27 Comments

This is somewhat tangential, but I have a broader version of the usual "But what about Long COVID?" question, partially inspired by the likelihood of many more breakthroughs likely to stem from Omicron.

I'm curious if you have a sense of why the topic is often not addressed when talking about headline COVID risks - both in the popular media and often among experts. There's obviously plenty of mentions and articles about it, but I would say the (vast) majority don't mention it or mention it as an afterthought.

I personally don't have a great answer, but my intuitive sense is that it likely means that the more serious cases aren't as common as some estimates have suggested. I.e. if longterm strong versions of Long COVID (e.g. severe chronic fatigue & brainfog) were quite common (say in 10% or even 30% of cases) then we would be hearing about it MUCH more from many corners and with increased urgency.

An alternative explanation could be that the lack of good data makes it hard to write a good story about it but I'm not fully convinced by that one.

Expand full comment

To me, if the virus shows a prolonged ability to outrun our vaccines, Long COVID in breakthrough cases is the most critical unknown. We're going to be able to suppress severity of acute disease even if we get hit with a whole series of Omicrons, but we might end up in the worst case in a condition where everyone can expect to suffer a breakthrough case every two or three years. If even a small percentage lead to serious long-term consequences it could be a big deal. We really need to know what Long COVID is and how frequent it is in breakthrough cases for people in different conditions and age groups. If there's not much or if it always resolves after a year or so, that's one thing, if we start accumulating people with long-term disabilities, that's another.

Expand full comment

I think doing a quality study of the effect of vaccination on Long COVID is of high importance. SARS-CoV-2 has a fairly robust ability to infect tissues outside of the respiratory tract, and in this it more resembles measles and chicken pox than flu. Measles, especially, can lead to long-term symptoms including suppression of the immune system with damage to the immune system repaired over one or two years. SARS-CoV-2 will target cells expressing Ace2 (which is different from measles) so it's logical it could cause long-term problems, but likely with a different profile.

It seems to me that maintaining high antibodies (or boosting them after infection) is far easier in the bloodstream than in the upper respiratory tract, and that therefore immunization would be massively more effective against Long COVID than against mild breakthrough infections. Even if antibody levels had dropped a secondary immune response will raise them much faster and higher than a primary immune response to initial infection. But this is all gut feelings rather than based on evidence, so I'm fairly concerned about claims that Long COVID is common after breakthrough infections, but also not sure if real long-term harm has been demonstrated as a common consequence. I have heard of one person known to some of my relatives to have a breakthrough infection after vaccination, and then to suffer severe fatigue for months afterwards. But we need good statistics including how many uninfected people suffer similarly for different reasons, especially at advanced ages.

The problem with existing studies is that many look at fatigue, loss of taste/smell, other symptoms starting as early as five weeks after infection which really muddies things. If we start looking at six months, we can be more confident that we are seeing unusual and concerning symptoms rather than simple slow recovery. Right now, I just saw a study using 5 weeks as a cutoff for Long COVID. That is fine, but they should use a different name related to slow recovery or something like that.

I'm not sure this matters for boosters. As far as I'm concerned, boosters for younger people is entirely about maintaining community protection, which also protects older people, some of whose immune systems may fare poorly even after boosting.

Expand full comment

+1 to Can's comment. At some point, the absence of evidence does seem to suggest that long covid is not nearly as prevalent as initially feared (which is great news).

I don't doubt that there are people with long-term effects, and individual cases sound very frustrating in a medical environment that doesn't work well for symptoms without a clear cause. That said, the worry about long covid also seems to be a major reason that many people are still very concerned about getting infected (even mild infections after vaccination + booster). As Can said, this worry is likely to increase if Omicron is able to cause more infections in vaccination people, even if the acute symptoms are mild.

I would love to see a post digging more into the known and unknowns about long covid at this point, to help understand how to think about its risks in the context of everything else we know.

Expand full comment

I *really* don't want to minimize the risk to older people etc. but societally I think it's important to know whether the risk of serious longterm effects for a triple vaccinated healthy person under say 50 is extremely small or actually significant. And to add to my first post I would argue that the non-booster advocates are effectively assuming the former.

Expand full comment

One of my more silly thoughts on this is that I feel it should be possible for anyone with some time on their hands to get a vague guesstimate for this very easily. Find some list of people who've had COVID, the level of randomization/sophistication and size can depend on how accurate you want to be, and call them and ask if they have *serious* Long COVID. I feel calling even 30-50 would be somewhat useful, a few hundred would be very much so. It sounds overly basic but when the best estimates are "10-50% of people have some form of Long COVID of various severity and length" it's really not hard to do better. Using some kind of official database would be great, but honestly reasonably social people in e.g. NY could probably just ask all their FB friends to re-share it and answer (e.g. Nate Silver said 50% of people he knows have had COVID so he could probably just do it himself if he wanted to).

Expand full comment

Disposable N95 respirators aren't that great; they generally provide a poor fit and that makes them prone to leaks. If everyone wore the things, this leakiness wouldn't be a problem, since they'd still prevent aerosols from accumulating indoors despite leaks. Unfortunately, most people aren't going to wear them (thanks to the CDC/WHO/most public health officials).

Fortunately, way better PPE exists in the form of elastomeric respirators. Elastomerics offer better fit, filtering (N100 filters), and comfort than N95s. They last forever too. Their only disadvantage is that they muffle speech somewhat.

https://www.eurekalert.org/news-releases/584008

https://www.ncbi.nlm.nih.gov/books/NBK540078/#effi1

https://www.doi.org/10.1016/j.jamcollsurg.2020.05.022

https://www.doi.org/10.1001/jama.2008.894

Expand full comment

If South Africa is only 25% vaccinated and the hospitals aren't overwhelmed yet, wouldn't that support a "more mild" hypothesis?

Expand full comment
author

South Africa had big outbreaks and by some estimates, almost unvaccinated everyone has prior infection. Hard to tell for sure, but the number is definitely high. So it is not a population without immunity.

Expand full comment

Here's a study from Malawi which suggests very high previous exposure in another African country (and almost 100-fold undercounting of cases). The numbers will be an underestimate since: 1) antibodies are difficult to detect after many asymptomatic cases especially many months after infection, 2) the study ended early in the Delta wave, so most Delta infections would have been missed.

https://bmcmedicine.biomedcentral.com/articles/10.1186/s12916-021-02187-y

Expand full comment

It is still way too early to make any kind of assessment on Omicron. Remember that pretty much every optimistic scenario for Covid has been wrong. Vaccines were amazingly efficacious on wild strain covid, but those times are long past.

Not to say it will be a disaster, but anybody with any degree of caution should bunker down, get boosted, double mask, avoid as many social interactions as possible, be prepared for another winter of shelter in place.

I don't have any confidence that our health care bureaucracies or government agencies have learned anything from this debacle, so take any recommendations with a grain of salt.

Oh, and have a Merry Christmas

Expand full comment

Within the United States, Omicron isn't going to be much of a factor before Christmas. It would be surprising if it makes up as many as 1% of cases in the United States by Christmas.

If you're in a place where Delta is surging (New England, the Great Lakes, Kansas, New Mexico, Arizona), then you definitely want to be taking all the precautions against Delta. And if you haven't been boosted yet, then getting boosted now will prepare you for when Omicron does come. But in the parts of the United States where Delta is not currently surging (the South, the Northwest, California), we can take basic precautions for now while seeing friends and family for the holidays, and do the deeper hunkering down as January approaches.

Expand full comment

The CDC says Omicron is at 3% as of December 11th. It went from 0.4% to 2.9% in one week. That's almost 3 doublings. If we conservatively assume two doublings per week going forward, it's about two more weeks to 50%. Coincidently two weeks from 12/11 is Christmas. The wonder of exponential growth. https://covid.cdc.gov/covid-data-tracker/#variant-proportions

Expand full comment

Hmm, I should look more into that! I might be behind the times, and with exponential growth, a week off can be a big deal. I'm still a bit skeptical of those numbers, because there's a lot of intentional over-sampling of travelers from Africa in recent sampling. But that over-sampling can cut either way in terms of over-estimating or under-estimating the growth rate.

Expand full comment

Thanks for this, the available media can be confusing, and as always I'm grateful for your help in finding solid resources.

Something I am still confused about, though, and maybe you can point to a good summary, is the effect of variability in the spike protein. The cartoon model of the infection process doesn't really have variability here, it posits a kind of lock-and-key metaphor, where, on the one hand, the spike protein allows the virus to enter the body's cells and do its damage, and/but also antibodies matched to the spike protein allow the immune system to detect and neutralize the virus.

But, as many resources and your text above point out, the emergence of variants complicates this picture. Changes to the spike protein open a gap between what the virus is and what the vaccines alert the immune system to look for, but also maybe open a similar gap between what the virus brings to the fight and what the receptors in your body will accept, and with what degree of enthusiasm.

I guess my question is, are these entirely empirical questions where we have to wait for the data, or is there some history or modeling or something that can provide some clue about how much variability we need to see before it changes the game entirely?

Expand full comment
author

Have you read this? It is admittedly a broad overview, but deals with this question directly. (What you are referring to is called "antigenic evolution." https://www.theinsight.org/p/novelty-means-severity-the-key-to

Expand full comment

I read that at the time, but having re-read it at your prompting, have a better appreciation of the latter section, so thanks! (again!)

The answer seems to be that these are indeed empirical questions, the degree to which variations are more or less immune-evading (either prior adaptive immunity or vaccine-induced adaptive immunity) is largely a matter of waiting and seeing.

Expand full comment

Since boosters are critical to not getting Omicron, I studied one of the upward-ramping bar charts that shows the higher and higher numbers of people who have gotten two shots, or the one of J&J. Then I made a vertical mark at 6 months ago and a horizontal mark across to today. I looked at how many of the additions were within the past 6 months. It looked like about 20%.

Given that we now know that the protectiveness of the Pfizer and Moderna vaccines declines after 6 months anyway, it seems to me that the number to focus on now is vaccinated and, if more than 6 months ago, boosted.

Do any of you know of charts that show either the number boosted, or the number of more recent than 6 (or 2 for J&J) vaccination plus booster?

It seems to me that this is the number we should be focusing on. Ideally we would incorporate those with the initial full vaccination plus a case of covid, but, since we do not have that latter number, I'm looking for the best data we have on what our society's immunity level is now.

Expand full comment

So a study out of South Africa now says 2 dose Pfizer is 70% effective vs hospitalization for Omicron, compared to 93% for Delta:

https://www.reuters.com/business/healthcare-pharmaceuticals/pfizer-vaccine-protecting-against-hospitalisation-during-omicron-wave-study-2021-12-14/

I have a couple of questions about this study that the article doesn't answer:

1. Does it measure drop in conditional P(hospitalized| infected) or in raw P(hospitalized)?

2. Does it correct effectively for the likely increased proportion of "incidental" infections among the hospitalized, i.e. people who are positive for Covid and maybe symptomatic but aren't hospitalized *because* of the severity of their COVID symptoms?

These questions seem more consequential this time around than before and I wish the articles would make them clearer.

Expand full comment

In the best-case scenario in which Sars-CoV-2 remains widespread, lets assume it turns into a mild cold, even for initial infections and that the new strains never cause long Covid. We'll still be able to track it, and will have a vast machinery to do so. After lockdowns and losing relatives, lots of people have phobias. I'm wondering if it could remain a severe problem for many years both for individuals who are scared of COVID and with regard to sticky government regulations, with both likely tied to politics in some complicated way? Thus was not the case after the 1918 flu because it was only decades after that the different strains of flu could be distinguished and tracked.

As a disclaimer, this is a thought experiment with some assumptions that might not be realistic. I was thinking how to turn it into a pure psychology/political/sociological question.

Future severity reduction could well depend entirely on pre-existing immunity and might require widespread vaccination of children. We'll need answers on Long COVID after vaccination breakthrough cases (whether long-term symptoms arise with enough frequency to pose a threat), and for that studies are needed over relatively long timeframes. Also, we'll need to see not much worse threats after breakthrough cases in the vaccinated aged than what nursing homes were dealing with pre-COVID with the various viruses already going around. All these things may or may not veer in the good direction.

Probably, for balance, I should sketch out a worst-reasonable-case scenario, but it's too hard to feel out what the boundaries for this virus are, other than postuating having to frequently update vaccines, lots of nasty breakthrough cases with Long COVID, continuing severe problems in nursing homes, and having to be vigilant with children. I think a Stephen King scenario would require a new virus. The rapid Omicron spread does raise the possibilty though we may need to become really fleet-footed with vaccines as it seems to be able to spread faster than our current ability to update.

Expand full comment

Here's a question from a recent patient I didn't have an answer to:

"I'm a 50 year old, fully vaccinated (no booster) male. My 70 year old mother-in-law with a heart condition is also (with the booster). What is the likelihood I will contract COVID (of any variant) and spread it to her over the holidays?

Expand full comment

If your patient was vaccinated more than 6 months ago (2 for J&J) his likelihood of contracting the virus is much higher than it would be if your patient would get the booster. We know that vaccinated people who get the virus can spread it, so getting boosted if over the recommended time (and not getting the covid at all) is the key. Apparently vaccinated people who get covid can infect others for a shorter period of time.

I would also recommend that he take a covid rapid test the day before the visit, and if it is a multi-day visit, take one every day. I got my test at CVS, which had a better supply than the other stores. The Abbot labs Binax Now is FDA recommended. I don't think all of the varieties on the shelves there were on the recommended list, although some of them were less expensive.

Please thank him for watching out for us elders.

Expand full comment

The usual concept with immune escape is that the virus has optimal binding to start with to a target on the cell it wants to infect. The binding site on the viral protein is a very good antibody target because the antibody will form a physical barrier, making it impossible for the virus protein to bind its target on the cell. So, this part of the virus protein has to be changed to escape antibody targeting, and that's difficult to do without losing some degree of binding, so there could be tradeoffs, with a bias towards becoming milder.

However, what we've seen with Sars-CoV-2 is that each of the subsequent strains through Delta has massively better binding to Ace2 than the original virus. That's still untested, to my knowledge, for Omicron, but it shares mutations with Alpha, Beta, Gamma and Delta that enhance binding to Ace2 in these viruses. So the virus has been able to repeatedly have its cake and eat it too. This suggests to me that Spike might have been poorly optimized for human Ace2 to start with, and while that was the case it was able to improve binding and lose some antibody binding at the same time. That road should be only so long, and after that we'll certainly still see variants appear, but maybe there's hope the tempo will be slower. But this is just guessing on my part.

Expand full comment
author

Omicron seems to be really different than Alpha, Beta and Gamma, though so I don't think we know its fusogenicity yet...

Expand full comment

Disappointing that there were no policy recommendations is only contingent: if the data show X do x' if the date show Y do Y'. For example should companies be working on getting Omicron variant-optimized vaccines approved? If not now under what conditions?

Expand full comment

I feel that much of what is concerning (or confusing) about omicron is clarified by these illustrations of potential relationships between severity, transmissibility/immune evasion, and cases or deaths.

https://twitter.com/nataliexdean/status/1468988174693289994.

https://twitter.com/CFR_org/status/1349760893874278400

https://twitter.com/GosiaGasperoPhD/status/1469456805491138560

Expand full comment
author

I love Natalie's explanations. The other two are good, but I think they applied more to Alpha like situations: just more transmissibility. Here, we are dealing with antigenic shift, so it's not the same: it's not necessarily some percent of the mild will get severe disease, but just more of them, though in the end, with the elderly being most vulnerable, the numbers do add up so the broad point stands.

Expand full comment