There are specific neutralizing antibodies to identified specific SARS-CoV-2 proteins. And those that are non-specific common to the other human coronaviruses that cause common colds frequently and recently in children.
The non-specific ones spike first. Which makes sense. Those who end up with minimal symptoms may never get much measurable specific antibody responses. Those individuals seem to have a stronger cellular (T cell) response. Are they less immune or do they have stronger more lasting immunity? Not yet known but I’d guess the latter.
Needless to say it complicates things. It will be harder to tell for sure how often kids were infected if their infections end up resolving without specific antibodies in their wake and never having gotten sick enough to test when sick.
Still unconfirmed in any definitive way is the apparent lack of children being very contagious when infected. If so understanding why seems important.
Median age in NYC is about 36, which is about on par with the national average. So it isn’t like NYC is a geriatric city like a retirement community in Florida.
Don’t know what this says for the IFR overall, but it has to be higher than 0.1% minimum.
Are there populations where we would expect nearly everyone to have been exposed to it before the lockdown was put in place? E.g.: ships, prisons, hospitals (what else?). The first corona cases to come to those places wouldn’t have been isolated because no one knew how contagious it was, people would just think it was the common flu. If we find a place where we can reasonably presume that 90%+ of the people were exposed, we can see the hospitalization, ICU/ventilators and death rates. We might need to see many of those places but just 1 could give an idea.
As a guess, would military ships be a good place to look?
Aside from the USS Theodore Roosevelt, there is also the Charles De Gaulle aircraft carrier in France. On that one, about half the sailors have tested positive.
You’d assume the R0 would be higher on military ships due to close quarters, shared eating areas, and the fact that sick people likely still have to report to work.
But on the Roosevelt, I believe its closer to 15-20% who got it. not the 50% on the De Gaulle.
Yes, although military ships might give overly optimistic stats given the demographic. Hospitals overly pessimistic. I was thinking prisons might give a good crosssection for a given gender.
Is it possible for someone to have been exposed to a virus and have fought it off so well that they test negative? Like the virus just “bounced off” them?
Any possible reasons why the De Gaulle got 50% while Roosevelt got 15 to 20%?
Possible explanations include how many individuals first seeded the infection, for how long it went on before being noted/responded to (the article notes that some of those ill of the De Gaulle were not thought to be noteworthy because having “15 to 20 sailors with symptoms like coughing every day … at the time was not out of the ordinary”), and specifics to the living conditions.
It is definitely documented for some to have fought of the infection without developing SARS-CoV-2 specific antibodies. Unclear if that means their main process was a cell-mediated response or that non-specific antibodies of sufficient amount responding quickly were sufficient. But it seems to happen more in those with milder infections and/or are younger.
“General Lecointre said on Sunday that officials had considered canceling the trip to Brest but that it was an important logistical stop. Sailors were ordered to comply with social distancing guidelines and to stay away from night clubs and gatherings of over 50 people.”
Alright sailors, you’re getting shore leave but no partying.
It seems like in an aircraft carrier, nearly everyone must have got a sniff. How do you avoid getting exposed when more than half of the people you share a tin can with have it if it’s that contagious?
Something about the asymptomatic cases. In a simplistic view, they should be just as infectious as the symptomatic. We would expect them to be generating symptomatic infections, all the way up to fatal infections. In which case, contact tracing should have turned them up.
Here in Oz we do pretty aggressive contact tracing. Only 10% of known cases cannot be sheeted home to another person. So whilst community infection is real, it has never been anything more than a minor component. This seems inconsistent with any more than a small portion asymptomatic infections.
Here where I live we have managed two days in a row of no new infections detected, despite having opened up testing to essentially anyone that wants one. I guess it is possible that our lockdown has snuffed out the entire asymptomatic infection without it ever having been detected. But that almost all of our known infections were traced back and the entire chain is known - right back to the person that imported the infection, the existence of a significant asymptomatic group just doesn’t fit out numbers.
However, the number, if they are known, of people with asymptomatic infections detected during contact tracing are not published. That is a problem. Whether they were asymptomatic or pre-symptomatic should be known. Even so, the numbers of untraceable contacts doesn’t seem to support there being many.
Unless asymptomatic cases are actually less infectious in some curious manner. One could posit a regime where symptomatic cases are delivering a much higher initial quantity of virus to the infectee, and they are thus more likely to go on to develop a more serious infection, and that asymptomatics deliver a much lighter hit, and the infectee is more likely to be asymptomatic. Essentially creating a multiple tiered set of infections. This seems hard to make stable.
OTOH, the idea that the initial infecting dose is an unusually important predictor of progression of the infection might have some basis.
Maybe because asymptomatics cough/sneeze less? If you have a mild case but it’s not getting out of you much, you’ll be less infectious than if you’re coughing into the air/hands/objects, especially since asymptomatics are likely washing their hands and maintaining social distance.
Less virus in their saliva/mucus which would mean lower probability of infecting others or lower viral load, as you say. It might well be that there could be multiple tiers, that’s an insightful possibility. As you say, we’d expect the asymptomatic tier to whittle down quickly if it’s less infectious as it would act as a kind of variolation.
Not a simplistic view that many experts would ever hold. There is instead a general belief that there is some degree of correlation between peak level of symptoms/virulence and an individual’s contagiousness, with asymptomatic individuals being significantly less contagious than more symptomatic ones. The physiology is that symptoms correlate with the degree to which the virus has taken over a mass of the host’s cells to job of creating more of their copies and spraying them out. It is not so curious.
Most of the models used have picked some arbitrary guess for that. (For example the Imperial College study chose to assume that “symptomatic individuals are 50% more infectious than asymptomatic individuals.”) They can drive spread even if much less contagious still simply because there may be many more of them and, under baseline conditions, having more close contacts because they do not appreciate their need to distance themselves throughout the entire period (whatever it might be) that they are contagious (to whatever, likely varying degree, that actually is).
Note also that this is about “asymptomatic” not “presymptomatic” … who may be even most contagious just before they become more symptomatic. In that regard the following may be of note to you.
Let’s state the same conclusion of the first sentence slightly differently:
For a reproductive number of 2.5, contact tracing and isolation is more likely to be successful if less than 30% of transmission occurred before symptom onset, or if >90% of the contacts can be traced.
Note further that lockdowns, and behaviors that occur by choice before lockdowns, significantly lower the effective reproductive number.
Not sure if any of that helps or serves to only further muddy the issues.
Interesting.
They note that viral load was independent of severity of the disease. But very interesting that peak shedding is in the pre-symptomatic phase. Then again, if they are pre-symptomatic, they won’t be coughing everywhere. Again, there are no asymptomatic cases. They only have stats for those admitted for one assumes reasonably severe illness.
I think our contact tracing managed the >90% part. We have been lucky in that infection has been mostly limited to a few specific groups. One of the groups was the Ruby Princess, our local plague ship. That and the other two groups (air baggage handlers and a tourist group) were easy to trace and accounted for just about everyone. Other parts of Oz have had a mixed bag, but today four states reported no new cases.
Overall I have the feeling there is something reasonably significant yet to be found. The numbers are not yet closing on a complete understanding.
Dirt cheap does not mean they are available in far far larger numbers. You can pay far more for tests to free up money for markets to try to improve supply. The market still takes time to adjust and supply things needed to conduct the test. We are only 18 days from when the first test got emergency FDA approval. We can’t just expect supply to appear out of thin air as soon as the demand is created.
Toilet paper, for example, is also cheap. We did not just invent it two and a half weeks ago. The supply chain is still struggling to keep it in stock weeks after the initial spike in demand.
My hunch is that it’s probably a bit of both: COVID-19 has been spreading longer than thought, and it is likely pretty contagious. NYC is also a densely populated area with (I’m guessing) higher use of public transport and a place in which infected persons can more easily spread it to others.
This seems like it could just be measurement error. Because cases cluster, someone who is in contact with an asymptomatic case is likely also in contact with symptomatic cases. But if you aren’t testing everyone, then you don’t know who’s sick.
Example: A and B, who are both infected, go to a party. A becomes symptomatic a few days later, B never does. Seven people who go to that party later get sick. Now, let’s say in reality 5 of those people were infected by B. He’s a loud talker and a bit of a spittle sprayer.
But B never gets tested because he never gets sick. A eventually gets tested, as do some of the people at that party.
If you do contact tracing with the data you have, you’re going to end up attributing 7 cases to A, and 0 to B. The only people you’re going to attribute to asymptomatic patients are people who only came into contact with asymptomatic patients.
He says 12.5% of the Roosevelt tested positive. That is 12.5% of the total on the ship. But only 94% were tested at the time of the recording. To assume the other 6% are negative seems wrong.
He says this data shows the beauty of having full testing. He claims that we’ll know exactly who to separate and quarantine. But that assumes a perfect test, which we know we don’t have. Allowing someone who tested negative back to normal duty could be doing nothing but exposing the other 4200 on the ship to the infection. This is a big reason why I don’t see how more testing is going to save us and allow the country to open back up, even if we could test everyone everyday.
If they are testing for active virus, there could be many more that are no longer shedding. So the infected rate is, at a minimum, 600/(0.94*4800) = 13.2%