5 ms·

Alex Tabarrok has been arguing in Marginal Revolution for "First Doses First" since mid-December. Back then, media commentators and the medical establishment o

Alex Tabarrok has been arguing in Marginal Revolution for "First Doses First" since mid-December. Back then, media commentators and the medical establishment objected "But the clinical trials were for two doses". Sadly, very few people acknowledged that prudence had a very real cost in additional infections, deaths, risk of mutations, slower economic recovery. The consensus only shifted to "First Doses First" after the faster spreading UK and South African variants emerged, and the risk of inaction became more obvious. If we step back from Covid-19, this kind of one-sided thinking is all too common in daily life. We need more training (at school, university, in media debates) that the choice is never "Do A or not". It is "do A instead of B (and compare the likely costs and benefits of both)". Or "do the first bit of A, learn some more, and do A' or go back to B". Etc. Real options improve expected outcomes!

fallingknife 6y ago
We don't need more training in school. That's where this type of procedural, bureaucratic thinking comes from in the first place.
- stevesimmons 6y ago
  My point, expressed more fully, is: "teach people to design projects with measurable outcomes versus the status quo; calculate an NPV to make your implicit assumptions and biases explicit; look for ways to restructure your project plan into intermediate goals where you can reassess your strategy (aka "real options"). And your projects will, on average, cost less and be more successful." My meta-point is: these are useful skills that can and should be taught more widely. Your argument is essentially "No", is it not? I'm not sure how that helps improve the situation.
  - dalbasal 6y ago
    This is basically the primary curriculum of current management training. The more people "train" in it, the more we get to a stupid "ruled by metrics" world. IRL, you can't actually be scientific outside of certain cases. Bad/fake science can be worse than no science. Fake objectivity is usually worse than subjectivity. The whole things gets bogged down into "searching under the streetlamp" problems. If you can find a problem that fits the solution you (+ most bogus management consultants of the last 10 years) are advocating, it'll work. Mostly, it gets bolted on to whatever people are doing as an additional layer of bullshit. Nonsense analytics + politics.
- dragontamer 6y ago
  Procedural bureaucratic thinking happens to be a very effective way at finding concensus. Make sure everyone speaks their turn, make sure everyone has a voice. That takes bean counting and nitty gritty arguments. No way around that. Robert's rules suggests that everyone at a meeting gets two chances to speak on any subject, no more. A big issue with today's discussion is that a few powerful and full of themself individuals take up all the oxygen in the room, removing other people's voices from consideration. Say your point, then one followup later as other points come up. Then your turn is over and other people talk.
dalbasal 6y ago
I disagree. Without weighing in on the merits of each strategy too much, I think there is/was a case to be made for each. Also, as you say, consensus did shift. It wasn't a totally fixed mindset. In any case, a decision had to be made. There were always going to be dissidents with a case, regardless of the decision. I don't think the problem is actually a decision making problem. The 2021 problem is that we're handling disagreements poorly. Whichever decision got made, there would be some dissent. If you're in a dissident "First Doses First" camp, the other camp is corrupt, pigheaded pseudoscientific bastards that need to go back to school and learn how to think. If/when "the establishment" comes around to your thinking, it's just more proof of this... this article is a case in point. The establishment side (2 doses, in this case), feels obliged to defend their position, often over defending, making things worse. It quickly devolves into rivalrous rage, both trying to delegitimize the other. Laying piles of bodies at each others' feet. Ad hominems. Accusations of quackery,misinformation and attempts to shut the other up. You end up here regardless of the decision taken. All roads seem to lead here atm. >"This is evidence of what I call magical thinking–an undue focus on the clinical trial design as having incantatory power" Sheesh! Magical thinking... Incantatory power? What earns the FDA these insults? In December, they're rolling out two-shot strategy, conforming to the clinical trial design. There is some reason to think that a one-shot first might work better. The "right" answer is complicated, (a) it depends on unknowns that'd need more clinical trials to learn and (b) it depends even more on execution issues related to vaccinated hundreds of millions of people. Different, hard to quantify risks involved in each strategy. Two-three months later, they're have a one-shot strategy as well. Presumably, this has been in the works for the entire period. Meanwhile, Alex makes them out to be witches trying to cure covid with willow bark. It's grossly unfair. Maybe they were wrong and you were right, but Magical thinking & Incantatory powers... that is vitriolic bullshit.
- DiggyJohnson 6y ago
  > I disagree. I think you and GP have nailed it actually. To use the educational perspective from the top comment, I would want this to be the discussion content for a lesson/section/chapter on the principle of the “existence of candidate decisions” (this is the most dense sentence I’ve ever written).
- 6y ago
tshaddox 6y ago
> Sadly, very few people acknowledged that prudence had a very real cost in additional infections, deaths, risk of mutations, slower economic recovery. > The consensus only shifted to "First Doses First" after the faster spreading UK and South African variants emerged, and the risk of inaction became more obvious. But isn’t this exactly what should happen if we’re comparing risks? It seems entirely possible that there was a well-understood risk to giving only 1 shot when the trials were for 2, and that risk was higher than the risk of virus outcomes before the new variants, but then the risk was lower than the risk of virus outcomes after the new variants. I don’t know if this was the case, but it seems very plausible and reasonable. Other things could change these relative risks too, like how quickly we have been able to get vaccinations done.
raverbashing 6y ago
True. While it is important for the advancement and validation of treatments, Evidence Based Medicine and the quest for "perfect RCTs" has become the new cult for a lot of people RCTs are a validating procedure, it doesn't mean that people are having zero protection after one dose (of course it is great if we can quantify that and see if it's worth it or not) As an example, see how for the AZ vaccine the people that got half a dose accidentally at first got (apparently) more protection than from whoever has gotten a first full dose, however, only the full dose was approved (though on the other hand, the J&J vaccine has more virions than the AZ one) > Sadly, very few people acknowledged that prudence had a very real cost in additional infections, deaths, risk of mutations, slower economic recovery. Exactly. Now we just need to convince people that maybe, just maybe, it would be a good idea to not have susceptible populations be Vitamin D deficient during a pandemic. You know, just in case.
- refurb 6y ago
  This criticism of “a quest for perfect RCTs” makes me chuckle. There are so many shit trials out there that sponsors do their best to torture desired results out of its kind of embarrassing. Instead of saying “prudence has real costs” it would be more accurate to say “having reliable information take time”
  - raverbashing 6y ago
    The existence of crap trials does not exclude that people are over-critical sometimes, and people take the crap trials that "validate the null (or preferred) hypothesis" at face value frequently as well.
    - refurb 6y ago
      A big risk I don't see flagged is the risk of losing the public's confidence in the vaccine approval process. People are already skeptical of the safety, despite pretty robust trials. Imagine if they went with one dose and it turned out to be marginally effective? Imagine how many people will just opt out of vaccines entirely?
nojito 6y ago
Advocating first dose first before we got clinical trial data to support it is a gross slap in the face of every potential dose recipient.
- leesalminen 6y ago
  And there’s the overly charged language that makes everything worse.
  - nojito 6y ago
    Using the general public as guinea pigs is not ethically sound at all.
    - leesalminen 6y ago
      Agreed!
bsder 6y ago
> Real options improve expected outcomes! And sometimes they don't. The biggest problem with ALL of these advocates for the "non-trial protocols" is that they want to violate the protocol but nobody wants to put the processes in place to keep an eye out for if something goes wrong. Read that again: None of the "first doses first" are considering that they might be wrong. And, correspondingly, they have no signals or systems to watch for that. Okay, this is one thing that ALL the trials did that you want to short circuit. And that's not cool. For example, what happens if the single dose effects allow a subgroup of people to carry the virus asymptomatically? Oops. You created an immune variant that will evolve to evade all the vaccines as well as lots of spreaders who don't know it. The human body is stupidly unique and you can get really weird effects. This is why we have trials. Derek Lowe (https://blogs.sciencemag.org/pipeline/ https://blogs.sciencemag.org/pipeline/) pointed this out earlier: "first doses first" vs "cache for second doses" has no medical justification--but you're going to have to make a choice anyway. And it's a shitty choice to have to make. But that's okay. Such is life, you sometimes go with what you have and not what you want--just don't cloak it as a medical decision because you aren't backed by the data. And you have to understand that you might be wrong and you have to have something in place to watch for that.
- maxerickson 6y ago
  Many of the earliest voices were calling for single dose trials in December, not just going off trial and doing one dose. I guess you can elide them from your "first doses first" group, but that kind of undermines the whole criticism.
  - bsder 6y ago
    Sure. But that trial takes time and money to run. And you know that the results are going to be inferior. So, the trial is of limited value once you have enough vaccine. This is the kind of thing that the federal government should have stepped in and funded (one-off, unprofitable, time-limited). But, in the US, you had a boorish asshat and his seditious bootlickers in charge who promulgated that Covid was a hoax. You were not going to get something so useful funded by them.

fallingknife 6y ago

We don't need more training in school. That's where this type of procedural, bureaucratic thinking comes from in the first place.

stevesimmons 6y ago
My point, expressed more fully, is: "teach people to design projects with measurable outcomes versus the status quo; calculate an NPV to make your implicit assumptions and biases explicit; look for ways to restructure your project plan into intermediate goals where you can reassess your strategy (aka "real options"). And your projects will, on average, cost less and be more successful." My meta-point is: these are useful skills that can and should be taught more widely. Your argument is essentially "No", is it not? I'm not sure how that helps improve the situation.
- dalbasal 6y ago
  This is basically the primary curriculum of current management training. The more people "train" in it, the more we get to a stupid "ruled by metrics" world. IRL, you can't actually be scientific outside of certain cases. Bad/fake science can be worse than no science. Fake objectivity is usually worse than subjectivity. The whole things gets bogged down into "searching under the streetlamp" problems. If you can find a problem that fits the solution you (+ most bogus management consultants of the last 10 years) are advocating, it'll work. Mostly, it gets bolted on to whatever people are doing as an additional layer of bullshit. Nonsense analytics + politics.
dragontamer 6y ago
Procedural bureaucratic thinking happens to be a very effective way at finding concensus. Make sure everyone speaks their turn, make sure everyone has a voice. That takes bean counting and nitty gritty arguments. No way around that. Robert's rules suggests that everyone at a meeting gets two chances to speak on any subject, no more. A big issue with today's discussion is that a few powerful and full of themself individuals take up all the oxygen in the room, removing other people's voices from consideration. Say your point, then one followup later as other points come up. Then your turn is over and other people talk.

dalbasal 6y ago

I disagree. Without weighing in on the merits of each strategy too much, I think there is/was a case to be made for each. Also, as you say, consensus did shift. It wasn't a totally fixed mindset. In any case, a decision had to be made. There were always going to be dissidents with a case, regardless of the decision. I don't think the problem is actually a decision making problem. The 2021 problem is that we're handling disagreements poorly. Whichever decision got made, there would be some dissent. If you're in a dissident "First Doses First" camp, the other camp is corrupt, pigheaded pseudoscientific bastards that need to go back to school and learn how to think. If/when "the establishment" comes around to your thinking, it's just more proof of this... this article is a case in point. The establishment side (2 doses, in this case), feels obliged to defend their position, often over defending, making things worse. It quickly devolves into rivalrous rage, both trying to delegitimize the other. Laying piles of bodies at each others' feet. Ad hominems. Accusations of quackery,misinformation and attempts to shut the other up. You end up here regardless of the decision taken. All roads seem to lead here atm. >"This is evidence of what I call magical thinking–an undue focus on the clinical trial design as having incantatory power" Sheesh! Magical thinking... Incantatory power? What earns the FDA these insults? In December, they're rolling out two-shot strategy, conforming to the clinical trial design. There is some reason to think that a one-shot first might work better. The "right" answer is complicated, (a) it depends on unknowns that'd need more clinical trials to learn and (b) it depends even more on execution issues related to vaccinated hundreds of millions of people. Different, hard to quantify risks involved in each strategy. Two-three months later, they're have a one-shot strategy as well. Presumably, this has been in the works for the entire period. Meanwhile, Alex makes them out to be witches trying to cure covid with willow bark. It's grossly unfair. Maybe they were wrong and you were right, but Magical thinking & Incantatory powers... that is vitriolic bullshit.

DiggyJohnson 6y ago
> I disagree. I think you and GP have nailed it actually. To use the educational perspective from the top comment, I would want this to be the discussion content for a lesson/section/chapter on the principle of the “existence of candidate decisions” (this is the most dense sentence I’ve ever written).
6y ago

tshaddox 6y ago

> Sadly, very few people acknowledged that prudence had a very real cost in additional infections, deaths, risk of mutations, slower economic recovery. > The consensus only shifted to "First Doses First" after the faster spreading UK and South African variants emerged, and the risk of inaction became more obvious. But isn’t this exactly what should happen if we’re comparing risks? It seems entirely possible that there was a well-understood risk to giving only 1 shot when the trials were for 2, and that risk was higher than the risk of virus outcomes before the new variants, but then the risk was lower than the risk of virus outcomes after the new variants. I don’t know if this was the case, but it seems very plausible and reasonable. Other things could change these relative risks too, like how quickly we have been able to get vaccinations done.

raverbashing 6y ago

True. While it is important for the advancement and validation of treatments, Evidence Based Medicine and the quest for "perfect RCTs" has become the new cult for a lot of people RCTs are a validating procedure, it doesn't mean that people are having zero protection after one dose (of course it is great if we can quantify that and see if it's worth it or not) As an example, see how for the AZ vaccine the people that got half a dose accidentally at first got (apparently) more protection than from whoever has gotten a first full dose, however, only the full dose was approved (though on the other hand, the J&J vaccine has more virions than the AZ one) > Sadly, very few people acknowledged that prudence had a very real cost in additional infections, deaths, risk of mutations, slower economic recovery. Exactly. Now we just need to convince people that maybe, just maybe, it would be a good idea to not have susceptible populations be Vitamin D deficient during a pandemic. You know, just in case.

refurb 6y ago
This criticism of “a quest for perfect RCTs” makes me chuckle. There are so many shit trials out there that sponsors do their best to torture desired results out of its kind of embarrassing. Instead of saying “prudence has real costs” it would be more accurate to say “having reliable information take time”
- raverbashing 6y ago
  The existence of crap trials does not exclude that people are over-critical sometimes, and people take the crap trials that "validate the null (or preferred) hypothesis" at face value frequently as well.
  - refurb 6y ago
    A big risk I don't see flagged is the risk of losing the public's confidence in the vaccine approval process. People are already skeptical of the safety, despite pretty robust trials. Imagine if they went with one dose and it turned out to be marginally effective? Imagine how many people will just opt out of vaccines entirely?

nojito 6y ago

Advocating first dose first before we got clinical trial data to support it is a gross slap in the face of every potential dose recipient.

leesalminen 6y ago
And there’s the overly charged language that makes everything worse.
- nojito 6y ago
  Using the general public as guinea pigs is not ethically sound at all.
  - leesalminen 6y ago
    Agreed!

bsder 6y ago

> Real options improve expected outcomes! And sometimes they don't. The biggest problem with ALL of these advocates for the "non-trial protocols" is that they want to violate the protocol but nobody wants to put the processes in place to keep an eye out for if something goes wrong. Read that again: None of the "first doses first" are considering that they might be wrong. And, correspondingly, they have no signals or systems to watch for that. Okay, this is one thing that ALL the trials did that you want to short circuit. And that's not cool. For example, what happens if the single dose effects allow a subgroup of people to carry the virus asymptomatically? Oops. You created an immune variant that will evolve to evade all the vaccines as well as lots of spreaders who don't know it. The human body is stupidly unique and you can get really weird effects. This is why we have trials. Derek Lowe (https://blogs.sciencemag.org/pipeline/ https://blogs.sciencemag.org/pipeline/) pointed this out earlier: "first doses first" vs "cache for second doses" has no medical justification--but you're going to have to make a choice anyway. And it's a shitty choice to have to make. But that's okay. Such is life, you sometimes go with what you have and not what you want--just don't cloak it as a medical decision because you aren't backed by the data. And you have to understand that you might be wrong and you have to have something in place to watch for that.

maxerickson 6y ago
Many of the earliest voices were calling for single dose trials in December, not just going off trial and doing one dose. I guess you can elide them from your "first doses first" group, but that kind of undermines the whole criticism.
- bsder 6y ago
  Sure. But that trial takes time and money to run. And you know that the results are going to be inferior. So, the trial is of limited value once you have enough vaccine. This is the kind of thing that the federal government should have stepped in and funded (one-off, unprofitable, time-limited). But, in the US, you had a boorish asshat and his seditious bootlickers in charge who promulgated that Covid was a hoax. You were not going to get something so useful funded by them.