In a decision that will likely save many lives, Biden Administration officials have indicated they will stop holding back second doses and vaccinate as many people as possible as soon as possible. Second doses will be provided as soon as manufacturing catches up. Here’s CNN:
President-elect Joe Biden will aim to release nearly every available dose of the coronavirus vaccine when he takes office, a break with the Trump administration’s strategy of holding back half of US vaccine production to ensure second doses are available.
Releasing nearly all vaccine doses on hand could quickly ratchet up the availability of coronavirus vaccines by allowing more people access to a first dose. It could also be a risky strategy as both Pfizer/BioNTech and Moderna’s vaccines require two doses, administered at specific intervals, and vaccine manufacturing has not ramped up as rapidly as many experts had hoped.
There’s a small risk, but in my view the potential gains outweigh the risks by more than 10 to 1. And more vaccines seem to be on the way.
I see this an another big win for the Tabarrok/Cowen school of epidemiology. (As well as other bloggers like Scott Aaronson.)
READER COMMENTS
Dylan
Jan 8 2021 at 12:55pm
Scott,
I’m happy to see this and think the situation calls for some risk taking, but I am curious how you feel confident to quantify what those risks are? Some of the smartest people I know work in the field and they have no idea how to quantify the risk/benefit based on the information we have. Most I’ve spoke with seem to think this is the right call as well, but they are pretty darn humble about how much we know. This isn’t meant as disrespect, but why should I trust a macro economist on risk assessment in this area?
Kevin Dick
Jan 8 2021 at 2:01pm
I’m curious as to what “in the field” means? If you mean epidemiology, virology, or immunology, these are the wrong people to ask.
My graduate degree is specifically in decision analysis. There is a pretty standard way of approaching this type of decision problem (you get input from subject matter experts of course). And I finished in 1991, so I imagine the field has advanced considerably.
The federal or any state government could have paid a pittance to my old department or any of the similar departments in universities across the US to be on standby to analyze these types of large, uncertain covid policy decisions.
Dylan
Jan 8 2021 at 3:18pm
Sorry, didn’t mean to turn that into an appeal to authority. The people I’m referring to are just people that have done a lot of work in drug development in various areas. The one thing that gets hammered in pretty quickly in that career is, it doesn’t matter what you think you know about how something is supposed to work, 90% of the time you’re going to be wrong.
Mark Z
Jan 8 2021 at 3:37pm
Of course, the ‘engineer’ response is that you don’t need to know how something works to know that it works. It seems there are many cancer drugs which the number show have clinical benefit, but about which the biologists and chemists who developed them were wrong about why: https://stm.sciencemag.org/content/11/509/eaaw8412?intcmp=trendmd-stm
Sometimes I think we treat domain scientists (e.g. biochemists) as the most relevant experts in situations where perhaps we should treat statisticians as the most relevant experts.
Brian
Jan 8 2021 at 4:35pm
But I think the problem here is that if the trials tried really hard to convince people to be available to get the second shot, then we don’t have statistics on single shot case subjects.
Dylan
Jan 8 2021 at 5:00pm
I worded that poorly, wasn’t trying to make a claim about how something works so much, as if it will work at all. You;’re right, that many drugs (not just in cancer) have unclear mechanisms of action. They got developed because someone noticed something maybe working and later trials confirmed that it does indeed work, but we don’t know exactly why. Sometimes we figure it out later other times we don’t.
But lots more things seem like they should work. We have what seems like a clear understanding of the biological pathways, they are confirmed with pre-clinical and animal data that shows that they are indeed working the way we expect them to and have the intended result in animal models of the disease. Then a lot of the time they seem to work when we give it to small patient populations. And then we ramp that up in a Ph III trial and it all just disappears and we don’t know why.
And, btw, most of the industry people I talk to are stats people to one degree or another, since there job is often to design clinical trial protocols.
Scott Sumner
Jan 8 2021 at 2:04pm
I don’t think you should “trust” me, but I have been a long time critic of the medical establishment, and specifically medical ethics. Thus I’ve advocated the creation of a market for kidney transplants, which is opposed by the medical establishment. I’ve looked at the arguments they’ve used to oppose a kidney market, and these arguments are very weak. I believe that some day the market will exist.
Of course vaccines are an entirely different issue, but I use this as an example of why I don’t trust the medical establishment. On March 1, I did a post questioning the medical establishment’s view of masks for the general public. A few months later the establishment basically admitted that I was right and they were wrong. So that’s another reason to be skeptical of medical experts.
Then you have critics like Alex Tabarrok and Tyler Cowen, who have hammered away on one issue after another for the past 12 months. In almost every case, Alex and Tyler have turned out to have been right and the establishment was wrong. Look at their very prescient critique of the way that Covid tests were being regulated, for instance.
There are studies by economists that suggest the FDA is much more worried about errors of commission than errors of omission, in a way that costs lives.
So yes, we should be skeptical of everyone, including economists. And there is some chance that I will turn out to be wrong on this issue. But I’d say that so far economists are coming out of this looking a lot better than medical bureaucrats.
Dylan
Jan 8 2021 at 3:14pm
Thanks for the reply, but it doesn’t really answer the question. What I’m curious about is the quantification element, how you go about calculating a 10 to 1 gain to risk ratio? What probabilities are you assigning to the various potential outcomes?
What percent do you assign to the possibility that vaccination pushes development of a vaccine resistant strain, and it takes us another 3 years before we develop a vaccine for that? Are you basing your estimate on any understanding of the science of virology or from conversations with people who do understand it? From this side, it reads as if the estimate was mostly pulled from the air based on gut feeling.
I don’t want to imply that economists (and others!) don’t have anything to add to the discussion. Economists in particular are well trained to think through costs and benefits in a way that is kind of foreign to lots of people…but that doesn’t mean they don’t need input from others to help fill in all the missing pieces in their knowledge.
Maybe I’ve listened to too much Russ Roberts, but I like to encourage us all to be humble in what we think we know.
Chris
Jan 8 2021 at 4:21pm
Tyler links to the models to support his insistence on First Doses First and then goes on to ask opponents to show their work.
You don’t have to know what the risk is – you have to find where the assumptions break down where it is better to wait – what assumptions are those models making that are wrong? And does it matter?
Dylan
Jan 8 2021 at 5:08pm
Well, I’m barely qualified to be a layman on this, and overall I agree with Tyler’s post, but where I think he doesn’t really give enough credence is the idea of vaccination increasing the probability of escape mutants. I do agree with him that if that is your worry, you should show your math. I’d just like to see someone who works with this stuff for a living provide those estimates, rather than back of the envelope calculations that don’t even seem to include this term (at least if I’m reading things right).
Scott Sumner
Jan 8 2021 at 6:21pm
Not vaccinating also increases the risk of mutants, as we’ve learned recently. AFAIK, there is no scientific consensus on which approach imposes the greater mutation risk.
Mutants are likely to be much more easily addressed with vaccines than the first virus. In addition, evidence suggests that existing vaccines continue to work against the recent mutation.
I’ve also read that if a new vaccine is necessary, we can tweak the RNA approach in such a way as to develop and test a new vaccine vastly faster than the first one.
But yes, it was mostly a guesstimate, based on various things I’ve read.
Again, I believe the overriding challenge is getting through the next three months; after that we’ll have vastly more capacity to address this problem, if we are smart.
Justin
Jan 9 2021 at 12:55pm
Dylan,
I think the risk of a dangerous vaccine resistant strain that we can’t create a new vaccine against is probably very low (<2%). My understanding is that the vaccine allows the immune system to recognize COVID-19’s spike protein, which is how the virus attaches to and infects our cells. To become vaccine resistant, the spike protein itself would have to change, but that change would likely make the virus lose the ability (or dramatically hinder the ability) to infect our cells. Also, if a new variant was found, we could easily conjure up a new variation of the COVID-19 mRNA vaccine to target some feature of that virus. It would likely not require nearly as much testing to get through as it’s a new version of an existing vaccine, similar to the annual flu vaccine.
I also wonder about the math and long term risks. If you have 30 million doses, is it better to give 30 million people 50% protection or 15 million 95% protection? I think you’d need a dynamic computer model to sort that one out. Also, if people go without the second dose for 2 or 3 months, do they risk just losing all of their protection whereas those with both shots will still have it? Maybe these questions have been answered, I’m not sure.
Justin
Jan 9 2021 at 1:00pm
One other issue about ‘first dose first’ is about how the typical patient will respond behaviorally. If people have a false sense of greater security, they may behave in a more risky fashion, going out and about more often. There was a story in Newsweek yesterday about a Nurse who received her first shot was later diagnosed with COVID and specifically said that she felt a false sense of security due to the first shot. From the article, it sounds like the nurse ended up infecting much of her family as well. It’s quite possible that had she not been given a shot at all, she would have continued to be more cautious and maybe wouldn’t have caught COVID.
Very hard to calculate the behavioral risk and see whether it’s enough to focus on two shots or not.
Jerry Brown
Jan 8 2021 at 2:22pm
What CNN says the Biden administration will do is not exactly what Alex and Tyler have been advocating- but it is closer to that. And I think it is a good policy to use as many shots that are currently available as soon as possible and trust that supply for the second shots will be forthcoming.
Some other proposals that Alex and Tyler have made are more risky in my opinion and I don’t agree with them. Like giving half doses to twice as many people. It might turn out that would be just fine, but that is not something that I would trust an economist on.
Anyways, they have been very vocal about this and I think they deserve some credit if not quite a school of epidemiology being named for them.
Scott Sumner
Jan 8 2021 at 6:23pm
Keep in mind that they’ve been right about a wide range of Covid issues, not just this. Look at their early posts on testing. Or challenge studies. Or subsidizing manufacturing.
Jerry Brown
Jan 9 2021 at 2:43am
They have been right about a lot of things in my opinion. And I applaud them. Just not sure on some of it. And really, nobody is sure about some of this.
But getting that first dose out to as many as you can seems to be a really good idea.
Dale Doback
Jan 9 2021 at 1:54am
We have short term solutions to stop spreading Covid before herd immunity. The 1 dose vs 2 dose framing of this argument obscures this fact. Admittedly people aren’t following these short term solutions well, but we have solid evidence that they work. We also know the long term solution of mass vaccination in 2 timed doses works. I would advocate that we follow known solutions and not risk screwing this up.
If evidence emerges and 1 dose goes well in UK then by all means switch it up.
MarkW
Jan 9 2021 at 9:57am
The next thing they need to do is switch to a something like a first-come-first-serve approach so providers can stop wasting doses to make sure nobody gets one out of order. So, for example, give priority to at-risk folks for appointments on a given day, but if they’re not all taken, fill the remaining slots from a ‘standby list’. First doses first won’t solve the current problem of not dispensing all of the doses available.
Steve
Jan 9 2021 at 11:29am
Is this really a first doses first policy? Or are states still recommending that people get their second dose at the suggested interval (difference now being that the federal government is no longer holding back any doses, so it’s up to the states to manage their doses to accomplish this)?
Fred_in_PA
Jan 9 2021 at 11:52am
I have no qualifications except reading a lot, coupled with old age forgetfulness, but . . .
I seem to recall that we have end-of-the-conveyor problems with Covid doses being thrown away because they passed their expiration date before we could get them into someone’s arm. My recollection is that there were multiple problems in; (1) recruiting enough patients with today’s specified qualifications, (2) having enough trained staff on a particular day to administer the shots, (3) delays due to the time investment in the requisite record keeping, and even (4) a shortage of appropriate syringes.
If, when the vaccine reaches the north end of the conveyor we throw it on the ash heap, it doesn’t seem to make much sense to speed up our loading it onto the south end.
Scott Sumner
Jan 9 2021 at 5:42pm
If we are physical not able to vaccine people at a more rapid rate, that would be a scandal that is 100 times worse than the recent invasion of the Capitol building (which was pretty bad). Indeed if we are unable to physically vaccine people as fast as the vaccines are being produced it might be the worst American scandal in the past 100 years. I mean, why didn’t we prepare for the vaccine?
My hunch is that we are physically able to vaccinate people much more quickly, we simply are not doing so for bureaucratic reasons, which itself is a pretty bad scandal.
Fred_in_PA
Jan 9 2021 at 9:13pm
Some of what I’m reading has an odor of “learned helplessness” about it. It troubles me that many people in the system seem to be passively waiting for someone else to decide what they should do. (In their defense, our bureaucratic society might very well punish them if they did take unauthorized initiative.)
Fred_in_PA
Jan 9 2021 at 8:56pm
I found the source that I was vaguely recalling above. It’s an article in January 1’s Wall Street Journal here .
J Mann
Jan 11 2021 at 9:38am
My understanding is that a big part of the problem is the logistics of getting everyone signed up in the correct order – it’s tough to get 10 people from Group 1-A there and ready for shots at the time you open the vial, so that slows down the number of vials you can open.
One solution would be to loosen the limitations on who can get the shots, at least for the last few doses in the vial, but then you’ll see shots go to younger people and people with connections to the medical provider, which seems to be better than a slow roll-out if your primary goal is to stop Covid, but seems worse to some people’s feelings of justice.
Comments are closed.