!), but this still seems like a bloody dangerous thing to have done. I understand they had an initial vaccine shortage and they were desperate, but the gamble could have easily not paid off.
I’m just glad the US had enough vaccine and capacity that we weren’t forced to do something that desperate.
I have a really hard time believing you would not drive extremely cautiously and avoid trips in a rust bucket, and drive very carefully in an SUV. Which is what Nate is saying.
I wonder if Alex understands why we don't have a vaccine against RSV. It turned out in studies it made the disease worse by some antibody dependent enhancement.
Rob, I read Nate’s comment differently. He’s not saying go wild, he’s saying you should be willing to take more risks on the simple logic of the costs of risky behavior have gone down. Before you could be infected or infect others. Now you can only infect others. Since that’s a lower burden to society, shouldn’t your behavior change?
And as far as wearing a mask inside the store because others don’t feel comfortable, I always looked at it differently. If I see a person without a mask, I assume that’s a sign that are vaccinated and they are signaling to others the extent of the ability to get back to normal if you do get vaxxed!
Alex does not appear to have any appreciation of the challenges of drug development. I call his approaches the Star Trek theory of pharmaceutical development, that people magically would develop the perfect cure in one hour and everyone gets better, buy the bad FDA makes them fill out paperwork.
The trial conditions were arbitrarily chosen. If spreading doses further apart, or going to half doses leads to far more people being vaccinated, then we should use our ability to reason to figure out which is likely to lead to more saved lives. We should not insist on doing all new, time consuming trials, *WHILE BANNING ANYONE WHO WANTS TO TAKE THE RISK FROM BUYING IT*, when the cost is tens of thousands dead.
The trial conditions weren’t arbitrarily chosen. They were the best guess at what would be effective based on what they knew at the time. They designed the phase 3 trials based on their best guess, and got pretty rapid FDA approval once the trials were completed.
Suppose, based on what we know now, we decided that half doses could work. Maybe it would be just as effective, or maybe it would be a lot less effective. Maybe there is a subpopulation where half doses provide no protection. Or maybe there is a completely weird side effect that we couldn’t possibly predict, given that we have barely any experience with mRNA vaccines.
So we can’t just roll out half doses. We need to test this first. We need to find thousands of people to give half doses to, and monitor the outcomes. We then need to run down every case where something unexpected happened, and figure out if it was due to the vaccine or an unrelated issue. That means we need to do a complete phase three trial for this.
Now it actually gets hard. The initial phase three trials showed that the full dosage was effective, and saved lives compared to not having a vaccine. That’s why they were able to get Emergency Use Authorization so quickly.
But we need to compare the half dose regimen to the full dose regimen. Even with millions of doses we still don’t have solid efficacy numbers on the full doses. It would require A LOT more data to determine if a half dose is just as effective.
Even if we did all that, then we would probably find that a half dose is almost as effective. But is it ethical to give patients a half dose, when we know that a half dose provides less protection? How many people should die on a half dose, that wouldn’t have died on a full dose, is acceptable? Is it ethical to give substandard treatment to some, in order to save other lives? Are these the kind of questions we want the FDA to make?
Is it ethical to even do a phase 3 trials of halve dose regimen, when we don’t even know it will work, when we know full doses work, when we know how long the whole process needs to take?
Maybe it would have been worth it. I’m just damn glad that I don’t have to be the person making these insane hard calls. I have no idea.
But I am pretty damn sure that Tabarrok hasn’t thought this through, and is just talking out his ass.
> is it ethical to give patients a half dose, when we know that a half dose provides less protection?
If it's more than half as much protection, and we know we can get it to twice as many people, then yes.
> How many people should die on a half dose, that wouldn’t have died on a full dose, is acceptable?
As long as it's fewer than the people who die on zero doses, that wouldn't have died on the half dose they would be able to receive.
> Is it ethical to give substandard treatment to some, in order to save other lives?
Yes. That is the essence of triage. Actual bioethicists work out the details of how precisely and when it should be done, but in many conditions, particularly emergencies, medical professionals have plenty of standardized procedures of when to deny care to some so that others can have it, or when to give only partial care, and how precisely to prioritize.
> Are these the kind of questions we want the FDA to make?
Probably not the FDA itself - they should just be determining which things are safe enough and effective enough that they are permissible, and some other body should be determining actual implementation. (For infectious diseases, this might naturally be the CDC, or the National Institute for Allergy and Infectious Disease, or something else - though in practice it probably usually ends up being insurance companies.)
President Biden got the Pfizer shot very early. He is almost 79 year old and is exposed to a lot of people. It is only a matter of time before Delta meets him and he is hardly protected at that phase. Deadly breakthrough cases at that age after 8-9 months are not rare at all as you could see in Israel before the booster campaign. Prime minister Bennet even said that those old people who got two shots but didn't get the booster shot are most at risk since they have the illusion of immunity. It is a matter of national security to start booster shots ASAP and to vaccinate the president first.
Regarding variant based shots, their approval and manufacturing will take some months. We need boosters now.
Agreed, especially because we are a month or two from the next winter wave. I'm more uncertain about exactly how much immunity is waning than the Israeli authorities are, but in that age group there is only upside to a third shot. Most people over 65 should probably lie and get a third shot if they aren't approved by mid-October; this is the advice I'm giving to my loved ones, anyway.
Pfizer's own positions are that any reduced effectiveness is due to waning immunity over time rather than lesser effectiveness against the Delta variant, and that Delta-specific boosters "almost certainly" won't be necessary:
I think the variant-specific boosters would be put into play if the vaccines had significantly reduced effectiveness against a dominant variant, but the people who would actually be marketing these boosters don't think that's the case here. Delta is more contagious across the board than original coronavirus, but that's a different issue.
You have a misunderstand of vaccine production complexities. Simply throwing more resources at it isn't a problem. The problem is spool-up time and expertise base - vaccine manufaturing is peak high-technology manufacturing. It requires very good specialists, perfect quality assurance, clean rooms, etc.
A major missing piece is not being able to rapidly trial for safety and efficacy. Any country with the means to do it could have started this already, coordinating a roll out in order to best learn from it. Recommending to press ahead with boosters (variant specific or otherwise) without a more experimental approach is like advancing oblivious to leaving enemy positions at your back. Same problem can be seen with childhood vaccination issues. It's a recurring problem of lacking the political will and organisation to move forward both rapidly and cautiously.
TLDR, Noah how should we roll out boosters while also learning if they work and are safe, in a short time scale?
On this count, we need to think about things in a Bayesian way. If the original shot is safe and effective, a slight variation on that shot is basically certain to be safe and overwhelmingly likely to be effective.
Unfortunately it would not be safe to assume that. Knowledge of biology is highly incomplete. And a single molecular difference can change something from lethal to harmless. If the protein the immune response that is stimulated targets is slightly wrong it may cause serious autoimmune disease.
Good question! They are live attenuated or inactive virus AFAIK. That would limit the possibilities for what is produced. Don't know we can be so confident with viral vector and mrna. There is also a good deal of monitoring of safety as they roll out in various countries seasons. I imagine there were a good deal of nerves the first time they updated without long term safety data. Now people are mostly quite comfortable as it is essentially tradition.
The modified mRNA is just copying whatever modification the virus made to its own mRNA, so it should be limited to exactly the same changes in effect that the virus itself has.
That said, it probably is worth doing some extra monitoring for the first couple updates before just rushing ahead, though it's probably *not* worth doing the same 30,000 participant trial waiting for 186 cases among the trial participants.
That seems sensible and close to the position FDA is signalling. Immunogenicity and safety in a smaller cohort. Like maybe a few thousand. Then enhanced surveillance as it goes out to more people. That would presumably add 2-3 months to any updated vaccine development. At least for the first few times.
I'm not sure the Delta mutations necessarily introduce new targetable epitopes, or that it's clear what these might be. It's like telling someone to target someone wearing an invisibility cloak. Sure, the cloak is unique but it's not easy to target.
The breakthrough we need now is in scaling up manufacturing while we figure out how to fine tune using the mRNA tech.
I think Delta is unique b/c of its speed, I am not sure just configuring it to search for a different protein wins this race, we need to take a different approach to counter the speed of Delta, I think we need a vaccine targeted at nasal and lung tissues, like a spray vaccine.
The idea is sound and we certainly should be working on it. However, two points should be kept in mind: First, the current vaccines offer a beautiful "template" allowing the same strategy for variants. However, without (at least) years of empirical experience, it would be dangerous to actually release one of these novel vaccines without substantial testing to verify efficacy and safety.
Second, the current vaccines do a great job of activating the B cell/IgG antibody system. However, they do a poor job of stimulating IgA, the primary immune molecule in the upper respiratory tract. That likely explains why they're more effective at preventing death than preventing infection or reducing viral load. Accordingly (though the point is consistent with your essay) it remains important to design new vaccine delivery systems, perhaps inhaled or subcutaneous mRNA.
One aspect to account for is supply chain issues. You will need separate production and distribution lines. For a company, it's easier to keep making loads of the same, rather than custom versions.
Excellent point. However in the US virtually all cases are no Delta (I assume this is similar in most places at this point) and (more speculatively) due to this most variants are likely to derive from Delta so they could just switch over completely for now. It may get more complicated in the future.
The pharma community used insanely advanced technology to create covid vaccines in under 18 months.
Any sufficiently advanced technology is indistinguishable from magic.
Magic in fantasy novels is wielded by either good wizards or evil wizards.
Good wizards in fantasy novels give the protagonist everything they want: advice, support, and the occasional magic sword.
Everyone thinks they are the protagonist.
Therefore, if the pharma community isn’t giving you exactly what you want, they must be evil. They have powerful magic and can do anything. Anything bad that happens must be deliberate on their part.
It did not take us 18 months to make the vaccine. It took Moderna a weekend. The murderous gang at the FDA succeeded only in delaying medicine going to the public. Calling out their delays ought not be shamed.
Many of those delays were just the time needed to get manufacturing capacity running at scale. The FDA not approving AstraZeneca, and perhaps their not approving Novavax, are murderousness. But the delays on Moderna and Pfizer has had basically no effect on the number of vaccinated people since April 2021.
The vaccine companies respond to incentives. If rushing to get things together costs more money, why should they rush, when they know they won’t be allowed to sell it for months? Obviously, yes, there were hurdles to overcome - we could not get things running immediately - but if they had had the ability to sell it months earlier, *then* they would have had the incentive to hurry and get their capacity together quickly.
It is a truism that work expands to fill the deadline provided, and I believe that happened here too.
I do agree the FDA places a higher weight on the chance that someone will be harmed by a treatment, and less weight on the chance that someone will be harmed by lack of treatment. I am more than fine with that. The last thing we want is a regulator getting inventive with the trolley cart problem.
I also agree that there are going to be lessons we learn from this, and that we can try to make the FDA more responsive. They have not been great in keeping their two major roles separate: communicating the science accurately in all its complexity, and summarizing the science to the public clearly and concisely.
But making actual change for the better has got nothing to do with insane rantings about evil murderous thugs based solely on ideological paranoia and total ignorance.
The FDA needn't make a decision for anyone. They do not have to deal with trolley problems and mandate or ban things. They should simply allow anyone to put anything they want into their body at anytime. If you want to wait for medicine to pass through the FDA's trials, you should be able to. But if you want to buy it before they think it's okay, you should be able to do that too.
When the FDA screws up, it is generally in the direction of letting pharma get away with stuff that it shouldn’t (aducanumab). The main pressure they face is pharma lobbying them to cut corners. They are doing a complicated job with budget limitations, but their performance has been pretty solid.
The process takes time because there is legitimately a lot of work to do. It simply takes a lot of time to set up and run the trials, and a ton of work to go through all the adverse reactions.
Do you recall when the FDA temporarily suspended J&J due to unexpected thrombocytopaenia in a handful of patients? It turned out that, after five million doses were delivered, that there were three unexplained deaths. Finding effects that small takes a ton of work.
You cite one instance where the FDA screwed up by being too loose, and also one instance where the FDA screwed up by being too tight. Medical professionals I know say that the latter is far more common (maybe because they're designed to try to resist pharma lobbying in any way possible, and thus to put way too much weight on the chance that someone will be harmed by a treatment, and not enough weight on the chance that someone will be harmed by lack of treatment).
I don't think the J&J decision is an example of them being too tight. I think their response was pretty calibrated and correct.
The issue was rare enough that there no chance they could have detected the issue in trials. They measured an adverse reaction rate of about two per million.
Bu the effect was concentrated in the subpopulation of young woman, and they may not have detected all the cases. The low sample size meant that it was not impossible that the actual risk was significantly higher.
Crunching the numbers they had at the time, the risk of being unvaccinated for an additional day was significantly less than the risk of thrombocytopaenia. So even if someone decided to delay vaccination, rather than switching to another vaccine, then it probably evened out.
I generally like this article but I think "But current vaccines do not stop you from transmitting Delta to others" is problematic enough that you should IMO change it.
1) The study in Eric's tweet is about AZ which is not available in the US and there seems to be good evidence that it's less effective against Delta than the mRNA vaccines.
2) You probably already know this and it's mostly hedging against anti-vaxxers (and potentially compatible with your sentence) but the vaccines reduce your chance of being infected in the first place so that reduces transmission.
3) There's decent emerging evidence that infected vaccinated people have less infectious virus (even though viral loads may be high) and that the period during which they are infectious is shorter.
I know you don't want to hear it but you should have left this question to experts. You and I absolutely don't know whether it would be medically useful or not. I suppose it depends on the minutiae of how differently the variants behave.
I am skeptical as I have not seen evidence of severe disease among the vaxed except among severely immunocompromised. If boosters are not medically useful and are merely a political project to make the vaxed feel better during an epidemic among the vaxed, that would be a waste and distraction from the real problem of getting more people vaxed.
But I'd pay attention to an expert making this recommendation. There just isn't any value in you weighing in here.
About
Very much, we have seen that the drop in any severe COVID in vaxxinated persons in UK. Only 1% of fully vaccinated people in UK make up COVID Delta deaths. Most were infected before second jab came into effect. Only 0.5% (256 people) were those who had full immunity. It shows that of those deaths, 76% were among people already deemed clinically vulnerable. So 25% of 0.5% is frankly, error margin under any other circumstance. At that point we're talking a grand total of 64 people. A sample size not good enough for even a pilot study.
Before we worry about boosters, let's prevent the spread of new variants by vaccinating the poorer nations.
This assumes we only care about deaths among the vaccinated. If we also care about deaths among the unvaccinated (which I grant you, is becoming increasingly difficult to do with every passing day), then this isn’t the right metric. And if we care about people not getting sick, it’s also not the right metric. I know quite a few people who had breakthrough cases. They were all completely fine in a few days thanks to the vaccine, but if I can get a booster to help me avoid what they went through, why not?
Also Noah specifically gives reasons to reject the idea that vaccinating the poorer nations needs to conflict at all with booster shots. You don’t have to disagree with him, but it’s a little strange to repeat the point he already addressed in the article without acknowledging you’re doing that.
How medically useful boosters would be is a separate issue. I think it's a fair summary of expert consensus that boosters will probably be useful soon for the most vulnerable segment of the population, and it's likely they'll be useful for increasingly broad segments as time goes on.
This article is arguing we need to develop variant-specific boosters. That's a different question, and one that I'm saying only virologists and especially those studying Covid variants could usefully comment on. And honestly my guess is that most of them would say they don't know yet.
Oops I mean during an epidemic among the unvaxed, of course.
If you don’t actually know what you are talking about, then maybe you should stay in your damn lane.
It is really easy to make random guesses that the consensus opinion is wrong, and occasionally by chance you are going to be right. Early in the pandemic everyone and their dog at some point wondered if masks would in fact help. But it wasn’t the endless tweets by armchair epidemiologists that discovered the problem, it was a bunch of actual scientists that did the work.
Delta specific covid shots seems like a good idea. It probably seems like a good idea to the army of people who have probably been working on the problem since delta emerged. You even link to a report that says that Pfizer is working on the problem.
How hard is it to understand that building a new vaccine, testing it, and making sure that it is safe and effective is an insanely difficult problem that takes a staggering amount of time, resources, work, and money? Do you actually think we can just tweak the sequence, and get the new dose into arms by the end of the week?
Are you actually implying that the constraints on vaccine production are due to artificial political scarcity, and not because we are running up against real physical limits?
What I find infuriating is your assumption that the roadblocks are political, rather than practical, and therefore the CDC et al must be incompetent or corrupt or playing larger political games. You have no evidence of any of that.
> Do you actually think we can just tweak the sequence, and get the new dose into arms by the end of the week?
Not by the end of the week. But it really looks like you can just tweak the sequence and get the new dose into arms within about 100 days or so: https://blog.jonasneubert.com/2021/01/10/exploring-the-supply-chain-of-the-pfizer-biontech-and-moderna-covid-19-vaccines/
You might want one round of that for some basic trials first to double-check that nothing weird happened, before making the new version the main version
“How hard is it to understand that building a new vaccine, testing it, and making sure that it is safe and effective is an insanely difficult problem that takes a staggering amount of time, resources, work, and money?”
How is this relevant? We’ve already done all of that. We’re talking about tweaking it slightly to match up with the current variant. The whole POINT of mRNA vaccines is that it is very easy to do this.
And yes, it was “a bunch of actual scientists that did the work” on masks…after a bunch of actual scientists did the work of lying to people by claiming masks aren’t important because they were scared we’d take masks away from medical professionals. This is the worst possible example to choose to prove your point.
There’s science, and there’s policy. People like Noah shouldn’t be in the lab working on the vaccines. But they absolutely can and should be involved in the conversation about what our policies and priorities should be. We all should.
If there are technical issues with what Noah proposes, than that’s a great topic for conversation. If the only issue is “I don’t like who it is who is saying it,” that’s not useful or productive.
Moderna's vaccine was created and delivered to NIH for testing between Jan 11th and Feb 24th, which is what - 44 days? And they're saying future mRNA vaccines will be even faster. It's actually totally feasible that they could sequence a variant and get shots into arms of the first trial group within a week.
https://www.cnbc.com/2021/07/03/how-moderna-made-its-mrna-covid-vaccine-so-quickly-noubar-afeyan.html
People don't complain about Nate Silver, Alex Taborek, and others because they 'don't stay in their lane', but because they can be insufferable arrogant at times. They have posted half baked critiques, created straw men, and not acknowledged errors. They also claim as their own brilliant insights, facts and ideas that have been well known by other experts.
Does anyone really believe that vaccine developers aren't attempting to improve the vaccine to tackle variants, as has been the standard practice for flu vaccines for years, as Alex T implies?
Has Alex T considered that the problem is that the challenges are greater than what he is aware of, instead of the researchers and public health people having only half his intellect?
“People don't complain about Nate Silver, Alex Taborek, and others because they 'don't stay in their lane', but because they can be insufferable arrogant at times. They have posted half baked critiques, created straw men, and not acknowledged errors. They also claim as their own brilliant insights, facts and ideas that have been well known by other experts.”
I just want to double emphasize this because it’s exactly right. The problem with Silver isn’t that he isn’t an expert; the problem with Silver is that he’s an asshole who is more interested in scoring cheap Twitter points than in making the world a better place.
What are some examples of errors Silver and Tabarrok have made?
These are basically guys who gained a following being really good at creating and understanding models. Eventually they decided that they were so super smart that they didn't need to create models. Eventually people just get lazy.
Nate Silver’s hottest take was
https://twitter.com/natesilver538/status/1389590782605205509
If had bothered to look at his implied risk model, then he would have spotted the flaw in his reasoning. For many people the personal cost of wearing a mask in public is smaller than the expected cost of getting covid and/or spreading it to others - even if they are vaccinated.
During the pandemic it was rational for them to wear a mask. Even after getting vaccinated, the expected risk is still higher than the cost of wearing a mask, so it is still rational for them to wear a mask.
Tyler Cowan's general schtick is
https://marginalrevolution.com/marginalrevolution/2021/09/our-anti-science-science-advisors-yet-again.html
Tyler is aware of how insanely difficult it is to fit economic models to the real world, and how hard it can be to tease out subtle effects. For some reason, however, he seems to think that evaluating the safety and effectiveness of vaccines can be done in Excel in an afternoon.
I don't read Tabarrok enough to follow all his idiocy, but a random sample is:
https://www.brunswickgroup.com/alex-tabarrok-covid-economist-i19161/
He doesn't understand why "people got fixated on the specific trial design of two-doses, three-to-four weeks apart". I don't understand why he doesn't understand that you can't just change the conditions of a trial, and assume you are going to get completely predicatable results. To change the dosage and scheduling, they would have needed to run another phase three trial, and prove that the new dosage had better results than the original dosage.
Re your first link, did Silver say anywhere that he was talking about masks? I think he would agree that people who aren't annoyed by masks could be rational to continue to use them, and their behavior changes should be made elsewhere by changing the behaviors that *are* costly to them.
Re Tabarrok, I think you and he just disagree about which of the FDA's procedures are grounded in scientific fact as opposed to bureaucratic inertia. Other countries like Canada have done things like altering dosage and scheduling, without bad results, because their take on the science was less legalistic than the FDA's.
I am reading Nate as saying that because the vaccine makes you safer, then you are irrational for not engaging in riskier behavior.
I choose to drive carefully when driving an old rustbucket, because I don’t want to get into an accident. If you do get into an accident, it is much safer when you are driving a big suv with airbags and crumple zones than the old ruskbucket.
It is perfectly valid for me to choose to drive carefully in the suv, even though my risk of death is now lower. You can’t say it is irrational to continue to drive carefully.
With Tabarrok, he is indeed assuming a priori that the FDA is acting slower than necessary. I think his base assumption is that all government is, by definition, always overly bureaucratic. He provides no evidence of this.
But he is going further than just saying that the FDA is overly bureaucratic. He is actively asking why we don’t just skip phase 3.
I don't think we should skip phase three. The government simply shouldn't *ban* anyone from buying it before phase three is completed. Second, the government shouldn't ban people from volunteering to be exposed to the disease, so that months can be shaved off the time needed for a trial.
Not only does figuring out if a half dose works as well as a full dose take a phase three trial, it takes a much much larger phase three trial than he original phase three trial.
I have zero problems with the FDA trying to stop people from injecting themselves with unproved treatments that the learned about on the internet.
I understand that there are a lot of different opinions about challenge trials, and this is probably something we need to work out sooner rather than later. I think it is pretty likely there will be other pandemics this decade, and this all might just be a trial run for what comes next.
But in this particular case, it is not clear that challenge trials would have actually saved us any time. A big percentage of covid cases are asymptomatic and unreported, so we still don’t have a solid idea of the true infection rate. That makes it really hard to use challenge trial data to infer efficacy rates, so we would need to do phase three trials anyway.
That said, I would love to have been able to get my greedy little paws on challenge trial data. Knowing exactly what percentage of people infected are asymptomatic, symptomatic, and are hospitalized would make modelling so much easier. I just don’t think the data it is worth the human cost.
Just glad it isn’t my decision to make.
It is certainly *valid* to choose to drive carefully in the SUV. But, given that a little bit of extra precaution provides more benefit in the rustbucket than in the SUV, it seems unlikely that someone who is both interested in safety and interested in reaching destinations would drive *precisely* as carefully in the SUV as in the rustbucket. Different balances of interest in safety and in reaching destinations will yield different behaviors in the same vehicle, and if you are driving more cautiously than me in the same vehicle, it likely just reflects our different values. But if we switch to a vehicle where caution has less of an effect on safety, and you still exercise just as much caution, then it seems that safety mustn't have been playing a role in your decision.
I don't know whether Tabarrok believes that all government is always overly bureaucratic. I think the question here is whether the American CDC reached the right level of caution, or whether the Canadian agencies did (who were willing to deviate from the protocol that was tested, because the evidence suggested that these deviations would likely save more lives).
I completely agree that most people would change driving habits in the SUV. But it would be silly to call someone “irrational” if they chose to drive with the same amount of care. Why would we want to insult people for insisting on driving safely?
I haven’t actually been following what Health Canada has been doing (bad Canadian!), but this still seems like a bloody dangerous thing to have done. I understand they had an initial vaccine shortage and they were desperate, but the gamble could have easily not paid off.
I’m just glad the US had enough vaccine and capacity that we weren’t forced to do something that desperate.
I have a really hard time believing you would not drive extremely cautiously and avoid trips in a rust bucket, and drive very carefully in an SUV. Which is what Nate is saying.
I wonder if Alex understands why we don't have a vaccine against RSV. It turned out in studies it made the disease worse by some antibody dependent enhancement.
Rob, I read Nate’s comment differently. He’s not saying go wild, he’s saying you should be willing to take more risks on the simple logic of the costs of risky behavior have gone down. Before you could be infected or infect others. Now you can only infect others. Since that’s a lower burden to society, shouldn’t your behavior change?
And as far as wearing a mask inside the store because others don’t feel comfortable, I always looked at it differently. If I see a person without a mask, I assume that’s a sign that are vaccinated and they are signaling to others the extent of the ability to get back to normal if you do get vaxxed!
Alex does not appear to have any appreciation of the challenges of drug development. I call his approaches the Star Trek theory of pharmaceutical development, that people magically would develop the perfect cure in one hour and everyone gets better, buy the bad FDA makes them fill out paperwork.
The trial conditions were arbitrarily chosen. If spreading doses further apart, or going to half doses leads to far more people being vaccinated, then we should use our ability to reason to figure out which is likely to lead to more saved lives. We should not insist on doing all new, time consuming trials, *WHILE BANNING ANYONE WHO WANTS TO TAKE THE RISK FROM BUYING IT*, when the cost is tens of thousands dead.
The trial conditions weren’t arbitrarily chosen. They were the best guess at what would be effective based on what they knew at the time. They designed the phase 3 trials based on their best guess, and got pretty rapid FDA approval once the trials were completed.
Suppose, based on what we know now, we decided that half doses could work. Maybe it would be just as effective, or maybe it would be a lot less effective. Maybe there is a subpopulation where half doses provide no protection. Or maybe there is a completely weird side effect that we couldn’t possibly predict, given that we have barely any experience with mRNA vaccines.
So we can’t just roll out half doses. We need to test this first. We need to find thousands of people to give half doses to, and monitor the outcomes. We then need to run down every case where something unexpected happened, and figure out if it was due to the vaccine or an unrelated issue. That means we need to do a complete phase three trial for this.
Now it actually gets hard. The initial phase three trials showed that the full dosage was effective, and saved lives compared to not having a vaccine. That’s why they were able to get Emergency Use Authorization so quickly.
But we need to compare the half dose regimen to the full dose regimen. Even with millions of doses we still don’t have solid efficacy numbers on the full doses. It would require A LOT more data to determine if a half dose is just as effective.
Even if we did all that, then we would probably find that a half dose is almost as effective. But is it ethical to give patients a half dose, when we know that a half dose provides less protection? How many people should die on a half dose, that wouldn’t have died on a full dose, is acceptable? Is it ethical to give substandard treatment to some, in order to save other lives? Are these the kind of questions we want the FDA to make?
Is it ethical to even do a phase 3 trials of halve dose regimen, when we don’t even know it will work, when we know full doses work, when we know how long the whole process needs to take?
Maybe it would have been worth it. I’m just damn glad that I don’t have to be the person making these insane hard calls. I have no idea.
But I am pretty damn sure that Tabarrok hasn’t thought this through, and is just talking out his ass.
> is it ethical to give patients a half dose, when we know that a half dose provides less protection?
If it's more than half as much protection, and we know we can get it to twice as many people, then yes.
> How many people should die on a half dose, that wouldn’t have died on a full dose, is acceptable?
As long as it's fewer than the people who die on zero doses, that wouldn't have died on the half dose they would be able to receive.
> Is it ethical to give substandard treatment to some, in order to save other lives?
Yes. That is the essence of triage. Actual bioethicists work out the details of how precisely and when it should be done, but in many conditions, particularly emergencies, medical professionals have plenty of standardized procedures of when to deny care to some so that others can have it, or when to give only partial care, and how precisely to prioritize.
> Are these the kind of questions we want the FDA to make?
Probably not the FDA itself - they should just be determining which things are safe enough and effective enough that they are permissible, and some other body should be determining actual implementation. (For infectious diseases, this might naturally be the CDC, or the National Institute for Allergy and Infectious Disease, or something else - though in practice it probably usually ends up being insurance companies.)
President Biden got the Pfizer shot very early. He is almost 79 year old and is exposed to a lot of people. It is only a matter of time before Delta meets him and he is hardly protected at that phase. Deadly breakthrough cases at that age after 8-9 months are not rare at all as you could see in Israel before the booster campaign. Prime minister Bennet even said that those old people who got two shots but didn't get the booster shot are most at risk since they have the illusion of immunity. It is a matter of national security to start booster shots ASAP and to vaccinate the president first.
Regarding variant based shots, their approval and manufacturing will take some months. We need boosters now.
Agreed, especially because we are a month or two from the next winter wave. I'm more uncertain about exactly how much immunity is waning than the Israeli authorities are, but in that age group there is only upside to a third shot. Most people over 65 should probably lie and get a third shot if they aren't approved by mid-October; this is the advice I'm giving to my loved ones, anyway.
Pfizer's own positions are that any reduced effectiveness is due to waning immunity over time rather than lesser effectiveness against the Delta variant, and that Delta-specific boosters "almost certainly" won't be necessary:
https://www.bloomberg.com/news/articles/2021-09-15/pfizer-says-covid-19-vaccine-efficacy-erodes-over-time
https://www.washingtoninformer.com/pfizer-updating-vaccine-for-delta-but-doesnt-anticipate-needing-it/
I think the variant-specific boosters would be put into play if the vaccines had significantly reduced effectiveness against a dominant variant, but the people who would actually be marketing these boosters don't think that's the case here. Delta is more contagious across the board than original coronavirus, but that's a different issue.
Noah please read this: https://www.science.org/content/blog-post/myths-vaccine-manufacturing
You have a misunderstand of vaccine production complexities. Simply throwing more resources at it isn't a problem. The problem is spool-up time and expertise base - vaccine manufaturing is peak high-technology manufacturing. It requires very good specialists, perfect quality assurance, clean rooms, etc.
You seem to believe that the FDA can and will approve the variant specific boosters quickly.
This seems quite optimistic if by quickly you mean as quickly as new variants pop up and spread, a few weeks or months...
They are good people but are imbeded in a very risk adverse organization
Appreciate your booster for boosters.
A major missing piece is not being able to rapidly trial for safety and efficacy. Any country with the means to do it could have started this already, coordinating a roll out in order to best learn from it. Recommending to press ahead with boosters (variant specific or otherwise) without a more experimental approach is like advancing oblivious to leaving enemy positions at your back. Same problem can be seen with childhood vaccination issues. It's a recurring problem of lacking the political will and organisation to move forward both rapidly and cautiously.
TLDR, Noah how should we roll out boosters while also learning if they work and are safe, in a short time scale?
On this count, we need to think about things in a Bayesian way. If the original shot is safe and effective, a slight variation on that shot is basically certain to be safe and overwhelmingly likely to be effective.
Unfortunately it would not be safe to assume that. Knowledge of biology is highly incomplete. And a single molecular difference can change something from lethal to harmless. If the protein the immune response that is stimulated targets is slightly wrong it may cause serious autoimmune disease.
If that's true, how do we know the updated flu vaccine is safe every year?
Good question! They are live attenuated or inactive virus AFAIK. That would limit the possibilities for what is produced. Don't know we can be so confident with viral vector and mrna. There is also a good deal of monitoring of safety as they roll out in various countries seasons. I imagine there were a good deal of nerves the first time they updated without long term safety data. Now people are mostly quite comfortable as it is essentially tradition.
The modified mRNA is just copying whatever modification the virus made to its own mRNA, so it should be limited to exactly the same changes in effect that the virus itself has.
That said, it probably is worth doing some extra monitoring for the first couple updates before just rushing ahead, though it's probably *not* worth doing the same 30,000 participant trial waiting for 186 cases among the trial participants.
That seems sensible and close to the position FDA is signalling. Immunogenicity and safety in a smaller cohort. Like maybe a few thousand. Then enhanced surveillance as it goes out to more people. That would presumably add 2-3 months to any updated vaccine development. At least for the first few times.
I'm not sure the Delta mutations necessarily introduce new targetable epitopes, or that it's clear what these might be. It's like telling someone to target someone wearing an invisibility cloak. Sure, the cloak is unique but it's not easy to target.
The breakthrough we need now is in scaling up manufacturing while we figure out how to fine tune using the mRNA tech.
Shouldn't we be laser-focused on the nasal-spray options . . . ?
I think Delta is unique b/c of its speed, I am not sure just configuring it to search for a different protein wins this race, we need to take a different approach to counter the speed of Delta, I think we need a vaccine targeted at nasal and lung tissues, like a spray vaccine.
I’m not sure that we could get variant Booster is out in time. I suspect broad spectrum vaccines will be more effective.
I’m working in Argentina and a week before I came here I went and got a Moderna Booster. I had Pfizer Mar/Apr.
The idea is sound and we certainly should be working on it. However, two points should be kept in mind: First, the current vaccines offer a beautiful "template" allowing the same strategy for variants. However, without (at least) years of empirical experience, it would be dangerous to actually release one of these novel vaccines without substantial testing to verify efficacy and safety.
Second, the current vaccines do a great job of activating the B cell/IgG antibody system. However, they do a poor job of stimulating IgA, the primary immune molecule in the upper respiratory tract. That likely explains why they're more effective at preventing death than preventing infection or reducing viral load. Accordingly (though the point is consistent with your essay) it remains important to design new vaccine delivery systems, perhaps inhaled or subcutaneous mRNA.
One aspect to account for is supply chain issues. You will need separate production and distribution lines. For a company, it's easier to keep making loads of the same, rather than custom versions.
Excellent point. However in the US virtually all cases are no Delta (I assume this is similar in most places at this point) and (more speculatively) due to this most variants are likely to derive from Delta so they could just switch over completely for now. It may get more complicated in the future.
Diversity may be helpful, though. Antibodies for the original and the Delta may help mitigate better future variants.
The pharma community used insanely advanced technology to create covid vaccines in under 18 months.
Any sufficiently advanced technology is indistinguishable from magic.
Magic in fantasy novels is wielded by either good wizards or evil wizards.
Good wizards in fantasy novels give the protagonist everything they want: advice, support, and the occasional magic sword.
Everyone thinks they are the protagonist.
Therefore, if the pharma community isn’t giving you exactly what you want, they must be evil. They have powerful magic and can do anything. Anything bad that happens must be deliberate on their part.
It did not take us 18 months to make the vaccine. It took Moderna a weekend. The murderous gang at the FDA succeeded only in delaying medicine going to the public. Calling out their delays ought not be shamed.
Many of those delays were just the time needed to get manufacturing capacity running at scale. The FDA not approving AstraZeneca, and perhaps their not approving Novavax, are murderousness. But the delays on Moderna and Pfizer has had basically no effect on the number of vaccinated people since April 2021.
The vaccine companies respond to incentives. If rushing to get things together costs more money, why should they rush, when they know they won’t be allowed to sell it for months? Obviously, yes, there were hurdles to overcome - we could not get things running immediately - but if they had had the ability to sell it months earlier, *then* they would have had the incentive to hurry and get their capacity together quickly.
It is a truism that work expands to fill the deadline provided, and I believe that happened here too.
I do agree the FDA places a higher weight on the chance that someone will be harmed by a treatment, and less weight on the chance that someone will be harmed by lack of treatment. I am more than fine with that. The last thing we want is a regulator getting inventive with the trolley cart problem.
I also agree that there are going to be lessons we learn from this, and that we can try to make the FDA more responsive. They have not been great in keeping their two major roles separate: communicating the science accurately in all its complexity, and summarizing the science to the public clearly and concisely.
But making actual change for the better has got nothing to do with insane rantings about evil murderous thugs based solely on ideological paranoia and total ignorance.
The FDA needn't make a decision for anyone. They do not have to deal with trolley problems and mandate or ban things. They should simply allow anyone to put anything they want into their body at anytime. If you want to wait for medicine to pass through the FDA's trials, you should be able to. But if you want to buy it before they think it's okay, you should be able to do that too.
When the FDA screws up, it is generally in the direction of letting pharma get away with stuff that it shouldn’t (aducanumab). The main pressure they face is pharma lobbying them to cut corners. They are doing a complicated job with budget limitations, but their performance has been pretty solid.
The process takes time because there is legitimately a lot of work to do. It simply takes a lot of time to set up and run the trials, and a ton of work to go through all the adverse reactions.
Do you recall when the FDA temporarily suspended J&J due to unexpected thrombocytopaenia in a handful of patients? It turned out that, after five million doses were delivered, that there were three unexplained deaths. Finding effects that small takes a ton of work.
You cite one instance where the FDA screwed up by being too loose, and also one instance where the FDA screwed up by being too tight. Medical professionals I know say that the latter is far more common (maybe because they're designed to try to resist pharma lobbying in any way possible, and thus to put way too much weight on the chance that someone will be harmed by a treatment, and not enough weight on the chance that someone will be harmed by lack of treatment).
I don't think the J&J decision is an example of them being too tight. I think their response was pretty calibrated and correct.
The issue was rare enough that there no chance they could have detected the issue in trials. They measured an adverse reaction rate of about two per million.
Bu the effect was concentrated in the subpopulation of young woman, and they may not have detected all the cases. The low sample size meant that it was not impossible that the actual risk was significantly higher.
Crunching the numbers they had at the time, the risk of being unvaccinated for an additional day was significantly less than the risk of thrombocytopaenia. So even if someone decided to delay vaccination, rather than switching to another vaccine, then it probably evened out.
I don’t think the FDA was an outlier on that.