PlagueWatch 2020-23 - Coronavirus

[throatybeard]

Anyway, I assume Biden’s team is developing a plan for international distribution but intends to keep its eye on the American ball for at least another 6-8 weeks, which seems to be when every American who wants a vaccine will be able to get one—actual insight welcome!

I hope I'm wrong, but I can't imagine states like Missouri getting anywhere near vaccinating everyone who needs/wants one in six to eight weeks. The state government here has been actively obstructing distribution of vaccine proportional to population density, and I'd be surprised if we were anywhere close to saturation in mid May.

[CDu]

We will probably have to agree to disagree about what constitutes a nit, but I think it matters that prevention of hospitalizations and deaths were not included as primary outcomes in any of the trials and were secondary outcomes in only two of the trials (J&J and AZ). Given the relatively small number of hospitalizations and deaths in the placebo groups, we probably shouldn't be telling people that the studies indicate that the vaccines create near-100% protection against death. Yes, we have good reason to believe that all of the vaccines are very effective at preventing hospitalizations and deaths but that's not what the studies were designed to measure, so some modesty is warranted.

You're, right, we'll have to agree to disagree.

One person's nit is evidence for another person's conspiracy theory. If we start seeing sporadic reports of people dying after full vaccination there will be a chorus of "I told you so's" from anti-vaxxers.

Crazy people are gonna do crazy regardless. If you tell the anti-vaxxers that it isn't close to 100% (which would be false based on all the evidence that we have to date, and all the evidence seems to suggest that the effect gets bigger with time, not smaller), those people will be even LESS inclined to get the vaccine. So I don't see any value in equivocating on the effectiveness.

[CDu]

Meanwhile, on the vaccine administration front, we topped 92 million doses yesterday. Over 60 million people (18.1%) have gotten at least 1 dose, and 31.3 million (9.4%) have gotten the full regimen. We're up to 23.5% of the adult population having gotten at least 1 dose. We should pass 100 million doses by the end of the week, and hopefully we should pass 10% of the total population fully vaccinated by week's end as well.

We're now at the point where we should finally start seeing a little of the benefit of vaccination, as probably about 5% of the population are now experiencing the full value of the vaccine (>7 days post second dose).

Still a LONG way to go, but the vaccine administration has really picked up steam more recently. And with the companies of the 3 vaccines approved in the US supposedly planning to ship another 150-200 million doses this month, hopefully things continue to improve on the administration front.

[Bostondevil]

We will probably have to agree to disagree about what constitutes a nit, but I think it matters that prevention of hospitalizations and deaths were not included as primary outcomes in any of the trials and were secondary outcomes in only two of the trials (J&J and AZ). Given the relatively small number of hospitalizations and deaths in the placebo groups, we probably shouldn't be telling people that the studies indicate that the vaccines create near-100% protection against death. Yes, we have good reason to believe that all of the vaccines are very effective at preventing hospitalizations and deaths but that's not what the studies were designed to measure, so some modesty is warranted.

One person's nit is evidence for another person's conspiracy theory. If we start seeing sporadic reports of people dying after full vaccination there will be a chorus of "I told you so's" from anti-vaxxers. FWIW, it's been one week since my first dose of Pfizer and my wife gets the J&J vaccine tomorrow. My wife had a choice of Moderna at a CVS 45 minutes away or J&J at a CVS 1:45 away. She chose J&J for the convenience of not having to schedule a 2nd shot but both of us would have happily taken any of the 3 vaccines.

Prevention of hospitalizations and deaths were not included as PRELIMINARY outcomes. Both were included as primary outcomes, in every trial, even if the media coverage of those trials focused on prevention of infection.

https://www.vox.com/22273502/covid-v...fficacy-deaths

[Bostondevil]

Yeah, it probably isn’t fair to expect her to be an expert across tons of disciplines. But if she doesn’t understand infectious disease research and vaccine trials well enough, she really shouldn’t have written this article (not arguing with you, as I think we agree; just saying it out loud).

. . .

The framing of the article makes sense from the lens that she is a consumer advocate. But her argument in this setting feels pretty hollow and her references to the data seem flawed.

She is not alone. I maintain that all journalism degrees should include a statistical methods course requirement and if I ran news outlets, I would employ statistical fact checkers (with bonafide statistical expertise) to go over all reporting on news items of scientific, medical, or public health importance.

[Skydog]

Who are the authors?

It's one author: Hilda Bastian. Who should know better.

Yes, she probably should. I've looked at her background and she's very much a healthcare consumer advocate but I don't see much infectious disease experience, so, maybe she doesn't have adequate experience to write this particular article.

Yeah, it probably isn’t fair to expect her to be an expert across tons of disciplines. But if she doesn’t understand infectious disease research and vaccine trials well enough, she really shouldn’t have written this article (not arguing with you, as I think we agree; just saying it out loud).

In summary, I think:
1. No, these vaccines aren’t 100% effective against death and hospitalizations. However, they are proving to be really close to it following the second dose. Arguing over the minor difference is very nit picky, especially when one misinterprets the evidence.
2. No, they aren’t identically effective, but they appear to be close enough to each other in prevention of truly severe disease that these differences aren’t really worth noting. See point 1.
3. No, they aren’t identical in AE risks. But the differences in risks and severity of the AEs experienced is small enough that it doesn’t really make much difference.
4. Most importantly, the consumer isn’t likely to have a choice on vaccine. Especially not the consumer who would benefit most from the vaccine. So it seems silly to even argue about the differences between them. If one is available to you, you should take it ASAP.

The one thing that WOULD have been a sensible argument would be to focus on the first month following the first dose. There IS a meaningful discussion to be had in pointing out that you don’t see much benefit in the first few weeks after the first dose. So it IS important to emphasize that you aren’t suddenly in the clear, especially after the first dose. But that isn’t a real differentiator between vaccines, rather a limitation of all of them.

The framing of the article makes sense from the lens that she is a consumer advocate. But her argument in this setting feels pretty hollow and her references to the data seem flawed.

I propose another theory about why her article is poor - it was shaped to fit a narrative that would draw eyeballs.

[Bluedog]

I propose another theory about why her article is poor - it was shaped to fit a narrative that would draw eyeballs.

That's pretty unusual for The Atlantic in my experience. (Not saying it's not the case here though necessarily).

[Bostondevil]

That's pretty unusual for The Atlantic in my experience. (Not saying it's not the case here though necessarily).

Or it wasn't The Atlantic with the motive.

[rsvman]

....
The one thing that WOULD have been a sensible argument would be to focus on the first month following the first dose. There IS a meaningful discussion to be had in pointing out that you don’t see much benefit in the first few weeks after the first dose. So it IS important to emphasize that you aren’t suddenly in the clear, especially after the first dose. But that isn’t a real differentiator between vaccines, rather a limitation of all of them...

Maybe this is also a nit that I am about to pick, but it isn't a limitation of the vaccines at all. It's a property of the human immune system. No active vaccination can bypass the way the immune system functions. (Passive vaccination provides immediate protection, though, which is why I made sure to include the word "active" in the sentence.)

[cato]

Hey CDu, since you’ve already done the work, why not pose your questions/criticisms to the author of The Atlantic article? It would be interesting to see some responses from the journalist. To the point you and BD raised, perhaps she is missing something in her analysis and I would be interested in knowing what her thoughts would be if she looked at those issues.

As for the article, it did not strike me as either clickbait or egregiously off-message. There are a ton of things that have not been discussed openly and in detail as a result of various aspects of this pandemic, not the least of which are how messaging and public health intertwine (or at least how people think they will intertwine).

I would appreciate an open discussion of the relative pluses and minus of the various vaccines, knowing full well this discussion will have zero impact on which shot gets stuck in my arm.

[House P]

Prevention of hospitalizations and deaths were not included as PRELIMINARY outcomes. Both were included as primary outcomes, in every trial, even if the media coverage of those trials focused on prevention of infection.

https://www.vox.com/22273502/covid-v...fficacy-deaths

As far as I can tell, both Moderna and Pfizer reported hospitalized patients in their FDA submissions -- but they didn't rprovide statistics regarding the magnitude of benefit (VE) or degree of uncertainty (confidence interval) associated with the benefit. It appears that Moderna reported 3 hospitalized patients in the placebo group and 0 in the vaccine group. Pfizer reported 2 hospitalizations in the placebo group and 0 in the vaccine group. I am not a statistician, but I suspect that the confidence intervals would be large based on these sample sizes.

I would welcome a rigorous analysis of the best estimate and degree of uncertainty we have regarding the vaccines' effectiveness for preventing hospitalization and death. I am not sure the Atlantic article provided such an analysis. However, my takeaway from the article was that we don't yet know how close to 100% the vaccines will be for preventing hospitalization/death (so we may wish to be cautious about statements which imply that they are 100% effective).

On the other hand, I can also understand why the simple, compelling message of "100% effective" may be better than the message "likely almost 100% effective" from a health policy perspective, but that is a different discussion.

[mph]

Prevention of hospitalizations and deaths were not included as PRELIMINARY outcomes. Both were included as primary outcomes, in every trial, even if the media coverage of those trials focused on prevention of infection.

https://www.vox.com/22273502/covid-v...fficacy-deaths

I may be misunderstanding something but the article you linked seems to say the exactly opposite and supports the distinction made in the Atlantic article.

Every clinical trial defines in advance a primary clinical endpoint: the outcome researchers are primarily focused on studying. The study is designed to have a large enough sample size to detect differences in this primary outcome between the trial group and the placebo group. For the vaccine trials, the primary clinical endpoint has largely been symptomatic Covid-19 infection (however mild or severe), or, in some cases, a positive Covid-19 PCR test.

It makes sense for the studies to set that endpoint. Designing a study to measure rare events like hospitalization or death is much harder than designing a study to measure relatively common events like infection. Covid-19 is a scary disease, but less than 1 percent of people who get it will die from it. That means that a study designed with death from Covid-19 as a primary clinical endpoint would need to be enormous, with potentially hundreds of thousands of participants. Pinning down the exact frequency of rare events requires a lot of data, and vaccine developers could only include so many people in their first clinical trials because recruiting and coordinating participants is expensive and time-consuming.

So instead, most studies primarily measure how many cases of Covid-19 the vaccines prevent. They do collect data on hospitalizations and deaths, but it’s not the primary outcome their studies are designed to measure, and that has been reflected in press releases, media coverage, and the popular understanding of how well the vaccines work.

[House P]

I may be misunderstanding something but the article you linked seems to say the exactly opposite and supports the distinction made in the Atlantic article.

My understanding is that, regardless of whether data on hospitalizations or deaths were collected, the studies were not designed/powered to identify a statistically significant reduction in deaths/hospitalizations. As such, they also weren't designed to identify whether any reduction in hospitalizations/deaths was greater than, the same as, or less than the 95% reduction in symptomatic cases.

There may be other reasons why we would expect a certain level of performance for reducing hospitalizations/deaths, but the FDA studies weren't designed to evaluate this with any meaningful level of confidence.

[rsvman]

I may be misunderstanding something but the article you linked seems to say the exactly opposite and supports the distinction made in the Atlantic article.

If you think about it, though, it is a much bigger hurrdle for the vaccine to prevent symptomatic infection than it is to prevent hospitalization and death. And the patients who are hospitalized with or die from covid are a subset of all patients who get symptomatic covid. It's not like there is this mystery group that gets hospitalized or dies from covid who is not counted as symptomatic covid.
Therefore, making symptomatic covid the primary outcome is actually a taller hurdle. So, if anything, it biases the studies in favor of the vaccines, not against them. Because any vaccine that successfully prevents symptomatic covid is automatically also going to prevent hospitalization and death, as those outcomes are nothing more than a more severely affected subset of the primary outcome.
I understand that because serious infection and death represents a minority of cases, strictly speaking the studies weren't powered to prove scientifically that they prevent that outcome, but it is not biologically plausible to imagine that they would not, given that they prevent symptomatic disease very effectively.

[Bostondevil]

Uhm, you folks do realize that these studies are ongoing - yes? Data are still being collected on the people in these trials. Due to the nature of the pandemic, there was pressure to look at things early which in this case is a good thing.

[rsvman]

Uhm, you folks do realize that these studies are ongoing - yes? Data are still being collected on the people in these trials. Due to the nature of the pandemic, there was pressure to look at things early which in this case is a good thing.

I can't speak for the others, but yes, I am aware.

I respect you a great deal, but i can't really get behind the tone of this post. Maybe I misinterpreted it, but it comes off a little condescending.

[Bostondevil]

I can't speak for the others, but yes, I am aware.

I respect you a great deal, but i can't really get behind the tone of this post. Maybe I misinterpreted it, but it comes off a little condescending.

I wasn't responding to your post. It was mostly in response to the comment about the studies lacking power to detect a difference in hospitalizations which, early on yes, but as we get more data, no.

[sagegrouse]

Monday over Monday, deaths (7-day ave.) are down 19 percent and new cases down 15 percent.

[mph]

Uhm, you folks do realize that these studies are ongoing - yes? Data are still being collected on the people in these trials. Due to the nature of the pandemic, there was pressure to look at things early which in this case is a good thing.

I don’t know what you read in my post that necessitated the snark, but, yes, I’m aware. You posted a link that, in the context of your post, appeared to be offered as evidence of the specific claim that hospitalizations and deaths were “primary outcomes” in the vaccine trials. The article does not support that claim.

I doubt anyone would deny that data collected from ongoing vaccinations will result in more reliable estimates of hospitalization and death rates over time—a point the Atlantic article never denies. Both the Atlantic and Vox articles use Israel as the best source of current data and both agree that the data is very encouraging. But, Israel is using Pfizer and Moderna which are similar vaccines with similar mechanisms. Which brings us back to the original point of the Atlantic article which is, at the moment, neither the stage 3 trials nor the real world data support claims that all of the vaccines are essentially equal in preventing hospitalizations and deaths. That seems like a reasonable and persuasive and point to me, but it seems clear that you and I just see this issue differently.

[CDu]

I don’t know what you read in my post that necessitated the snark, but, yes, I’m aware. You posted a link that, in the context of your post, appeared to be offered as evidence of the specific claim that hospitalizations and deaths were “primary outcomes” in the vaccine trials. The article does not support that claim.

I doubt anyone would deny that data collected from ongoing vaccinations will result in more reliable estimates of hospitalization and death rates over time—a point the Atlantic article never denies. Both the Atlantic and Vox articles use Israel as the best source of current data and both agree that the data is very encouraging. But, Israel is using Pfizer and Moderna which are similar vaccines with similar mechanisms. Which brings us back to the original point of the Atlantic article which is, at the moment, neither the stage 3 trials nor the real world data support claims that all of the vaccines are essentially equal in preventing hospitalizations and deaths. That seems like a reasonable and persuasive and point to me, but it seems clear that you and I just see this issue differently.

The Israel data most certainly are NOT the best source of data on anything after 28 days... at least not yet. They eventually will be. But as of the preprint of the Israel study, very few of the 600,000 participants had reached 40 days. The phase 3 trials for Pfizer and Moderna dwarf the sample size of Israel’s data beyond 35 days, and those trials are the only source with 100+ days of follow up (Israel maxed at 42 days). This is an important point, as the separation increases with time since second dose. So the Israel data currently in preprint is woefully underpowered and underfollowed relative to either the Pfizer or Moderna trial data.

In a few months, Israel’s data will be the best source. But it is not there yet.