As a government employee, none of what's happened surprises me. My agency is full of people who find it all to easy to say "no" or "my hands are tied". It really has become an immovable object that the anti-government types complain about, and they're right.
If I tried to envision what would happen if they wanted to make some of the structural changes necessary to make things better, some combination of politics and media stories seems to be the thing that would either stop progress or make the person who stuck their neck out sorry. Suffice to say this is not the same government that landed on the moon or fought WWII. I don't know exactly how we got ourselves into this mess, but untying the knot seems harder every day.
"But just because a trial was conducted a certain way doesn’t mean we have to slavishly imitate it."
Yes, shortly after Moderna set up a trial with four weeks between doses and declared people fully vaccinated two weeks after the second dose, Pfizer set up its trial with three weeks between doses and fully vaccinated one week after the second dose -- saving a total of two weeks vs. Moderna. The reason was that Pfizer wanted to beat Moderna to reporting results and get all the public relations boost associated with being first.
There was nothing medical about that decision, but we have to follow these protocols. When Walgreens decided to use a four-week interval for both brands, for scheduling ease, the CDC told the company to stop doing that:
"Walgreens Not Following U.S. Guidance on Pfizer Vaccine Spacing
"After complaints from customers and the C.D.C., the pharmacy chain will start scheduling doses three weeks apart....
"Kate Grusich, a spokeswoman for the C.D.C., said the agency had asked Walgreens to stop using a longer-than-recommended period between doses.
"The company’s vaccine-scheduling system by default schedules all second doses four weeks after the first. Doses of Moderna’s vaccine, which Walgreens is also administering, are supposed to be spaced four weeks apart. Using the same gap for both vaccines was 'the easiest way to stand up the process based on our capabilities at the time,' Dr. Kevin Ban, Walgreens’ chief medical officer, said in an interview."
I’m heading to Argentina for work next month. I had two Pfizer shots. Last in April. I’m heading over the border to Oregon this weekend to get another shot of Moderna (Idaho has a state vaccine database) as a booster.
I got onto argument on Twitter with the 100 mask in school people. They are fanatics. They were complaining about parents opting out.
My daughter goes to a boarding school, for math and science. It is a highly highly selective school, filled with smart, mostly progressive kids. It is in a state where masking is optional. I asked her how many of the kids choose to wear masks. She said very very few.
I was in the military for 22 years. Enlisted. Once it’s mandatory, everyone will get it.
I got into another argument a couple of weeks ago on Twitter saying that mandatory vaccine, is less intrusive than mandatory masks.
Delta has the idea by charging a surcharge for medical insurance for people that choose not to be vaccinated.
That’s the route we need to take, just make life extremely difficult for unvaccinated people. The FA should just require it if people fly. The federal government should require it for all contractors.
There is something incomplete in Matt's analysis of why people in the military were slow to get vaccinated. The feeing of "I'm invulnerable" could mean no fear of side effects from the vaccine. And the ethos of courage facing an enemy in defense country should have pushed for vaccination as a way of protecting fellow citizens.
I think we also need to ask why the press has not been asking these questions. Even when reporters are aware of failings like the reluctance to allow screening testing, they do not ask DFA/CDC to explain why these mistakes happened? And if the principles will not talk, (not even off the record?) they could ask "knowledgeable observers" to speculate on why DFA/CDC keep getting it wrong.
Now I could speculate that in 2020 with Trump's politicizing policy, MSM might have been reluctant to be seen "piling on" institutions already under attack, but why now?
I think the core issue with institutional flexibility in the US gov't is that it is very hard for a bureaucrat to get fired for Just Doing His Job, but quite easy for them to get tagged as a shady character who should not be promoted and maybe should be quietly pushed out the door, by challenging procedure or rocking the boat.
You can't change that without very intense pressure from the top down to care about results more than the formal procedures that lead to the results.
One other think to keep in mind. Your opinion may be vastly different from that of the public. As one example at the end of the day when there were leftover vaccine drug stores, hospitals, super markets, etc. were doing PA announcements - if you want a shot come and get it. Now I would say that makes perfect sense. Obviously you don’t want to throw away vaccine.
But when they polled the question +60% of people were opposed - some very opposed - to that plan.
When making these proposals as an official, it’s not only about what you personally think, it’s what the public thinks and accepts.
I agree that in this case at least the FDA has been inexcusably slothful. I mean, what was the range of possible decisions for them? With closing in on 400 million shots given in the US, were they going to come out and say, "Nope, no approval. It's too dangerous"? So if they *had* to approve the shots, why the hell just not do it earlier?
But what I don't quite get is how to distinguish this from other cases. Is scale the *only* thing that matters? Going through a rigorous process for approving a cancer drug means that people will likely die who wouldn't have if the drug had been approved more quickly. Far fewer than COVID, of course, but that's cold comfort to them. And unlike the COVID vaccinations, these drugs are typically not fast tracked so cancer patients with bad timing are simply out of luck. Sorry about that.
Now we know they go through the process with the cancer drug because it's the only way of knowing it's effective enough and safe enough: the trial is the only source of information. Not the case with the COVID shots, with hundreds of millions of people getting the shots outside of any tests. While this is not rigorously-acquired information (although the VAERS data actually look pretty good to me), it's a *mountain* of information. Why not use that and truly speed the process? There's a lot to be said for using a Bayesian approach in cases like this!
>>Not the case with the COVID shots, with hundreds of millions of people getting the shots outside of any tests. While this is not rigorously-acquired information (although the VAERS data actually look pretty good to me), it's a *mountain* of information. Why not use that and truly speed the process? <<
We didn't have "hundreds of millions" of shots' worth of data until quite a few months after the introduction of the vaccines.
Military vaccination is also not something you can blame on the FDA. The military can order its members literally to walk into gunfire. It doesn't need FDA's approval to require them to take a vaccine.
"Then there are tons of other questions about acquired immunity in recovered Covid patients, about the efficacy of different kinds of masks, the utility of HEPA filters, etc. where the quality of the evidence just isn’t that good."
It just boggles the mind that basic scientific questions like this don't have answers based on RCT trials, and no-one seems particularly curious.
I'm not in this field, but my sense is that In normal times, research activity is prioritized from the bottom up, with various academics deciding which research questions they want to work on. This is of course subject to funding availability, which could be used to nudge priorities somewhat.
OTOH, during an emergency I think you want a top-down model where leaders/experts prioritize the most important questions we need answered and then find researchers willing/able to do the work.
“One RCT tested these masks that so many high-profile public-health officials have touted. This “first RCT of cloth masks,” in the trial’s own words (it is apparently still the only one), was a 2015 study by MacIntyre, et al. in Hanoi, Vietnam. A relatively large study, with over 1,100 participants, it tested cloth masks against surgical masks and did not feature a no-mask control group. The trial tested the protection of health-care workers, instructing them to wear a two-layer cloth mask at all times on every shift (“except in the toilet or during tea or lunch breaks”) across four weeks.”
This study strongly suggests that masks can make a big difference. Not sure why the author of this article only uses it to endure the cloth masks. If masks weren’t helpful, there shouldn’t have been a difference.
Matt has an extremely common, and somewhat bonkers opinion. The FDA is doing precisely what it is supposed to be doing here.
Since April there have been about 200 covid deaths per million in the US. So, if you chose to be unvaccinated for the last five months, then that was roughly your increased chance of death from covid.
The vaccines also have a risk. There have been roughly 35 deaths per million of people who have had the vaccine in that time frame according to VAERS, but the vast majority of those deaths were almost certainly from another source. A covid shot is not going to protect you from an unrelated heart attack.
There has been a total of 3 deaths due to thrombosis with thrombocytopenia syndrome … about 0.02 per million… that are directly attributable to the J&J vaccine. As far as we know, those are the only deaths. But it is not impossible that there are more.
So _on average_, we now know that it is vastly safer to take the vaccine then to be unprotected. The FDA had enough evidence to determine this with Moderna last December, which is why it authorized it for emergency use. And they turned out to be right! The EUA vaccines are vastly safer than not being vaccinated.
But we don’t actually know that the vaccines are safe and effective for all sub-populations. The J&J vaccine, for example, has a higher risk for younger women. The FDA, correctly, decided to briefly halt administration of the J&J vaccine until it could figure out if it was safe for this subpopulation.
Determining that a vaccine is safe and effective on average is extremely difficult. Determining that it is safe and effective for all subpopulations is even more difficult. That’s what the FDA has to figure out before it can give full approval.
Yes, In all likelihood the FDA is going to grind through their data and the bureaucracy will eventually determine that Moderna et al as safe. And then Matthew Yglesias, Tyler Cowen, and everyone on this comment section will crow about being right.
But if the FDA did bow to pressure to become “more flexible”, and they approved the vaccines before being certain, and it turned out that a vaccine caused harm to left-handed Scotsmen (for example), then I doubt any of you would lose too much sleep.
But the FDA would take a massive hit, and the left-handed Scottish antivaxxers would scream vindication, and everyone who received mandatory vaccinations would sue. And then what happens in the next pandemic, when the FDA has already shown it can’t be trusted?
It’s one thing to argue that, as a political question, we should encourage everyone to be vaccinated. But we can’t base our politics on fact, and then use politics to influence the facts.
We know that the vaccines are safe for general use, because we've seen *hundreds of millions of people use them*. Yes, we followed the outcomes of those people with much less specificity than we would for a typical drug trial, but on the other hand there are like thousands of times as many people who have used them as would in a typical drug trial.
There was never going to ever, ever, ever be a drug trial for a generally permitted drug that would suss out whether that drug was going to have a bad effect on sub 0.1% of the population who might use it. Normal FDA policies would not find out if a drug was bad for left-hand Scotsmen.
What the FDA has done is not be flexible about seeing the results of a badly controlled, badly administered trial that none the less has a lot of power because, again, it's got like probably at least 5,000, maybe 50,000 times as many people in it than a typical phase III trial.
The number of people affected by the J&J issue was about 30% of the population, not 0.1%. Even then it took ten million people before there was enough data to spot the issue. Vaccine side effects are really, really rare, and it takes a ton of data and careful analysis to spot issues.
Left handed men with Scotish ancestry make up about .5% of the US population. We may just have enough data to spot an issue with this group, if we thought there might be an issue with them. But it would take trillions of doses for it to show up on a clustering model.
We know the FDA isn’t waiting for arbitrary small subgroup results, because there will never be enough data for that.
But what about African-American men with heart disease? Latino women with high blood pressure? Cancer survivors? We can and should make sure we understand the effects of those subpopulations.
It is important for the FDA to do due diligence to make sure that what they claim is safe and effective, is safe and effective based on what can be reasonably known. They aren’t idiots.
The FDA has a ton of rules. The cliché is that all those rules are written in blood: someone didn’t think to look for something, and people died. Then a lot of people spent a lot of time trying to make sure that it didn’t happen again.
And yeah, the FDA is more worried about deaths from people who choose to take the vaccine then deaths of people who choose not to take the vaccine. What kind of monsters could act differently?
I read this and feel like you missing the entire point that Matt is trying to make.
The FDA and every other government authority has been urging all the left handed Scotsman to get the vaccine! There is no sub population (of adults) that they are suggesting hold off until there is more data!
The real question is - since they are not doing the normal process and they are actually doing an emergency process, why aren't they actually treating it like an EMERGENCY? They are asking almost the entire country to literally derail months/years of their lives, but aren't willing to work around the clock or work through a holiday?
If this is really an emergency that requires shutting down, stay at home, endless masking, etc. then why aren't all the government agencies screaming for funding/flexibility so they can do everything possible to get this over ASAP?
Rules are written in blood, but that doesn't make them good rules, just rules that were written in response to a bad event. "We must do something, this is something, ergo we must do this" is a seductive syllogism but one that I think institutions should try to avoid. "What is the cost, what is the benefit, what is the likely outcome of doing nothing, and how can we do better next time" are the questions that I think we should be trying to ask here, especially when the subject is 4.5M and counting dead humans.
I agree that is a possible failure mode for agencies. There is an argument that that happened with the Nuclear Regulatory Commission, which lead to increasingly costly regulation, which eventually killed the industry. The NRC bore no costs for imposing regulations and building requirements, and massive reputational costs if an accident did happen. Imposing regulations allowed them to push back against constant activist charges of regulatory capture.
I don’t know of any evidence that that dynamic occurs at the FDA. If a drug turns out to have adverse effects, then the pharma company is blamed rather than the FDA. There is no activist community arguing that the FDA is cutting corners.
If there is any regulatory capture, it is by pharma companies that are intensely interested in cutting regulation to reduce their costs. The FDA’s capitulation in aducanumab shows that it needs more internal safeguards, not less.
Note that Matt’s entire argument is about the political effects of the FDA’s decision. The opinions of “Public health twitter” is not evidence.
Classic FDA defense. That’s pretty well the whole story — “there’s a chance doing things differently would hurt the FDA’s recitation!” That it. Abolish the FDA.
People have views on the FDA based on their politics and background. Some people trust the FDA, and some don’t.
The FDA makes decisions on whether pharmaceuticals are safe and effective or not. Sometimes they make the right call (covid vaccines), and sometimes they don’t (aducanumab).
These are completely different questions. The FDA’s reputation for trustworthiness is not the same as the FDA’s actual trustworthiness. It is just insane not to realize that these are different questions.
Where the FDA does fuck up… and it does fuck up (aducanumab)…. Is when it tries to find the middle line between political pressures and the available evidence.
Pharma is hard… Epistemology is hard… Anyone who tells you differently is selling something.
>>>Relatedly, they continue to not recommend that people who got the Johnson & Johnson shot get an mRNA booster even though there is clear data to support that idea. The issue, as Leana Wen of the Milken Institute School of Public Health explains, is that nobody has done a specific safety trial on the mix-and-match<<<
My sense after reading this article is that quite a few of the examples cited by Matt (I've excerpted one, above) aren't issues of lack of flexibility so much as lack of state capacity. Given the gravity of the crisis that confronted the country in the winter of 2020, we should have marshaled much greater resources toward fighting the pandemic. More studies, more "trying everything," more use of available technologies (like smartphone-based contact tracing software), more use of the Defense Production Act. In short, a titanic national effort. And yes, greater capacity would have enabled greater flexibility.
Also, in the realm of public health, I can't help but suspect the country's lack of genuinely universal, government-guaranteed healthcare coverage also plays a role.
American governance as a whole — not just its public health apparatus — isn't operating at a very high level by rich country standards, from what I can tell. I believe the US is suffering a genuine crisis of governance.
I don't totally disagree with this, but did other developed countries do more 'trying everything', contact tracing, etc.? My impression is that a small number of them did use draconian restrictions (South Korea, Taiwan, now Australia)- but, they also seem to have populations that accept that, so it's tough to say that the military locking down whole streets would 'work' in America. In fact, I'm very confident that it wouldn't.
What are the other 'try everything' methods that other developed countries with stronger state capacity did, that the US didn't?
>>I don't totally disagree with this, but did other developed countries do more 'trying everything', contact tracing, etc.?<<
In the main, no, but A) The USA by rights *ought* to have a far greater capacity to do things than other high income countries (4x GDP of our closest competitor, Japan); B) many (certainly not all) of them mounted a much more thorough, serious and unified response to the crisis early on, and so were in a better position to absorb the odd mistake as time went by. America's lack of nimbleness was magnified by some highly egregious failings in the early days, under Trump (incoherent borders policy, badly botched development of covid tests, lack of proper quarantine protocols, widespread civil disobedience — often with the encouragement of government officials — vis-a-vis social distancing measures, and so on.
I believe America's generally chaotic, shambolic and lumbering response to the pandemic from the beginning, under Trump — and the terrible delays and wasted time this caused (we should have been doing MUCH more in January-February-early March) — are the biggest cause of the country's travails; not generalized "lack of flexibility" as such (though to be sure every mistake adds up).
excellent. I think what it boils down to is we need competent government. And while we certainly do have some good people in government, by and large that's not what we have.
My wife used to work in property management for the government. I asked her how long it should take to get a project done if everyone just did their jobs, but still protecting the environment etc. She said 1 to 2 months.
Then I asked her how long it really takes. 1 to 2 years. That's due to multiple problems. A large part is people not doing their jobs. Civil service protections need to be greatly reformed. People that aren't doing their jobs need to be fired.
In a similar vein, red tape that isn't accomplishing anything also needs to go.
Yes we do need government to do all types of things. But that government needs to be responsive and competent. That's not what we are getting.
And if can't depend on government to get it's act together in a pandemic then when???
I can see your point about civil service protections needing to be reformed, but how? Do you think a spoils system would be better? We’d have all Trump appointees for a while, then all Biden ones?
Agreed, that would be wonderful, but from experience with civil service jobs I'd say can't be achieved. And progressives are strongly pushing to get that "fire only for cause" standard applied to everyone -- they think it is wonderful if it becomes impossible to get rid of people who aren't working out.
Over the course of this Covid epidemic many issues have come up where people who lack knowledge about how things work propose solutions that won't get anything done faster and in many cases make things worse. They do not understand production techniques or quality control exigencies or supply chains or anything. Scaling up production is a difficult and exacting task. Having a bunch of people or organizations chasing after the same production equipment is foolish. The proposed alternative, to have the government take control and allocate resources to maximize production seems nice although some legal impediments apply. The more compelling fact is that a lot of the supply chain for production equipment and necessary raw materials lie outside American jurisdiction. And that isn't even getting started on mandating disclosure of proprietary information or the acquisition of patents. Now all of this takes time to do anyway and there is no reason to think that a World War Two style effort to generate vaccine production was going to succeed. Especially when you are asking the world to serve American customers first.
When it comes to a critique of Mathew's beliefs about how to do to science and evaluate data from trials I doubt he would take my word on why is is wrong. Fortunately there is a very good source he might believe and that is the very able Vox journalist Julia Beluz. She can instruct him about the permissibility of changing test hypotheses in mid study or even, as is often regrettably done, abandoning the hypothesis to publish findings that weren't being tested. It's a no no.
> Honestly, if you made masks mandatory forever in grocery stores with the stated goal of reducing the spread of the common cold, I wouldn’t really mind.
Not that you've explicitly said so here, but it's a thing I see a lot around the internet. I think a lot of people are confusing the effects of wearing a mask to the grocery store with the effects of closing all the offices and many of the schools.
Let me see if I understand Matt’s point. The FDA authorized the vaccine using the emergency process due to the existence of an emergency. But they shouldn’t have down that. They should have claimed they used the regular process but in reality only used the emergency process?
Matt's point is that while most US institutions made radical (arguably previously unfathomable) changes to adapt to an extremely urgent and very different situation. Public health institutions did not do this. You are bringing up the existing EUA process as an example and I'm glad they used it for the vaccines. It's hard to believe that there was nothing else that could've been done, especially given several pieces of evidence that the FDA did not try greatly to speed things up or change the way they are doing things (e.g. their recent statement that they'll prioritize full approval over other matters while complaining that this will have some costs).
I think this is where what happened is blinding people to what was most probable to happen. We got two vaccines that worked really well, and another couple that worked okay very fast! We were incredibly fortunate! The most probable outcome given how we responded was one vaccine that had a much lower effectiveness rate and took much longer.
In a world where we don't have a successful vaccine in the first 9 months, and its more like 18 - 24 months, the fact that the FDA, CDC, etc. did not act with greater urgency would be even more egregious. And that I think is the point that Matt is trying to make. If we have another recurrence where we don't get so lucky, having government agencies that can respond with greater flexibility will be an even greater imperative.
Imagine the devastation to the world if the virus is slightly more deadly but sticks around for 20 months before vaccine is successful - and then we take another year+ to roll it out significantly.
The UK beat the USA to the start of vaccinations by a grand total of five days. I agree American public health authorities have made mistakes, but I share BronxZooCobra's skepticism of this particular critique of Matt's. We smashed our own record for vaccine development!
The biggest counter-example of how the US managed its trials is China. And yes, they started jabs months before the first Pfizer dose was administered. But as the recipient of two Sinovac jabs of rather dubious effectiveness, America's M.O. is looking pretty good to me.
And since we are dealing with a subject group where half will be dead in a year. How do you respond to the headline - 11.5% of people who received the vaccine died within 3 months.
“ IIRC the safety testing in adults took <10 weeks. It should've been authorized for use in nursing homes the next day.”
With no clue of it even works? The public isn’t going to go for that for a second. I agree it makes sense. You obviously agree that it makes sense. But the public won’t like your use grandma as a guinea pig plan.
As a government employee, none of what's happened surprises me. My agency is full of people who find it all to easy to say "no" or "my hands are tied". It really has become an immovable object that the anti-government types complain about, and they're right.
If I tried to envision what would happen if they wanted to make some of the structural changes necessary to make things better, some combination of politics and media stories seems to be the thing that would either stop progress or make the person who stuck their neck out sorry. Suffice to say this is not the same government that landed on the moon or fought WWII. I don't know exactly how we got ourselves into this mess, but untying the knot seems harder every day.
"But just because a trial was conducted a certain way doesn’t mean we have to slavishly imitate it."
Yes, shortly after Moderna set up a trial with four weeks between doses and declared people fully vaccinated two weeks after the second dose, Pfizer set up its trial with three weeks between doses and fully vaccinated one week after the second dose -- saving a total of two weeks vs. Moderna. The reason was that Pfizer wanted to beat Moderna to reporting results and get all the public relations boost associated with being first.
There was nothing medical about that decision, but we have to follow these protocols. When Walgreens decided to use a four-week interval for both brands, for scheduling ease, the CDC told the company to stop doing that:
"Walgreens Not Following U.S. Guidance on Pfizer Vaccine Spacing
"After complaints from customers and the C.D.C., the pharmacy chain will start scheduling doses three weeks apart....
"Kate Grusich, a spokeswoman for the C.D.C., said the agency had asked Walgreens to stop using a longer-than-recommended period between doses.
"The company’s vaccine-scheduling system by default schedules all second doses four weeks after the first. Doses of Moderna’s vaccine, which Walgreens is also administering, are supposed to be spaced four weeks apart. Using the same gap for both vaccines was 'the easiest way to stand up the process based on our capabilities at the time,' Dr. Kevin Ban, Walgreens’ chief medical officer, said in an interview."
https://www.nytimes.com/2021/04/05/business/covid-vaccine-pfizer-walgreens-three-weeks.html
I’m heading to Argentina for work next month. I had two Pfizer shots. Last in April. I’m heading over the border to Oregon this weekend to get another shot of Moderna (Idaho has a state vaccine database) as a booster.
I got onto argument on Twitter with the 100 mask in school people. They are fanatics. They were complaining about parents opting out.
My daughter goes to a boarding school, for math and science. It is a highly highly selective school, filled with smart, mostly progressive kids. It is in a state where masking is optional. I asked her how many of the kids choose to wear masks. She said very very few.
I was in the military for 22 years. Enlisted. Once it’s mandatory, everyone will get it.
I got into another argument a couple of weeks ago on Twitter saying that mandatory vaccine, is less intrusive than mandatory masks.
Delta has the idea by charging a surcharge for medical insurance for people that choose not to be vaccinated.
That’s the route we need to take, just make life extremely difficult for unvaccinated people. The FA should just require it if people fly. The federal government should require it for all contractors.
There is something incomplete in Matt's analysis of why people in the military were slow to get vaccinated. The feeing of "I'm invulnerable" could mean no fear of side effects from the vaccine. And the ethos of courage facing an enemy in defense country should have pushed for vaccination as a way of protecting fellow citizens.
I think we also need to ask why the press has not been asking these questions. Even when reporters are aware of failings like the reluctance to allow screening testing, they do not ask DFA/CDC to explain why these mistakes happened? And if the principles will not talk, (not even off the record?) they could ask "knowledgeable observers" to speculate on why DFA/CDC keep getting it wrong.
Now I could speculate that in 2020 with Trump's politicizing policy, MSM might have been reluctant to be seen "piling on" institutions already under attack, but why now?
I think the core issue with institutional flexibility in the US gov't is that it is very hard for a bureaucrat to get fired for Just Doing His Job, but quite easy for them to get tagged as a shady character who should not be promoted and maybe should be quietly pushed out the door, by challenging procedure or rocking the boat.
You can't change that without very intense pressure from the top down to care about results more than the formal procedures that lead to the results.
One other think to keep in mind. Your opinion may be vastly different from that of the public. As one example at the end of the day when there were leftover vaccine drug stores, hospitals, super markets, etc. were doing PA announcements - if you want a shot come and get it. Now I would say that makes perfect sense. Obviously you don’t want to throw away vaccine.
But when they polled the question +60% of people were opposed - some very opposed - to that plan.
When making these proposals as an official, it’s not only about what you personally think, it’s what the public thinks and accepts.
I agree that in this case at least the FDA has been inexcusably slothful. I mean, what was the range of possible decisions for them? With closing in on 400 million shots given in the US, were they going to come out and say, "Nope, no approval. It's too dangerous"? So if they *had* to approve the shots, why the hell just not do it earlier?
But what I don't quite get is how to distinguish this from other cases. Is scale the *only* thing that matters? Going through a rigorous process for approving a cancer drug means that people will likely die who wouldn't have if the drug had been approved more quickly. Far fewer than COVID, of course, but that's cold comfort to them. And unlike the COVID vaccinations, these drugs are typically not fast tracked so cancer patients with bad timing are simply out of luck. Sorry about that.
Now we know they go through the process with the cancer drug because it's the only way of knowing it's effective enough and safe enough: the trial is the only source of information. Not the case with the COVID shots, with hundreds of millions of people getting the shots outside of any tests. While this is not rigorously-acquired information (although the VAERS data actually look pretty good to me), it's a *mountain* of information. Why not use that and truly speed the process? There's a lot to be said for using a Bayesian approach in cases like this!
>>Not the case with the COVID shots, with hundreds of millions of people getting the shots outside of any tests. While this is not rigorously-acquired information (although the VAERS data actually look pretty good to me), it's a *mountain* of information. Why not use that and truly speed the process? <<
We didn't have "hundreds of millions" of shots' worth of data until quite a few months after the introduction of the vaccines.
It’s quite possible for a cancer drug to harm patients.
The NYT article on vaccinations in the military says exactly the opposite what MY claims it does. Uptake among military service members has not been fairly low; it was (as of the time of the article) higher than the civilian population. Moreover, since service members skew younger, the vaccination rate is significantly higher. In the general population of 18-39 year olds, 38% had been vaccinated by July (https://www.wsj.com/articles/young-americans-arent-getting-vaccinated-jeopardizing-covid-19-fight-11625995802#:~:text=Some%2038%25%20of%20people%20ages,vaccination%20rate%2C%20at%2080%25.), compared to 77% overall in the Navy and 70% in the Army (with the Air Force and the Marines being significantly lower).
Mandating DoD vaccinations is all to the good, but we're talking fairly marginal returns here.
Military vaccination is also not something you can blame on the FDA. The military can order its members literally to walk into gunfire. It doesn't need FDA's approval to require them to take a vaccine.
"Then there are tons of other questions about acquired immunity in recovered Covid patients, about the efficacy of different kinds of masks, the utility of HEPA filters, etc. where the quality of the evidence just isn’t that good."
It just boggles the mind that basic scientific questions like this don't have answers based on RCT trials, and no-one seems particularly curious.
Yeah, this is puzzling to me.
I'm not in this field, but my sense is that In normal times, research activity is prioritized from the bottom up, with various academics deciding which research questions they want to work on. This is of course subject to funding availability, which could be used to nudge priorities somewhat.
OTOH, during an emergency I think you want a top-down model where leaders/experts prioritize the most important questions we need answered and then find researchers willing/able to do the work.
A lot of the research on masks is super old, and looks like there hasn't been much work on covid specifically.
https://www.city-journal.org/do-masks-work-a-review-of-the-evidence
It also looks like people don't want to look because they may not like the answer and they'll also get a lot of political grief.
“One RCT tested these masks that so many high-profile public-health officials have touted. This “first RCT of cloth masks,” in the trial’s own words (it is apparently still the only one), was a 2015 study by MacIntyre, et al. in Hanoi, Vietnam. A relatively large study, with over 1,100 participants, it tested cloth masks against surgical masks and did not feature a no-mask control group. The trial tested the protection of health-care workers, instructing them to wear a two-layer cloth mask at all times on every shift (“except in the toilet or during tea or lunch breaks”) across four weeks.”
This study strongly suggests that masks can make a big difference. Not sure why the author of this article only uses it to endure the cloth masks. If masks weren’t helpful, there shouldn’t have been a difference.
Matt has an extremely common, and somewhat bonkers opinion. The FDA is doing precisely what it is supposed to be doing here.
Since April there have been about 200 covid deaths per million in the US. So, if you chose to be unvaccinated for the last five months, then that was roughly your increased chance of death from covid.
The vaccines also have a risk. There have been roughly 35 deaths per million of people who have had the vaccine in that time frame according to VAERS, but the vast majority of those deaths were almost certainly from another source. A covid shot is not going to protect you from an unrelated heart attack.
There has been a total of 3 deaths due to thrombosis with thrombocytopenia syndrome … about 0.02 per million… that are directly attributable to the J&J vaccine. As far as we know, those are the only deaths. But it is not impossible that there are more.
So _on average_, we now know that it is vastly safer to take the vaccine then to be unprotected. The FDA had enough evidence to determine this with Moderna last December, which is why it authorized it for emergency use. And they turned out to be right! The EUA vaccines are vastly safer than not being vaccinated.
But we don’t actually know that the vaccines are safe and effective for all sub-populations. The J&J vaccine, for example, has a higher risk for younger women. The FDA, correctly, decided to briefly halt administration of the J&J vaccine until it could figure out if it was safe for this subpopulation.
Determining that a vaccine is safe and effective on average is extremely difficult. Determining that it is safe and effective for all subpopulations is even more difficult. That’s what the FDA has to figure out before it can give full approval.
Yes, In all likelihood the FDA is going to grind through their data and the bureaucracy will eventually determine that Moderna et al as safe. And then Matthew Yglesias, Tyler Cowen, and everyone on this comment section will crow about being right.
But if the FDA did bow to pressure to become “more flexible”, and they approved the vaccines before being certain, and it turned out that a vaccine caused harm to left-handed Scotsmen (for example), then I doubt any of you would lose too much sleep.
But the FDA would take a massive hit, and the left-handed Scottish antivaxxers would scream vindication, and everyone who received mandatory vaccinations would sue. And then what happens in the next pandemic, when the FDA has already shown it can’t be trusted?
It’s one thing to argue that, as a political question, we should encourage everyone to be vaccinated. But we can’t base our politics on fact, and then use politics to influence the facts.
But this is in fact wrong.
We know that the vaccines are safe for general use, because we've seen *hundreds of millions of people use them*. Yes, we followed the outcomes of those people with much less specificity than we would for a typical drug trial, but on the other hand there are like thousands of times as many people who have used them as would in a typical drug trial.
There was never going to ever, ever, ever be a drug trial for a generally permitted drug that would suss out whether that drug was going to have a bad effect on sub 0.1% of the population who might use it. Normal FDA policies would not find out if a drug was bad for left-hand Scotsmen.
What the FDA has done is not be flexible about seeing the results of a badly controlled, badly administered trial that none the less has a lot of power because, again, it's got like probably at least 5,000, maybe 50,000 times as many people in it than a typical phase III trial.
The number of people affected by the J&J issue was about 30% of the population, not 0.1%. Even then it took ten million people before there was enough data to spot the issue. Vaccine side effects are really, really rare, and it takes a ton of data and careful analysis to spot issues.
Left handed men with Scotish ancestry make up about .5% of the US population. We may just have enough data to spot an issue with this group, if we thought there might be an issue with them. But it would take trillions of doses for it to show up on a clustering model.
We know the FDA isn’t waiting for arbitrary small subgroup results, because there will never be enough data for that.
But what about African-American men with heart disease? Latino women with high blood pressure? Cancer survivors? We can and should make sure we understand the effects of those subpopulations.
It is important for the FDA to do due diligence to make sure that what they claim is safe and effective, is safe and effective based on what can be reasonably known. They aren’t idiots.
The FDA has a ton of rules. The cliché is that all those rules are written in blood: someone didn’t think to look for something, and people died. Then a lot of people spent a lot of time trying to make sure that it didn’t happen again.
And yeah, the FDA is more worried about deaths from people who choose to take the vaccine then deaths of people who choose not to take the vaccine. What kind of monsters could act differently?
I read this and feel like you missing the entire point that Matt is trying to make.
The FDA and every other government authority has been urging all the left handed Scotsman to get the vaccine! There is no sub population (of adults) that they are suggesting hold off until there is more data!
The real question is - since they are not doing the normal process and they are actually doing an emergency process, why aren't they actually treating it like an EMERGENCY? They are asking almost the entire country to literally derail months/years of their lives, but aren't willing to work around the clock or work through a holiday?
If this is really an emergency that requires shutting down, stay at home, endless masking, etc. then why aren't all the government agencies screaming for funding/flexibility so they can do everything possible to get this over ASAP?
(p.s. Hi Rob! LTNS!)
p.s. Hi N! It is so weird running into people I know in the normcore world. ;)
Rules are written in blood, but that doesn't make them good rules, just rules that were written in response to a bad event. "We must do something, this is something, ergo we must do this" is a seductive syllogism but one that I think institutions should try to avoid. "What is the cost, what is the benefit, what is the likely outcome of doing nothing, and how can we do better next time" are the questions that I think we should be trying to ask here, especially when the subject is 4.5M and counting dead humans.
I agree that is a possible failure mode for agencies. There is an argument that that happened with the Nuclear Regulatory Commission, which lead to increasingly costly regulation, which eventually killed the industry. The NRC bore no costs for imposing regulations and building requirements, and massive reputational costs if an accident did happen. Imposing regulations allowed them to push back against constant activist charges of regulatory capture.
I don’t know of any evidence that that dynamic occurs at the FDA. If a drug turns out to have adverse effects, then the pharma company is blamed rather than the FDA. There is no activist community arguing that the FDA is cutting corners.
If there is any regulatory capture, it is by pharma companies that are intensely interested in cutting regulation to reduce their costs. The FDA’s capitulation in aducanumab shows that it needs more internal safeguards, not less.
Note that Matt’s entire argument is about the political effects of the FDA’s decision. The opinions of “Public health twitter” is not evidence.
Classic FDA defense. That’s pretty well the whole story — “there’s a chance doing things differently would hurt the FDA’s recitation!” That it. Abolish the FDA.
This is what I find frustrating.
People have views on the FDA based on their politics and background. Some people trust the FDA, and some don’t.
The FDA makes decisions on whether pharmaceuticals are safe and effective or not. Sometimes they make the right call (covid vaccines), and sometimes they don’t (aducanumab).
These are completely different questions. The FDA’s reputation for trustworthiness is not the same as the FDA’s actual trustworthiness. It is just insane not to realize that these are different questions.
Where the FDA does fuck up… and it does fuck up (aducanumab)…. Is when it tries to find the middle line between political pressures and the available evidence.
Pharma is hard… Epistemology is hard… Anyone who tells you differently is selling something.
>>>Relatedly, they continue to not recommend that people who got the Johnson & Johnson shot get an mRNA booster even though there is clear data to support that idea. The issue, as Leana Wen of the Milken Institute School of Public Health explains, is that nobody has done a specific safety trial on the mix-and-match<<<
My sense after reading this article is that quite a few of the examples cited by Matt (I've excerpted one, above) aren't issues of lack of flexibility so much as lack of state capacity. Given the gravity of the crisis that confronted the country in the winter of 2020, we should have marshaled much greater resources toward fighting the pandemic. More studies, more "trying everything," more use of available technologies (like smartphone-based contact tracing software), more use of the Defense Production Act. In short, a titanic national effort. And yes, greater capacity would have enabled greater flexibility.
Also, in the realm of public health, I can't help but suspect the country's lack of genuinely universal, government-guaranteed healthcare coverage also plays a role.
American governance as a whole — not just its public health apparatus — isn't operating at a very high level by rich country standards, from what I can tell. I believe the US is suffering a genuine crisis of governance.
I don't totally disagree with this, but did other developed countries do more 'trying everything', contact tracing, etc.? My impression is that a small number of them did use draconian restrictions (South Korea, Taiwan, now Australia)- but, they also seem to have populations that accept that, so it's tough to say that the military locking down whole streets would 'work' in America. In fact, I'm very confident that it wouldn't.
What are the other 'try everything' methods that other developed countries with stronger state capacity did, that the US didn't?
>>I don't totally disagree with this, but did other developed countries do more 'trying everything', contact tracing, etc.?<<
In the main, no, but A) The USA by rights *ought* to have a far greater capacity to do things than other high income countries (4x GDP of our closest competitor, Japan); B) many (certainly not all) of them mounted a much more thorough, serious and unified response to the crisis early on, and so were in a better position to absorb the odd mistake as time went by. America's lack of nimbleness was magnified by some highly egregious failings in the early days, under Trump (incoherent borders policy, badly botched development of covid tests, lack of proper quarantine protocols, widespread civil disobedience — often with the encouragement of government officials — vis-a-vis social distancing measures, and so on.
I believe America's generally chaotic, shambolic and lumbering response to the pandemic from the beginning, under Trump — and the terrible delays and wasted time this caused (we should have been doing MUCH more in January-February-early March) — are the biggest cause of the country's travails; not generalized "lack of flexibility" as such (though to be sure every mistake adds up).
excellent. I think what it boils down to is we need competent government. And while we certainly do have some good people in government, by and large that's not what we have.
My wife used to work in property management for the government. I asked her how long it should take to get a project done if everyone just did their jobs, but still protecting the environment etc. She said 1 to 2 months.
Then I asked her how long it really takes. 1 to 2 years. That's due to multiple problems. A large part is people not doing their jobs. Civil service protections need to be greatly reformed. People that aren't doing their jobs need to be fired.
In a similar vein, red tape that isn't accomplishing anything also needs to go.
Yes we do need government to do all types of things. But that government needs to be responsive and competent. That's not what we are getting.
And if can't depend on government to get it's act together in a pandemic then when???
I can see your point about civil service protections needing to be reformed, but how? Do you think a spoils system would be better? We’d have all Trump appointees for a while, then all Biden ones?
I'm not suggesting replacing the entire bureaucracy every time we change presidents.
I am saying that managers need to be able to fire people that don't perform. And you shouldn't need years of documentation to do it.
Agreed, that would be wonderful, but from experience with civil service jobs I'd say can't be achieved. And progressives are strongly pushing to get that "fire only for cause" standard applied to everyone -- they think it is wonderful if it becomes impossible to get rid of people who aren't working out.
Over the course of this Covid epidemic many issues have come up where people who lack knowledge about how things work propose solutions that won't get anything done faster and in many cases make things worse. They do not understand production techniques or quality control exigencies or supply chains or anything. Scaling up production is a difficult and exacting task. Having a bunch of people or organizations chasing after the same production equipment is foolish. The proposed alternative, to have the government take control and allocate resources to maximize production seems nice although some legal impediments apply. The more compelling fact is that a lot of the supply chain for production equipment and necessary raw materials lie outside American jurisdiction. And that isn't even getting started on mandating disclosure of proprietary information or the acquisition of patents. Now all of this takes time to do anyway and there is no reason to think that a World War Two style effort to generate vaccine production was going to succeed. Especially when you are asking the world to serve American customers first.
When it comes to a critique of Mathew's beliefs about how to do to science and evaluate data from trials I doubt he would take my word on why is is wrong. Fortunately there is a very good source he might believe and that is the very able Vox journalist Julia Beluz. She can instruct him about the permissibility of changing test hypotheses in mid study or even, as is often regrettably done, abandoning the hypothesis to publish findings that weren't being tested. It's a no no.
> Honestly, if you made masks mandatory forever in grocery stores with the stated goal of reducing the spread of the common cold, I wouldn’t really mind.
Not that you've explicitly said so here, but it's a thing I see a lot around the internet. I think a lot of people are confusing the effects of wearing a mask to the grocery store with the effects of closing all the offices and many of the schools.
Let me see if I understand Matt’s point. The FDA authorized the vaccine using the emergency process due to the existence of an emergency. But they shouldn’t have down that. They should have claimed they used the regular process but in reality only used the emergency process?
They should have lied in other words…
Is that the proposal?
That's clearly not his point.
It sure seems like his point.
Matt's point is that while most US institutions made radical (arguably previously unfathomable) changes to adapt to an extremely urgent and very different situation. Public health institutions did not do this. You are bringing up the existing EUA process as an example and I'm glad they used it for the vaccines. It's hard to believe that there was nothing else that could've been done, especially given several pieces of evidence that the FDA did not try greatly to speed things up or change the way they are doing things (e.g. their recent statement that they'll prioritize full approval over other matters while complaining that this will have some costs).
Didn’t try and speed up things!?! It was in peoples arms in 1/4 the time as the previous record!!!
I think this is where what happened is blinding people to what was most probable to happen. We got two vaccines that worked really well, and another couple that worked okay very fast! We were incredibly fortunate! The most probable outcome given how we responded was one vaccine that had a much lower effectiveness rate and took much longer.
In a world where we don't have a successful vaccine in the first 9 months, and its more like 18 - 24 months, the fact that the FDA, CDC, etc. did not act with greater urgency would be even more egregious. And that I think is the point that Matt is trying to make. If we have another recurrence where we don't get so lucky, having government agencies that can respond with greater flexibility will be an even greater imperative.
Imagine the devastation to the world if the virus is slightly more deadly but sticks around for 20 months before vaccine is successful - and then we take another year+ to roll it out significantly.
The UK beat the USA to the start of vaccinations by a grand total of five days. I agree American public health authorities have made mistakes, but I share BronxZooCobra's skepticism of this particular critique of Matt's. We smashed our own record for vaccine development!
The biggest counter-example of how the US managed its trials is China. And yes, they started jabs months before the first Pfizer dose was administered. But as the recipient of two Sinovac jabs of rather dubious effectiveness, America's M.O. is looking pretty good to me.
How long would it have taken with you in charge?
Hahaha!! You think the public is going to buy that for one second?
Here take this - we don’t know that it works but try it anyway.
Like really - that’s what you think?
And since we are dealing with a subject group where half will be dead in a year. How do you respond to the headline - 11.5% of people who received the vaccine died within 3 months.
Also the average length of stay in a nursing home before death is 13 months. Proving efficacy in that environment might not be as easy as you claim.
“ IIRC the safety testing in adults took <10 weeks. It should've been authorized for use in nursing homes the next day.”
With no clue of it even works? The public isn’t going to go for that for a second. I agree it makes sense. You obviously agree that it makes sense. But the public won’t like your use grandma as a guinea pig plan.