1 Or 2 Vaccine Doses For People Who Had COVID?

[AsherO]

An antibody test can reveal something. Politicians are opposed to it. Just mandate vaccines. The more me merrier.

Maybe, maybe not. It boggles my mind that there isn’t much data on antibody levels of reinfected patients at time of infection. The cynics will say that such research is being suppressed for political reasons, I just think it’s hard data to isolate (recovered people who tested for antibodies recently before reinfection)

[Euclid]

I just think it’s hard data to isolate (recovered people who tested for antibodies recently before reinfection)

Yep, especially since (accurate) antibody testing requires a blood test and not something people would agree to do often.

[ExGingi]

Maybe, maybe not. It boggles my mind that there isn’t much data on antibody levels of reinfected patients at time of infection. The cynics will say that such research is being suppressed for political reasons, I just think it’s hard data to isolate (recovered people who tested for antibodies recently before reinfection)

The latter is due to the former. If there was interest in this (like the interest in getting vaccines approved) the data would be there.

#TheMostPoliticizedVirusInHistory

[biobook]

https://archive.is/RPzua

Haaretz article in Hebrew discussing antibody tests, last paragraph seems to indicate that medical experts say no reason for shot after six months of recovery. More likely to be protected for 10-12 months after recovery.

If I understood it correctly, it's not that medical experts say no reason for shot after six months of recovery, but rather that some medical experts say that, while some do recommend the shot after six months. 

I would guess that the reasoning of the latter group of experts is that if protection lasts 10-12 months on average, then you would want to provide the vaccine before protection begins to wane.  If you think of the length of protection for different people as falling along some sort of bell curve, some percentage will have a shorter than average length of protection, so a shot at 6 months would likely help the group of recoverees as a whole, even if it is not yet necessary for every individual within that group.

[biobook]

An antibody test can reveal something. Politicians are opposed to it. Just mandate vaccines. The more me merrier.

The latter is due to the former. If there was interest in this (like the interest in getting vaccines approved) the data would be there.

When you asked about this a few weeks ago, I answered that everyone would like this information, but the antibody tests we have cannot yet reveal whether someone is immune to infection.  You just posted an article from Haaretz that says the same thing, quoting medical personnel, not politicians. 

Some people with high antibody levels may still get infected.  Some people with low antibody levels may have better protection due to non-antibody factors, such as T cells.  We just don't know yet whether there is a certain level of serum antibodies that can be used to determine immunity. 

It's true that we can get a sense of immunity to measles by measuring antibody levels, but measles antibodies have been studied for decades, and COVID antibodies for less than two years.

Which part of this do you find confusing?

[Moshe123]

All the reinfections I heard of didn't have much antibodies left.

[biobook]

All the reinfections I heard of didn't have much antibodies left.

Can you expand on that?  When were they infected the first time?  When did they measure antibodies?  How much is not much?  How old? 

[biobook]

Maybe, maybe not. It boggles my mind that there isn’t much data on antibody levels of reinfected patients at time of infection. The cynics will say that such research is being suppressed for political reasons, I just think it’s hard data to isolate (recovered people who tested for antibodies recently before reinfection)

Do the math. 

August 2020 we were discussing reports of about 20 documented reinfections world-wide.  At that time, about 26 million covid cases had been reported, so let's say the chance of reinfection was 1 in a million.

August 2021 we discussed the article from Israel which looked at reinfections among members of the Maccabi kupat cholim.  It has not been published in peer-reviewed form, but it's all we've got to go on.  It was confusing, and I hope I got the numbers here right.
https://forums.dansdeals.com/index.php?topic=123776.msg2491027#msg2491027

They looked at people who'd had COVID in Jan-Feb 2021, and had a second COVID infection by the end of June, that is, about 5 months later.  Out of 16,000 cases, there were 19 reinfections, so about 1 out of 1000.

When they included people who'd had COVID earlier, from Feb 2020-Feb 2021, and had a second COVID infection by the end of June, that is, 5-17 months later, they found a higher reinfection rate.  Out of 46,000 cases, there were 108 reinfections, or about 1 out of 500.

I've rounded off numbers here, and not sure I got this right, but I think the order of magnitude is right.  That is, among people who've recovered from COVID, something like 1/500 to 1/1000 were getting a second case of COVID by the end of June. 

This was before the delta peak, so let's take the higher number, and assume 1 out of every 500 COVID recoverees will be infected a second time with COVID this month.  Once they're reinfected, they'll naturally increase production of antibodies, so if we want to know their level of antibodies BEFORE that infection, we would need to measure antibodies of 500 recoverees in order to find ONE person who is reinfected.  But to draw reliable conclusions, we would need to find more than one.  If we want to reach a sample size of 20 reinfections, we would have to examine 20x500, or about 10,000 recoverees, in order to find those 20 who will be reinfected. 

For each of those 10,000, we'd need to measure there antibodies frequently - perhaps once or twice a week - so as to catch their antibody levels before the infection.  As Euclid mentioned, it's not easy to get 10,000 people to agree to give frequent blood samples for something like this.

Instead of studying 10,000 people for one month, we could study 1000 people for ten months, but of course, that's longer than we want to wait for an answer. 

The point is that it TAKES TIME to get an answer to the question of whether a particular antibody level can be used to determine a person's immunity to covid.  Some recent studies show a different chemical in the blood of recoverees that seemed to correlate with immunity, so that's another direction the research is taking.
 

[4yourinfo]

Can you expand on that?  When were they infected the first time?  When did they measure antibodies?  How much is not much?  How old?

Don't you think with the millions of people getting infected there should be one person who recently tested for antibodies and got covid? I tested my antibody levels a month apart and it dropped from 39 to 33 I don't see why you need people to get tested 2x weekly for your study.. Seems to me getting an antibody test monthly can get a pretty good picture at what point your antibodies are not effective won't be exact but don't you think it's pretty good? Once a month not that crazy - how about we get some volunteers right here?

[gozalim]

Don't you think with the millions of people getting infected there should be one person who recently tested for antibodies and got covid? I tested my antibody levels a month apart and it dropped from 39 to 33 I don't see why you need people to get tested 2x weekly for your study.. Seems to me getting an antibody test monthly can get a pretty good picture at what point your antibodies are not effective won't be exact but don't you think it's pretty good? Once a month not that crazy - how about we get some volunteers right here?

not that many volunteers here are willing to risk the PCR test when they get symptoms

[biobook]

I tested my antibody levels a month apart and it dropped from 39 to 33 I don't see why you need people to get tested 2x weekly for your study.. Seems to me getting an antibody test monthly can get a pretty good picture at what point your antibodies are not effective won't be exact but don't you think it's pretty good?

Biological systems have a lot of variability, so if antibodies drop 6 points over a month in one person, they might drop 12 points in another person, and rise 10 points in a third person.  So we can't conclude that the change you saw in yourself is necessarily going to generalize to everyone else.

But to your more important point, yes, you're probably right that antibodies could be measured less frequently than twice weekly, and get results that are less precise, but would still give you a general idea of what particular range of antibody levels are needed for protection.  But you would still need to study a large number of individuals, for a long period of time.

Don't you think with the millions of people getting infected there should be one person who recently tested for antibodies and got covid?

Yes!  There was one!  And they DID have measurable antibodies just before getting infected.

Actually, there was more than one.  I posted a study that included some relevant data:

I've said before that there's no point to testing for antibodies, because we don't know what level of antibody is needed for immunity, that is, to prevent a covid infection.  The recent report from Israel provides a bit of relevant data.  This is the study that was in the news for showing a small number of breakthrough infections in vaccinated people.  For a few of those breakthrough cases, the researchers had taken blood in the week before or on the first day of a positive covid test, so they could say something about what antibody level was insufficient to prevent a covid infection.  Those who got infections had lower antibodies ON AVERAGE than those who didn't, but the difference was not statistically significant. 

Covid-19 Breakthrough Infections in Vaccinated Health Care Workers
https://www.nejm.org/doi/full/10.1056/NEJMoa2109072

Scroll down to Figure 2, where the thick black horizontal line shows that the average antibody levels for breakthrough cases are lower than the average for controls.  But the dots and squares show the values for the individuals in each of those groups, and there's clearly a lot of overlap.

Don't ask me to explain the article, which I didn't understand myself.  Here are some reader-friendly summaries:
https://www.nature.com/articles/d41586-021-02096-3
https://www.cidrap.umn.edu/news-perspective/2021/07/israeli-study-finds-26-covid-breakthrough-infection-rate

Here is one part of Fig. 2, see the linked article for legend and other details.  Each  circle is antibody level of one person who was positive for covid after being vaccinated, and colored circles were symptomatic, and each gray square is antibody level of matched control who did not get covid.  All people were health care workers, so exposed to covid patients.  Note that y-axis is logarithmic. 

[AsherO]

Thanks Biobook, I recently wondered if such data existed.

The x-axis is clear to me, higher means higher titers. What does the y-axis mean, if a data point is further to the right it is more what?

[biobook]

Thanks Biobook, I recently wondered if such data existed.

The x-axis is clear to me, higher means higher titers. What does the y-axis mean, if a data point is further to the right it is more what?

The x axis is the horizontal one.  I think you're asking about, for example, in the breakthrough cases, just above the heavy black line, there's a row of color-color-color-gray, and you're asking what it means that the gray is to the right.  Is that what you're asking?  I'm pretty sure the answer is it doesn't mean anything.  They just needed to put all those data points in the same place, and since it would be unclear to pile them on top of each other, they just spread it out.  Perhaps they decided to put symptomatic cases first, then asymptomatic.  And when they got to the row just beneath that, they realized they had one more symptomatic case, so added that at the end. 

[biobook]

On second thought, it could be that they drew the dots quantitatively, that is, the row of color-color-color-gray are around the 300 mark, but maybe they're actually in some kind of order, like 290-310-320-330.  IDK, but effectively it's not meaningful.

The point that I think the graph shows is that among those who got symptomatic infections, some had antibody levels in the 100-1000 range, and those with asymptomatic infections had antibodies as high as about 10,000.  These levels are within the same range as those who didn't get infected, so based on this, we can't say that antibody levels will give us a good idea of who will or won't get infected.

Note that this is the more relevant neutralizing antibody, measured in lab research.  The IgG antibody measured in clinical lab tests show a similar relationship, see the figure in the paper.

[ExGingi]

Not sure if this is the best thread for this, but since the "conclusion" drawn is that people who had Covid-19 should get vaccinated, I guess this is an appropriate thread.

https://www.cdc.gov/mmwr/volumes/70/wr/mm7044e1.htm?s_cid=mm7044e1_w

I didn't read the full PDF report, but I am rather skeptical of the findings in the summary. With Crown Heights and other Jewish neighborhoods that mostly have natural immunity we don't hear of much (if any) hospitalizations of people who recovered from covid-19, while we do hear of hospitalizations and deaths of people who only had a vaccine.

[Moshe123]

Not sure if this is the best thread for this, but since the "conclusion" drawn is that people who had Covid-19 should get vaccinated, I guess this is an appropriate thread.

https://www.cdc.gov/mmwr/volumes/70/wr/mm7044e1.htm?s_cid=mm7044e1_w

I didn't read the full PDF report, but I am rather skeptical of the findings in the summary. With Crown Heights and other Jewish neighborhoods that mostly have natural immunity we don't hear of much (if any) hospitalizations of people who recovered from covid-19, while we do hear of hospitalizations and deaths of people who only had a vaccine.

It's all modeled garbage. Not real actual data.