A fourth wave may hit in March, when the more contagious B117 strain from the UK takes over. Expect more shelter-in-place orders, school shutdowns, and a spike in cases at least the size of July's, maybe December's. That will last until May-ish, when the usual control system (more virus -> stricter lockdowns -> less virus -> looser lockdowns -> more virus) moves back into the "less virus" stage. Also coronavirus is seasonal and summer isn't its season. Also by that time a decent chunk of the population will be vaccinated. The worst consequences of the UK strain should burn themselves out by late spring.
Prediction: 75% chance that there will be a new wave peaking in March or April, with a peak at least half again as high as the preceding trough.
[EDIT: some people link new studies saying the B117 strain is less virulent than previously believed, and the US has been getting much better at vaccination since I checked, probably my prediction above is too high and we should worry less about this]
We should also be concerned about a fifth wave (possibly overlapping with the fourth wave; they may not have obviously separate peaks). Virologists have identified two new strains, one in South Africa, one in Brazil, which probably have "immune escape" - the ability to infect people who have already gotten, recovered from, and developed antibodies to the original strain (or been vaccinated against it). Both strains already have a few cases in the US. It will take them a few months to spread to the point where they're relevant, but they should eventually be the majority of new cases.
[EDIT: some people link new studies saying the B117 strain is less virulent than previously believed, and the US has been getting much better at vaccination since I checked, probably my prediction above is too high and we should worry less about this]
We should also be concerned about a fifth wave (possibly overlapping with the fourth wave; they may not have obviously separate peaks). Virologists have identified two new strains, one in South Africa, one in Brazil, which probably have "immune escape" - the ability to infect people who have already gotten, recovered from, and developed antibodies to the original strain (or been vaccinated against it). Both strains already have a few cases in the US. It will take them a few months to spread to the point where they're relevant, but they should eventually be the majority of new cases.
Prediction: 66% chance that sometime this year, the South African and Brazilian strains - or other new strains with similar dynamics - will be a majority of coronavirus cases in the US.
Some sources describe these strains as "vaccine resistant". This is a matter of degree. The UK strain is probably very slightly vaccine-resistant (most sources are describing it as not vaccine resistant, but if you look closely this is another "well we can't prove it is" situation, and the best point estimates suggest some tiny amount of extra resistance which probably doesn't make a big difference.). The South African strain is significantly vaccine resistant. The Brazilian strain is too new to know much about, but seems to be very similar to the South African strain and I would be surprised if its numbers differed very much.
In terms of preventing sympomatic infections, the best current data suggests that the Novavax vaccine is 96% effective against Coronavirus Classic, 86% effective against UK, and 60% effective against South Africa. AstraZeneca is something like 80% effective against Classic, 65% effective against UK, and the South African study was kind of bungled but our best guess is "seems pretty bad". Johnson and Johnson is 66-72%+ effective against Classic and 57% effective against South Africa. Pfizer/Moderna hasn't been tested against South Africa in real life yet, but lab studies suggest slightly decreased efficacy.
The good news is that vaccines which protect inconsistently against infection are probably still good at protecting against severe disease and death. For example, although the J&J vaccine is only 66-72% effective at preventing people from getting symptomatic disease, it's 85% effective at preventing severe disease, and (at least so far in studies) 100% effective at preventing deaths. In fact, most vaccine studies have shown 100% efficacy at preventing deaths. Probably some of this is that the trials are underpowered to detect rare outcomes, but the vaccines really do seem good at this, even with strains that have some level of vaccine resistance. Also, although I don't know of any studies investigating this, it makes sense to think that vaccinated people would also be less likely to transmit the virus to others if they do get it.
Some sources describe these strains as "vaccine resistant". This is a matter of degree. The UK strain is probably very slightly vaccine-resistant (most sources are describing it as not vaccine resistant, but if you look closely this is another "well we can't prove it is" situation, and the best point estimates suggest some tiny amount of extra resistance which probably doesn't make a big difference.). The South African strain is significantly vaccine resistant. The Brazilian strain is too new to know much about, but seems to be very similar to the South African strain and I would be surprised if its numbers differed very much.
In terms of preventing sympomatic infections, the best current data suggests that the Novavax vaccine is 96% effective against Coronavirus Classic, 86% effective against UK, and 60% effective against South Africa. AstraZeneca is something like 80% effective against Classic, 65% effective against UK, and the South African study was kind of bungled but our best guess is "seems pretty bad". Johnson and Johnson is 66-72%+ effective against Classic and 57% effective against South Africa. Pfizer/Moderna hasn't been tested against South Africa in real life yet, but lab studies suggest slightly decreased efficacy.
The good news is that vaccines which protect inconsistently against infection are probably still good at protecting against severe disease and death. For example, although the J&J vaccine is only 66-72% effective at preventing people from getting symptomatic disease, it's 85% effective at preventing severe disease, and (at least so far in studies) 100% effective at preventing deaths. In fact, most vaccine studies have shown 100% efficacy at preventing deaths. Probably some of this is that the trials are underpowered to detect rare outcomes, but the vaccines really do seem good at this, even with strains that have some level of vaccine resistance. Also, although I don't know of any studies investigating this, it makes sense to think that vaccinated people would also be less likely to transmit the virus to others if they do get it.
Prediction: 55% chance that later, when we have great evidence on this, we’ll find that P/M, Novavax, AZ, and J&J all cut deaths from all extant strains by at least four-fifths.
When the fifth wave strikes in late spring/early summer, some of the population (~50%?) will be vaccinated, another part of the population (~25%?) will have had the disease already, and the rest (~25%?) will be completely vulnerable. The new strains will probably cause a limited number of mild cases among the vaccinated/resistant, and a larger number of more severe cases among the vulnerable. Either way, the presence of the larger vaccinated/resistant contingent could potentially make this less severe than previous waves. Also, we may have learned more about treating severe COVID (with eg ivermectin, fluvoxamine), which might further decrease deaths. (...)
R in most US states right now is closely clustered around 1. Mutant strains are more contagious, enough to bring the R0 up to 1.5 or so. But having a lot of the population vaccinated will bring it back down again. Also, I'm acting like there's some complex-yet-illuminating calculation we can do here, but realistically none of this matters. It's not a coincidence that all US states are closely clustered around 1. It's the control system again - whenever things look good, we relax restrictions (both legally and in terms of personal behavior) until they look bad again, then backpedal and tighten restrictions. So we oscillate between like 0.8 and 1.2 (I made those numbers up, I don't know the real ones). If vaccines made R0 go to 0.5 or whatever, we would loosen some restrictions until it was back at 1 again. So unless we overwhelm the control system, R0 will hover around 1 in the summer too, and the only question is how strict our lockdowns will be.
In autumn, if we haven’t already vaccinated everyone there’s a risk things will get worse again because of the seasonal effect. Also, for all we know maybe the virus will have mutated even further and become even more vaccine resistant. Now what?
Vaccine companies say it should be pretty easy to create a vaccine targeted to the South African strain. Remember, it only took them two days to invent the original coronavirus vaccine. This one should be even easier, since we already know the principles involved. The vaccine is basically taking a part of the coronavirus' chemical code which functions as a "password" and telling it to the immune system so it can break its password and defeat it. The mutant coronaviruses haven't done anything fancy, they've just changed their password. The vaccine companies can plug in the new password to the vaccines they already have, and they'll work against the mutant strains.
But even if they have it tomorrow, that's...what? Another four months for studies, one month before the FDA is able to meet to discuss an approval (you can't rush meetings!), two months to ramp up production, and five months of Distribution Hell while we argue about who should be first in line and prosecute people for distributing vaccines too quickly. So maybe by this time next year you get a vaccine against the South African strain. And by that point the virus will have just changed its password again and we'll be right back where we started.
The problem is, all the virus has to do is change its chemical "password" - a simple one-step process. The people fighting the virus have to go through the entire FDA approval, production, and distribution pipeline each time - a seven million step process. This puts us at a bit of a handicap.
Best-case scenario, here's how we respond:
When the fifth wave strikes in late spring/early summer, some of the population (~50%?) will be vaccinated, another part of the population (~25%?) will have had the disease already, and the rest (~25%?) will be completely vulnerable. The new strains will probably cause a limited number of mild cases among the vaccinated/resistant, and a larger number of more severe cases among the vulnerable. Either way, the presence of the larger vaccinated/resistant contingent could potentially make this less severe than previous waves. Also, we may have learned more about treating severe COVID (with eg ivermectin, fluvoxamine), which might further decrease deaths. (...)
R in most US states right now is closely clustered around 1. Mutant strains are more contagious, enough to bring the R0 up to 1.5 or so. But having a lot of the population vaccinated will bring it back down again. Also, I'm acting like there's some complex-yet-illuminating calculation we can do here, but realistically none of this matters. It's not a coincidence that all US states are closely clustered around 1. It's the control system again - whenever things look good, we relax restrictions (both legally and in terms of personal behavior) until they look bad again, then backpedal and tighten restrictions. So we oscillate between like 0.8 and 1.2 (I made those numbers up, I don't know the real ones). If vaccines made R0 go to 0.5 or whatever, we would loosen some restrictions until it was back at 1 again. So unless we overwhelm the control system, R0 will hover around 1 in the summer too, and the only question is how strict our lockdowns will be.
In autumn, if we haven’t already vaccinated everyone there’s a risk things will get worse again because of the seasonal effect. Also, for all we know maybe the virus will have mutated even further and become even more vaccine resistant. Now what?
Vaccine companies say it should be pretty easy to create a vaccine targeted to the South African strain. Remember, it only took them two days to invent the original coronavirus vaccine. This one should be even easier, since we already know the principles involved. The vaccine is basically taking a part of the coronavirus' chemical code which functions as a "password" and telling it to the immune system so it can break its password and defeat it. The mutant coronaviruses haven't done anything fancy, they've just changed their password. The vaccine companies can plug in the new password to the vaccines they already have, and they'll work against the mutant strains.
But even if they have it tomorrow, that's...what? Another four months for studies, one month before the FDA is able to meet to discuss an approval (you can't rush meetings!), two months to ramp up production, and five months of Distribution Hell while we argue about who should be first in line and prosecute people for distributing vaccines too quickly. So maybe by this time next year you get a vaccine against the South African strain. And by that point the virus will have just changed its password again and we'll be right back where we started.
The problem is, all the virus has to do is change its chemical "password" - a simple one-step process. The people fighting the virus have to go through the entire FDA approval, production, and distribution pipeline each time - a seven million step process. This puts us at a bit of a handicap.
Best-case scenario, here's how we respond:
by Scott Alexander, Astral Codex Ten | Read more:
Image: Johns Hopkins
