3 ms·

The blog post is mainly about the difficulty of producing vaccines. I don't see much in it that is objectionable. But this line may be misleading: So no, we di

The blog post is mainly about the difficulty of producing vaccines. I don't see much in it that is objectionable. But this line may be misleading: So no, we did not “have the vaccine” in February. The author is arguing that, because it's difficult to produce of the Moderna vaccine, we couldn't have disseminated it on a wide scale in February or shortly thereafter. That argument may be fine, but it diverts attention from the critical role that the U.S. regulations in general, and the U.S. FDA in particular, played in slowing the process of widespread vaccine dissemination. Specifically: the process could have been sped by months if the U.S. had been willing to conduct human-challenge trials [1] and to integrate tests of safety (Phase 2) and efficacy (Phase 3) [2]. To my knowledge, understanding of how to integrate the safety and efficacy trials has only recently developed; perhaps the FDA shouldn't be faulted on that front. But the benefits of human-challenge trials have long been understood. [1] https://doi.org/10.1093/infdis/jiaa152 https://doi.org/10.1093/infdis/jiaa152 [2] https://doi.org/10.1198/016214504000001790 https://doi.org/10.1198/016214504000001790

angry_octet 6y ago
Firstly, it wasn't produced in large quantities in February. They have to do trials against the mass produced article, not lab samples. Secondly, it would be extremely unethical to overlap phase 2 & 3 trials, when there are effective NPIs. The lessons of unethical medical science, like Tuskegee and experimentation on armed forces personnel, must not be ignored. And yet that is precisely what your ref Eyal[1] is proposing, because practically none of the safeguards can be achieved. Efficacy trials require a much larger participant group, given the still low infection rates. I don't see how one can justify giving a candidate to 100k people before safety screening, especially when the downside risk is significant risk of death if they ever catch COVID.
- jbullock35 6y ago
  > Firstly, it wasn't produced in large quantities in February. They have to do trials against the mass produced article, not lab samples. Yes, I’m not disputing this point. My claim is that even after we account for the difficulties of mass production, the FDA-controlled licensing process should have taken less time than it did. > Efficacy trials require a much larger participant group, given the still low infection rates. This is the major benefit of human-challenge trials [1]: if you intentionally infect healthy, well-informed, low-risk volunteers with low doses of a virus, the size of the requisite participant group falls dramatically. Note also that “Any volunteers in whom infection was confirmed would receive excellent care for COVID-19, including priority for any scarce life-saving resources, in state-of-the-art facilities.” You can say that it’s unethical to run such trials. But it seems hard to make that argument if the volunteers are informed about the risks and if safeguards are in place (e.g., limiting the dose to which volunteers are exposed). Consider also the cost -- many thousands of lives -- of refusing to run such trials and thus delaying the licensing of a vaccine by months. > The lessons of unethical medical science, like Tuskegee [...] must not be ignored. And yet that is precisely what your ref Eyal[1] is proposing No, not at all. In the Tuskegee experiment, people infected with syphilis were promised medical care that they were later denied. When they were diagnosed with syphilis, they weren't even told of the diagnosis. (One of the purposes of the experiment was to study what happens when syphilis isn't treated.) These conditions are a world apart from what Eyal et al. are proposing in [1]. You don't need to take my word for it: the article is short and easy to read. [1] https://academic.oup.com/jid/article/221/11/1752/5814216 https://academic.oup.com/jid/article/221/11/1752/5814216
- orangecat 6y ago
  Secondly, it would be extremely unethical to overlap phase 2 & 3 trials, when there are effective NPI. The NPIs have enormous costs in both dollars and quality of life, and are in fact not terribly effective.
foobiekr 6y ago
We have a vaccine in February in the same way that we have 5nm fabs this year. After all, all the technology is there, it's just the investment and those pesky actual manufacturing process that's standing in the way.
- bluGill 6y ago
  Not exactly. If you have a 5nm fab you can produce chips today that you know work. (the yield might be terrible, but you can get some functional chips every day). In February what we had was something that completed the process, but we had no idea if it worked at all.
  - foobiekr 6y ago
    From a manufacturing perspective, the equipment was there because they've been working on it 20Y; whether the vaccine worked is a separate question. By analogy, it's one thing to have a fab, a working fab can make 5nm chips as designed at some yield, however, the fab is not responsible for whether the chip designs work or are valid. My point here really was that tech people, and especially people in the software industry, ignore that the bulk of the work isn't the equipment, it's the long tale of supply chain, manufacturing process and QA. All of the 5nm equipment exists today, it's the other parts that are the hold up. Essentially, in our own domain we know better, and we should really understand this and stop making this mistake in other domains.

angry_octet 6y ago

Firstly, it wasn't produced in large quantities in February. They have to do trials against the mass produced article, not lab samples. Secondly, it would be extremely unethical to overlap phase 2 & 3 trials, when there are effective NPIs. The lessons of unethical medical science, like Tuskegee and experimentation on armed forces personnel, must not be ignored. And yet that is precisely what your ref Eyal[1] is proposing, because practically none of the safeguards can be achieved. Efficacy trials require a much larger participant group, given the still low infection rates. I don't see how one can justify giving a candidate to 100k people before safety screening, especially when the downside risk is significant risk of death if they ever catch COVID.

jbullock35 6y ago
> Firstly, it wasn't produced in large quantities in February. They have to do trials against the mass produced article, not lab samples. Yes, I’m not disputing this point. My claim is that even after we account for the difficulties of mass production, the FDA-controlled licensing process should have taken less time than it did. > Efficacy trials require a much larger participant group, given the still low infection rates. This is the major benefit of human-challenge trials [1]: if you intentionally infect healthy, well-informed, low-risk volunteers with low doses of a virus, the size of the requisite participant group falls dramatically. Note also that “Any volunteers in whom infection was confirmed would receive excellent care for COVID-19, including priority for any scarce life-saving resources, in state-of-the-art facilities.” You can say that it’s unethical to run such trials. But it seems hard to make that argument if the volunteers are informed about the risks and if safeguards are in place (e.g., limiting the dose to which volunteers are exposed). Consider also the cost -- many thousands of lives -- of refusing to run such trials and thus delaying the licensing of a vaccine by months. > The lessons of unethical medical science, like Tuskegee [...] must not be ignored. And yet that is precisely what your ref Eyal[1] is proposing No, not at all. In the Tuskegee experiment, people infected with syphilis were promised medical care that they were later denied. When they were diagnosed with syphilis, they weren't even told of the diagnosis. (One of the purposes of the experiment was to study what happens when syphilis isn't treated.) These conditions are a world apart from what Eyal et al. are proposing in [1]. You don't need to take my word for it: the article is short and easy to read. [1] https://academic.oup.com/jid/article/221/11/1752/5814216 https://academic.oup.com/jid/article/221/11/1752/5814216
orangecat 6y ago
Secondly, it would be extremely unethical to overlap phase 2 & 3 trials, when there are effective NPI. The NPIs have enormous costs in both dollars and quality of life, and are in fact not terribly effective.

foobiekr 6y ago

We have a vaccine in February in the same way that we have 5nm fabs this year. After all, all the technology is there, it's just the investment and those pesky actual manufacturing process that's standing in the way.

bluGill 6y ago
Not exactly. If you have a 5nm fab you can produce chips today that you know work. (the yield might be terrible, but you can get some functional chips every day). In February what we had was something that completed the process, but we had no idea if it worked at all.
- foobiekr 6y ago
  From a manufacturing perspective, the equipment was there because they've been working on it 20Y; whether the vaccine worked is a separate question. By analogy, it's one thing to have a fab, a working fab can make 5nm chips as designed at some yield, however, the fab is not responsible for whether the chip designs work or are valid. My point here really was that tech people, and especially people in the software industry, ignore that the bulk of the work isn't the equipment, it's the long tale of supply chain, manufacturing process and QA. All of the 5nm equipment exists today, it's the other parts that are the hold up. Essentially, in our own domain we know better, and we should really understand this and stop making this mistake in other domains.