As always, interesting overview and very useful cost-benefit Fermis etc. As usual, I’m confused by some generalizing statements.
The WHO and EMA said there was no evidence there was an issue.
The EMA says:
EMA’s safety committee, PRAC, concluded its preliminary review of a signal of blood clots in people vaccinated with COVID-19 Vaccine AstraZeneca …
the vaccine is not associated with an increase in the overall risk of blood clots (thromboembolic events) in those who receive it;
...
however, the vaccine may be associated with very rare cases of blood clots associated with thrombocytopenia, i.e. low levels of blood platelets (elements in the blood that help it to clot) with or without bleeding, including rare cases of clots in the vessels draining blood from the brain (CVST).
These are rare cases – around 20 million people in the UK and EEA had received the vaccine as of March 16 and EMA had reviewed only 7 cases of blood clots in multiple blood vessels (disseminated intravascular coagulation, DIC) and 18 cases of CVST. A causal link with the vaccine is not proven, but is possible and deserves further analysis.
… Overall the number of thromboembolic events reported after vaccination, both in studies before licensing and in reports after rollout of vaccination campaigns (469 reports, 191 of them from the EEA), was lower than that expected in the general population. This allows the PRAC to confirm that there is no increase in overall risk of blood clots. However, in younger patients there remain some concerns, related in particular to these rare cases.
The Committee’s experts looked in extreme detail at records of DIC and CVST reported from Member States, 9 of which resulted in death. Most of these occurred in people under 55 and the majority were women. Because these events are rare, and COVID-19 itself often causes blood clotting disorders in patients, it is difficult to estimate a background rate for these events in people who have not had the vaccine. However, based on pre-COVID figures it was calculated that less than 1 reported case of DIC might have been expected by 16 March among people under 50 within 14 days of receiving the vaccine, whereas 5 cases had been reported.Similarly, on average 1.35 cases of CVST might have been expected among this age group whereas by the same cut-off date there had been 12. A similar imbalance was not visible in the older population given the vaccine.
The Committee was of the opinion that the vaccine’s proven efficacy in preventing hospitalisation and death from COVID-19 outweighs the extremely small likelihood of developing DIC or CVST. However, in the light of its findings, patients should be aware of the remote possibility of such syndromes, and if symptoms suggestive of clotting problems occur patients should seek immediate medical attention …
The PRAC will undertake additional review of these risks, including looking at the risks with other types of COVID-19 vaccines (although no signal has been identified from monitoring so far). …
Sorry for the lengthy quote, but I think it’s worthwhile to read this, and I think it does not fit your description. I think that’s not saying there was no evidence of an issue, it’s saying there maybe was an issue among younger people and PRAC should look into that issue, but cost-benefit analysis says vaccination is still much better.
Given the different age groups affected and analyzed, I would like to understand what your “So it’s not remotely fair to use the background population rate when you’re explicitly targeting your elderly population for vaccinations.” sentence means. Which background population rate was used by the authorities? (By the way, the media in Germany noted that the difference between UK and EU may be due to the fact that the age groups receiving AZ in these places are different. That is, AZ in Germany was seemingly given to young nurses, many of which are women, because it was restricted to people under 65.)
For your “sequence of events”, as always I’d be happy to know whether “there’s extensive reporting of anything that happens to people right after getting the vaccine” is actually true. Intuition tells me that there’s also extensive reporting of symptoms of COVID-19 in times of a COVID-19 pandemic, but in fact there’s a relevant amount of unknown cases additional to official numbers. If headaches are the symptom of the relevant blood clots, should we really expect overreporting? My intuition would be that people underreported this symptom, in particular because everyone has heard that you should expect to feel sick etc after being vaccinated. On the other hand, after this discussion and media coverage, I expect people to report headaches more often, and this would also happen without any government-imposed interruption of the vaccination campaign—maybe even more so.
Being a European, I guess I must have lost my mind, so I don’t really understand what “All of this due, effectively, to pure p-hacking, without even bothering to pretend otherwise.” is supposed to mean. “p-hacking” would be intentional behavior, in particular combined with the “pretend” part. So you imply that there was an intention by analysts in some agency to stop the vaccination? And “without even bothering to pretend otherwise”, that is, they also said so? (But then again, seeing the Samo Burja tweet and the text around it, I guess it’s not even necessary to present a plausible mechanism how such things work. “Malice”, “madness”, etc. I can imagine the government meeting: “How do we cover up our failure?” “Let’s stop vaccination by pointing out blood clots! We understand statistics perfectly, so we know that the experts in the Paul Ehrlich Institute are wrong, but due to our malice and madness, we follow their recommendation.”)
“but you have a legal obligation to these people that forces your hand, because ‘there could be legal consequences’? And there’s no way to, say, pass a new law to fix that, even if you should have fixed it long ago? So that’s it, nothing you could do, huh? ”
If I am not mistaken about the Bundestag procedures, the interruption of vaccination did not take long compared to the time it takes to change a law.
The EMA report saying that there may be an association with very rare blood clots (which would still imply far more blood clots prevented than caused because Covid causes blood clots + math) came out after I hit the publish button. I agree that they then changed their tune from the pure explicit ‘no evidence’ line to a new line of Very Serious Person language designed to make it easier for everyone to resume.
When I say p-hacking, I mean that the search function was identical to what happens when people p-hack, with identical results—they’re looking at all conditions and subconditions, in all regions and subregions, with any possible lag ranges, in order to find something that happened above rate. And for the same reason—people are highly motivated to find a positive result somewhere. I don’t think anyone in a meeting said the word “p-hack”, but no one has denied that the search took place in this fashion, either, nor did they make any attempt to account for it, or notice any issues after they identified what they suspected was the issue. And there’s still no mechanism.
I didn’t intend to explicitly say that the authorities are failing to use an adjusted background rate, but my prior is that they’re not doing so, because no one has mentioned doing the adjustment and in general no one silently does such adjustments when they make things seem more safe, because again everyone is on the ‘make the vaccines look unsafe’ team.
The EMA report saying that there may be an association with very rare blood clots (which would still imply far more blood clots prevented than caused because Covid causes blood clots + math) came out after I hit the publish button.
It was around five or at a maximum ten minutes. They had one press conference in which the first person that spoke for longer laid out both.
The EMA report saying that there may be an association with very rare blood clots … came out after I hit the publish button. I agree that they then changed their tune from the pure explicit ‘no evidence’ line …
I may be naive and sound like a broken record but I still think it would be helpful if claims about what some said or did were backed up by a link or something.
When I say p-hacking, I mean that the search function was identical to what happens when people p-hack, with identical results—they’re looking at all conditions and subconditions, in all regions and subregions, with any possible lag ranges, in order to find something that happened above rate. And for the same reason—people are highly motivated to find a positive result somewhere.
So when you say p-hacking, you don’t really mean people p-hack? Or maybe I don’t understand the aim of your word choice—is this just rhetorics, and not meant to be accurate? It seems related to the questions whether MLK was a criminal, and tax is theft?
I don’t think anyone in a meeting said the word “p-hack”, but no one has denied that the search took place in this fashion, either, nor did they make any attempt to account for it, or notice any issues after they identified what they suspected was the issue. And there’s still no mechanism.
So no one has denied that—was there any public accusation to deny it? Any discussion where it would have to be denied? Or was there at least a serious indication that “the search took place in this fashion”?
I didn’t intend to explicitly say that the authorities are failing to use an adjusted background rate, but my prior is that they’re not doing so, because no one has mentioned doing the adjustment and in general no one silently does such adjustments when they make things seem more safe,
“the authorities” seems like a word that doesn’t explain anything. The health minister possibly does not “use an adjusted background rate”, he relies on judgement by a specialized agency. Assuming that this agency does not use an adjusted background rate seems quite a stretch; of course it’s possible, but where is the evidence for that in your summary? Where is the evidence for the claim that “no one silently does such adjustments when they make things seem more safe”?
because again everyone is on the ‘make the vaccines look unsafe’ team.
I have no idea why “everyone” should be on that side, and again, I don’t see any evidence for that. Asked by Watson, 220 German politicians today publicly stated their trust in AZ. Your implicit model of politics (or of whatever, I still don’t know who “everyone” is) seems to be wrong.
Some comments: * Re: blood-clotting, I think you’ve bolded the wrong section. “it is difficult to estimate a background rate for these events in people who have not had the vaccine. However, based on pre-COVID figures” is the part to bold, which makes the rest of the sentence rather pointless. You cannot use pre-COVID figures to estimate expected blood-clotting when we’re during a pandemic which involves an illness that specifically causes blood-clotting. * Institutions use extensive amounts of caveats and other forms of blame-avoiding language as a matter of course, but this language doesn’t contain much information. That is, irrespective of how high the actual risk is, I wouldn’t expect the language to change much. For instance, the phrase “patients should be aware of the remote possibility” is a waste of time for me to read, and for them to write, unless it affects the agency’s actual public health guidelines. * The focus on this one particular symptom is arbitrary. It seems implausible that a drug that actually made people sick would do so only via rare blood clots and only in young people, whereas it’s commonplace in bad statistics to find arbitrary problems in arbitrary subgroups. Hence the accusation of p-hacking. This xkcd comic is a decent illustration of how such a thing can happen.
Now contrast that with the real harms caused by delaying vaccinations, as Zvi points out in his essay. Orders of magnitude more people will die due to delayed vaccinations, not to mention the second-order effect of harming vaccine acceptance worldwide for the foreseeable future.
Insofar as one accepts the notion that a) the risk of side-effects is not nearly high enough to warrant this response, and b) the harm to the vaccination effort and to vaccine acceptancy is orders of magnitude higher, then the actual political response in Europe looks like gross negligence, malfeasance, or outright malice—not of the form “let’s intentionally hinder vaccination and get people killed” (which I agree would be implausible comic-book villainy), but rather of the form “as a politician, I only care about avoiding blame; I don’t care if my (in)actions kill thousands, as long as I’m not blamed for this”, which has to me become an increasingly plausible lense through which to see politics. Here is one Zvi post on the politics of blame-avoidance and inaction.
For me personally, the part during Covid that soured me a ton on EU competence was this: The politicians were so worried of being blamed for wasting tax-payer money on expensive vaccines, that they negotiated lower prices in exchange for receiving vaccines months later. This calculation was so crazy and wrong-headed that you kind of need something like Zvi’s blame-avoidance model to make sense of it.
As always, interesting overview and very useful cost-benefit Fermis etc. As usual, I’m confused by some generalizing statements.
The EMA says:
Sorry for the lengthy quote, but I think it’s worthwhile to read this, and I think it does not fit your description. I think that’s not saying there was no evidence of an issue, it’s saying there maybe was an issue among younger people and PRAC should look into that issue, but cost-benefit analysis says vaccination is still much better.
Given the different age groups affected and analyzed, I would like to understand what your “So it’s not remotely fair to use the background population rate when you’re explicitly targeting your elderly population for vaccinations.” sentence means. Which background population rate was used by the authorities? (By the way, the media in Germany noted that the difference between UK and EU may be due to the fact that the age groups receiving AZ in these places are different. That is, AZ in Germany was seemingly given to young nurses, many of which are women, because it was restricted to people under 65.)
For your “sequence of events”, as always I’d be happy to know whether “there’s extensive reporting of anything that happens to people right after getting the vaccine” is actually true. Intuition tells me that there’s also extensive reporting of symptoms of COVID-19 in times of a COVID-19 pandemic, but in fact there’s a relevant amount of unknown cases additional to official numbers. If headaches are the symptom of the relevant blood clots, should we really expect overreporting? My intuition would be that people underreported this symptom, in particular because everyone has heard that you should expect to feel sick etc after being vaccinated. On the other hand, after this discussion and media coverage, I expect people to report headaches more often, and this would also happen without any government-imposed interruption of the vaccination campaign—maybe even more so.
Being a European, I guess I must have lost my mind, so I don’t really understand what “All of this due, effectively, to pure p-hacking, without even bothering to pretend otherwise.” is supposed to mean. “p-hacking” would be intentional behavior, in particular combined with the “pretend” part. So you imply that there was an intention by analysts in some agency to stop the vaccination? And “without even bothering to pretend otherwise”, that is, they also said so? (But then again, seeing the Samo Burja tweet and the text around it, I guess it’s not even necessary to present a plausible mechanism how such things work. “Malice”, “madness”, etc. I can imagine the government meeting: “How do we cover up our failure?” “Let’s stop vaccination by pointing out blood clots! We understand statistics perfectly, so we know that the experts in the Paul Ehrlich Institute are wrong, but due to our malice and madness, we follow their recommendation.”)
If I am not mistaken about the Bundestag procedures, the interruption of vaccination did not take long compared to the time it takes to change a law.
The EMA report saying that there may be an association with very rare blood clots (which would still imply far more blood clots prevented than caused because Covid causes blood clots + math) came out after I hit the publish button. I agree that they then changed their tune from the pure explicit ‘no evidence’ line to a new line of Very Serious Person language designed to make it easier for everyone to resume.
When I say p-hacking, I mean that the search function was identical to what happens when people p-hack, with identical results—they’re looking at all conditions and subconditions, in all regions and subregions, with any possible lag ranges, in order to find something that happened above rate. And for the same reason—people are highly motivated to find a positive result somewhere. I don’t think anyone in a meeting said the word “p-hack”, but no one has denied that the search took place in this fashion, either, nor did they make any attempt to account for it, or notice any issues after they identified what they suspected was the issue. And there’s still no mechanism.
I didn’t intend to explicitly say that the authorities are failing to use an adjusted background rate, but my prior is that they’re not doing so, because no one has mentioned doing the adjustment and in general no one silently does such adjustments when they make things seem more safe, because again everyone is on the ‘make the vaccines look unsafe’ team.
It was around five or at a maximum ten minutes. They had one press conference in which the first person that spoke for longer laid out both.
Assuming that you refer to a biological mechanism, there are people who claim to have found just that.
I may be naive and sound like a broken record but I still think it would be helpful if claims about what some said or did were backed up by a link or something.
So when you say p-hacking, you don’t really mean people p-hack? Or maybe I don’t understand the aim of your word choice—is this just rhetorics, and not meant to be accurate? It seems related to the questions whether MLK was a criminal, and tax is theft?
So no one has denied that—was there any public accusation to deny it? Any discussion where it would have to be denied? Or was there at least a serious indication that “the search took place in this fashion”?
“the authorities” seems like a word that doesn’t explain anything. The health minister possibly does not “use an adjusted background rate”, he relies on judgement by a specialized agency. Assuming that this agency does not use an adjusted background rate seems quite a stretch; of course it’s possible, but where is the evidence for that in your summary? Where is the evidence for the claim that “no one silently does such adjustments when they make things seem more safe”?
I have no idea why “everyone” should be on that side, and again, I don’t see any evidence for that. Asked by Watson, 220 German politicians today publicly stated their trust in AZ. Your implicit model of politics (or of whatever, I still don’t know who “everyone” is) seems to be wrong.
(Commenting from Germany.)
Some comments:
* Re: blood-clotting, I think you’ve bolded the wrong section. “it is difficult to estimate a background rate for these events in people who have not had the vaccine. However, based on pre-COVID figures” is the part to bold, which makes the rest of the sentence rather pointless. You cannot use pre-COVID figures to estimate expected blood-clotting when we’re during a pandemic which involves an illness that specifically causes blood-clotting.
* Institutions use extensive amounts of caveats and other forms of blame-avoiding language as a matter of course, but this language doesn’t contain much information. That is, irrespective of how high the actual risk is, I wouldn’t expect the language to change much. For instance, the phrase “patients should be aware of the remote possibility” is a waste of time for me to read, and for them to write, unless it affects the agency’s actual public health guidelines.
* The focus on this one particular symptom is arbitrary. It seems implausible that a drug that actually made people sick would do so only via rare blood clots and only in young people, whereas it’s commonplace in bad statistics to find arbitrary problems in arbitrary subgroups. Hence the accusation of p-hacking. This xkcd comic is a decent illustration of how such a thing can happen.
Now contrast that with the real harms caused by delaying vaccinations, as Zvi points out in his essay. Orders of magnitude more people will die due to delayed vaccinations, not to mention the second-order effect of harming vaccine acceptance worldwide for the foreseeable future.
Insofar as one accepts the notion that a) the risk of side-effects is not nearly high enough to warrant this response, and b) the harm to the vaccination effort and to vaccine acceptancy is orders of magnitude higher, then the actual political response in Europe looks like gross negligence, malfeasance, or outright malice—not of the form “let’s intentionally hinder vaccination and get people killed” (which I agree would be implausible comic-book villainy), but rather of the form “as a politician, I only care about avoiding blame; I don’t care if my (in)actions kill thousands, as long as I’m not blamed for this”, which has to me become an increasingly plausible lense through which to see politics. Here is one Zvi post on the politics of blame-avoidance and inaction.
For me personally, the part during Covid that soured me a ton on EU competence was this: The politicians were so worried of being blamed for wasting tax-payer money on expensive vaccines, that they negotiated lower prices in exchange for receiving vaccines months later. This calculation was so crazy and wrong-headed that you kind of need something like Zvi’s blame-avoidance model to make sense of it.