A model from the University of Washington predicts that new COVID-19 cases and deaths will drop to zero shortly after July 1: https://covid19.healthdata.org/projections

I assume that's because the virus runs out of new people to infect, if people who have had it before can't get it again, and if we assume it doesn't mutate.

But then I don't understand what is the point of developing a vaccine that won't be ready for 12-18 months anyway. What does it achieve, if all of the deaths from COVID-19 will have already occurred by this July? And if the reason the deaths level off in July is because so many people have the antibodies by then that we have herd immunity, then who is the vaccine for?

I understand the virus could mutate so that people who have antibodies from the current version will not be immune. But if the eventual vaccine is going to be based on the current version, doesn't that mean that if it mutates, people who get the vaccine won't be immune either?

I can think of some possible answers that would make sense (by which I only mean they would make logical sense as an answer; I don't mean that they make scientific sense):

1) The UW model represents a rogue view and people are working on the vaccine because they don't think it's correct.

2) Even if we achieve herd immunity at 75%, the vaccine is to protect the remaining 25%. (I would call this a special case of #1, contradicting the UW model, because it seems to assume there is a significant number of preventable deaths lying to the right of July 2020.)

3) People who get the SARS-CoV-2 virus will only be immune to the current version, but the vaccine can be developed in a way that it might also provide immunity against other mutated versions.

4) The vaccine is not going to be ready in time to make a difference against the current SARS-CoV-2, but we're working on a general strategy for identifying any mutated strains that might appear next year, so that we can respond quickly with a vaccine, similar to what we do with the flu.

5) Even though we think our bodies will retain the antibodies from SARS-CoV-2 indefinitely, there's a chance that they might not (similar to norovirus), so the vaccine is to cover that possibility.

Or something else?

  • 3
    "I assume that's because the virus runs out of new people to infect, if people who have had it before can't get it again, and if we assume it doesn't mutate." - You assume this. Did the modelers?
    – Bryan Krause
    Mar 30, 2020 at 0:44
  • @BryanKrause I am not assuming that; I am assuming they assumed it :) If the UW model shows no new cases after mid-July, doesn't that mean they're assuming that the virus did not mutate? Otherwise wouldn't it keep infecting people?
    – Bennett
    Mar 30, 2020 at 0:52
  • You can read their paper to learn about the assumptions they do make: healthdata.org/sites/default/files/files/research_articles/2020/… I think it's kind of rude to ask a question like you have here without even reading the resource you're quoting from.
    – Bryan Krause
    Mar 30, 2020 at 0:53
  • @BryanKrause I did read the paper; it does not explicitly state whether they are assuming the virus will or will not mutate. Did you not even click the link yourself to see if they mention mutation or not?
    – Bennett
    Mar 30, 2020 at 2:04
  • I don't know why you are talking about mutation, maybe I wasn't clear: I was referring to the broader assumption in the sentence I quoted: that it levels off because the virus runs out of people to infect. That is not at all what their model is based on, and in the beginning of the paper they mention that models based on waiting for enough people to get infected to build herd immunity would result in at least an order of magnitude more deaths than they predict. Their paper is based on the effects of social distancing and isolation based on the course in Italy and Wuhan.
    – Bryan Krause
    Mar 30, 2020 at 2:06

2 Answers 2


You've misinterpreted the modelling you refer to above. The modelling is based on the premise that the USA population adopts assuming full social distancing through May 2020. This means that the majority of the USA population will remain susceptible to infection because the social distancing has kept them away from the virus.

Furthermore, there are millions of people in other countries that will need vaccination. Take for instance China where the majority of the population remains immune naive because they have shut down the pandemic inside China.

Furthermore, the risks of future coronavirus pandemics remains high. The main reason the SARS vaccine was not developed was because they ran out of people to test the vaccine on. With a SARS-Cov-2 vaccine that will help the development of new vaccines. The influenza vaccine is able to be developed yearly because the expertise is there, but we lack that expertise at present with the novel coronaviruses.

  • OK. But that doesn't imply what I said in point #1 -- that there is a significant number of potential infections and deaths to the right of July 2020 on the timeline, and those are the ones that will be eventually prevented by the vaccine? Would it be correct to say that if you divide society into "strictly isolating" and "everybody else", then the curve that hits zero in July 2020 only depicts the "everybody else" segment, but we still want to minimize deaths after July 2020 in the "isolating" segment?
    – Bennett
    Mar 30, 2020 at 3:56
  • Vaccine scientists will do their own modelling, and what their employers tell them to do. The curve only applies to those people who can be affected. If you're isolated then clearly you can never be infected. Mar 30, 2020 at 4:04
  • I will note if you are agreeing that the virus will eventually sweep through a significant percentage of the non-isolating population, this goes against one of the assumptions in the paper -- the paper rejects the premise that "given current estimates of the basic reproductive rate... 25% to 70% of the population will eventually become infected... Based on reported case-fatality rates... millions of deaths in the United States". (In that case I think you're right and they're wrong; that the virus will infect >25% of people, but hopefully the infection fatality rate is less than CFR.)
    – Bennett
    Mar 31, 2020 at 3:26
  • Also, "If your isolated then clearly you can never be infected" -- but then why would you need the vaccine? :) I'm just trying to come up with a consistent set of assumptions that would lead to conclude that the vaccine benefits someone. If you're 100% isolated, you don't need it; if you're not 100% isolated, then you'll probably come in contact with the virus in the next 18 months before the vaccine is ready. Are you assuming someone might be 100% isolated for the next 18 months, but come into contact with society after that, and that's who the vaccine benefits?
    – Bennett
    Mar 31, 2020 at 6:53
  • China will control its borders until a vaccine is developed. That means everyone who isolated is allowed out and about without the risk of meeting someone virulent. NZ will try the same; we have a good chance of eradicating the virus inside our country. Anyone coming into the country will need to tested to make sure they're not carrying. Mar 31, 2020 at 6:56

If by the model being "correct", you mean that the model will correctly predict what happens, then my understanding is it's mostly your first point. But if you consider a model correctly predicting what would happen if its assumptions turn out to be accurate, but those assumptions turn out to be false, then that's a separate points. Other points:

It's downright foolish to not work on all possible responses simply because you think they won't be needed. Even if people thought that there was, say, a 60% chance of cases going to zero by July, it would be silly to say "Well, there's only a 40% chance we'll need this vaccine, so no point working on it." When you're facing the possibility of millions of deaths, the probability of that doesn't need to be very high for it to warrant working to address.

It doesn't take 18 months to develop a vaccine. It took a few weeks to create a vaccine, and then several more months to test it, and it will take several months after that to deploy it.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.