Omicron anecdata?

[Not_a_Number]

19 minutes ago, Ausmumof3 said:

https://www.cidrap.umn.edu/news-perspective/2022/11/kids-similar-risk-long-covid-adults-study-suggests
 

Long covid risk in kids is similar to adults 

“A large study today from Germany shows that kids and adolescents are at the same relative risk of experiencing COVID-19 symptoms 90 days or more after acute infection as adults are, according to findings in PLOS Medicine.”
 

I wish there were more longer term studies. Three months out is not great but could still resolve. Issues that carry over a year or more seem much more worrying to me.

Yes, I want longer term studies, and also more studies for Omicron and for vaccinated people. 

Studies are pretty hard to design at this point. Although I'm really... offended? annoyed? that there isn't a study that just tallies what happened right after the first giant US Omicron wave last December/January. It seems like it wouldn't have been hard to do. I've seen studies like that about the first NYC wave -- just tracking health outcomes immediately after. They are very convincing, because the timing is so right. It makes it obvious that COVID was at fault, even if it wasn't listed as a cause. 

Edited November 11, 2022 by Not_a_Number

[Not_a_Number]

5 hours ago, mommyoffive said:

Repeat COVID is riskier than first infection, study finds (msn.com)

The risk of death, hospitalization and serious health issues from COVID-19 jumps significantly with reinfection compared with a first bout with the virus, regardless of vaccination status, according to a study published on Thursday.

"Reinfection with COVID-19 increases the risk of both acute outcomes and long COVID," said Dr. Ziyad Al-Aly of Washington University School of Medicine in St. Louis. "This was evident in unvaccinated, vaccinated and boosted people."

Reinfected patients had a more than doubled risk of death and a more than tripled risk of hospitalization compared with those who were infected with COVID just once. They also had elevated risks for problems with lungs, heart, blood, kidneys, diabetes, mental health, bones and muscles, and neurological disorders, according to a report published in Nature Medicine.

People with repeat infections were more than three times more likely to develop lung problems, three times more likely to suffer heart conditions and 60% more likely to experience neurological disorders than patients who had been infected only once, the study found. The higher risks were most pronounced in the first month after reinfection but were still evident six months later.

 

 

Catching Covid more than once ‘doubles your risk of death’ (msn.com)

I'd be quite suspicious of this data, because I'd guess that the only second infections that make it onto health records are much worse than normal infections tend to be -- almost everyone I know tested for COVID at home and didn't let their doctor know. So then the "2 infections" data is self-selected, which has the standard self-selection issues. 

[popmom]

6 minutes ago, Not_a_Number said:

I'd be quite suspicious of this data, because I'd guess that the only second infections that make it onto health records are much worse than normal infections tend to be -- almost everyone I know tested for COVID at home and didn't let their doctor know. So then the "2 infections" data is self-selected, which has the standard self-selection issues. 

I was wondering about this, too, but I don’t have the mental energy to look into it. 

[Not_a_Number]

Just now, popmom said:

I was wondering about this, too, but I don’t have the mental energy to look into it. 

It is REALLY hard to run proper studies right now. Unlike before, when there were no tests, most people test with a Binax or equivalent at home. The doctor never knows unless it's a BAD infection and they need Paxlovid or something, which of course self-selects. 

You can also do questionnaires, but then you only get response from like 20%, and that's also self-selected. (I just saw a study like that yesterday!) So that's no good.

Or you can go by app data, but most people don't enter things into the app! I think you can probably see a pattern. 

What we really need is a study where people actually interview everyone and we get a response rate of, like, 70% instead of 20% or whatever the self-selected rate is. But that's much harder to run, and in the meantime, people need to publish or perish... 🙄

[KSera]

22 minutes ago, Not_a_Number said:

I'd be quite suspicious of this data, because I'd guess that the only second infections that make it onto health records are much worse than normal infections tend to be -- almost everyone I know tested for COVID at home and didn't let their doctor know. So then the "2 infections" data is self-selected, which has the standard self-selection issues. 

If you read the study and particularly the analysis and cohort selection methods, the limitations appear to primarily reduce the actual magnitude of effect rather than increase it. The fact that people who didn’t test would end up in the control group  (or the single infection vs reinfection group) would serve to make the control group slightly more like the infected group than it otherwise would have been. They also defined reinfection as a positive test at least 90 days after the previous one, which also would have left out many reinfections (again, reducing the actual effect), since we know those can happen within a month of a previous infections.

I think a bigger problem is that the VA population skews heavily older and male. Although it was compared to a similar VA cohort, so the effects found should hold up within those populations. I would personally like to know if the vaccine’s lack of protection against the long term sequelae holds just as true in younger people with more robust immune responses to the vaccine. I actually fear from what we’ve seen this far that it will 😔. 

[Ausmumof3]

Just now, KSera said:

If you read the study and particularly the analysis and cohort selection methods, the limitations appear to primarily reduce the actual magnitude of effect rather than increase it. The fact that people who didn’t test would end up in the control group  (or the single infection vs reinfection group) would serve to make the control group slightly more like the infected group than it otherwise would have been. They also defined reinfection as a positive test at least 90 days after the previous one, which also would have left out many reinfections (again, reducing the actual effect), since we know those can happen within a month of a previous infections.

I think a bigger problem is that the VA population skews heavily older and male. Although it was compared to a similar VA cohort, so the effects found should hold up within those populations. I would personally like to know if the vaccine’s lack of protection against the long term sequelae holds just as true in younger people with more robust immune responses to the vaccine. I actually fear from what we’ve seen this far that it will 😔. 

Yes. So many studies being run on the same VA data set. (Although I’ve seen a few comments that the assumption that they are all older is wrong as many veterans are younger as well. I haven’t fact checked that.) it seems like everyone runs studies on existing data rather than collecting new data at the moment.

[Not_a_Number]

14 minutes ago, KSera said:

If you read the study and particularly the analysis and cohort selection methods, the limitations appear to primarily reduce the actual magnitude of effect rather than increase it. The fact that people who didn’t test would end up in the control group  (or the single infection vs reinfection group) would serve to make the control group slightly more like the infected group than it otherwise would have been. They also defined reinfection as a positive test at least 90 days after the previous one, which also would have left out many reinfections (again, reducing the actual effect), since we know those can happen within a month of a previous infections.

It does NOT reduce the magnitude of the effect if the reinfections are not randomly selected. It increases it. 

[Not_a_Number]

16 minutes ago, Ausmumof3 said:

Yes. So many studies being run on the same VA data set. (Although I’ve seen a few comments that the assumption that they are all older is wrong as many veterans are younger as well. I haven’t fact checked that.) it seems like everyone runs studies on existing data rather than collecting new data at the moment.

Yeah, that bothers me. I can see why they want to use it, but you can only squeeze so much information out of ONE data set. 

[KSera]

Just now, Not_a_Number said:

Yeah, that bothers me. I can see why they want to use it, but you can only squeeze so much information out of ONE data set. 

Super important data that they’re able to glean a whole lot of useful info from. I think it’s very worthwhile because it’s one of the most complete data sets we have, they just also need to replicate in other populations. But just knowing the effects are that significant in that population tells us a whole lot.

[Not_a_Number]

Just now, KSera said:

Super important data that they’re able to glean a whole lot of useful info from. I think it’s very worthwhile because it’s one of the most complete data sets we have, they just also need to replicate in other populations. But just knowing the effects are that significant in that population tells us a whole lot.

Except, as I keep saying, it's not at all clear the data isn't self-selected. 

[KSera]

3 minutes ago, Not_a_Number said:

It does NOT reduce the magnitude of the effect if the reinfections are not randomly selected. It increases it. 

It does, because the entire population is the data set. So they are either in the reinfection population or in the control group. It makes the control group more similar to the infection group than it should be.

[KSera]

1 minute ago, Not_a_Number said:

Except, as I keep saying, it's not at all clear the data isn't self-selected. 

Have you read the study itself and read what the data is actually based on and how the groups are set up? That might be helpful in understanding the results.

[Not_a_Number]

Just now, KSera said:

It does, because the entire population is the data set. So they are either in the reinfection population or in the control group. It makes the control group more similar to the infection group than it should be.

This is really standard stats, and I can explain how this fails if you like. 

I may not do research in medical statistics (although, as I said, my husband does), but I do have a stats-adjacent PhD in probability and know what I'm talking about. 

[Not_a_Number]

Just now, KSera said:

Have you read the study itself and read what the data is actually based on and how the groups are set up? That might be helpful in understanding the results.

I don't think I'm the one who's having trouble understanding things. 

I'm finding your responses to me on this topic really rude. I understand that you don't agree with me, but that doesn't excuse talking to me like I don't know anything. Data is in fact very close to my area of expertise and I resent the tone. 

[KSera]

1 minute ago, Not_a_Number said:

I don't think I'm the one who's having trouble understanding things. 

I'm finding your responses to me on this topic really rude. I understand that you don't agree with me, but that doesn't excuse talking to me like I don't know anything. Data is in fact very close to my area of expertise and I resent the tone. 

Apologies on the tone. I think my tone tends to come across as overly terse in writing. It’s not intended as a talking down, I just couldn’t tell from your responses if you have read the study, that’s all. You responded to me quickly enough after I posted the link that it didn’t seem like there would’ve been time for you to have read it. 

[Not_a_Number]

Just now, KSera said:

Apologies on the tone. I think my tone tends to come across as overly terse in writing. It’s not intended as a talking down, I just couldn’t tell from your responses if you have read the study, that’s all. You responded to me quickly enough after I posted the link that it didn’t seem like there would’ve been time for you to have read it. 

Thanks. 

I've seen that dataset before, that's all. As people pointed out upthread, people keep using it. 

[Not_a_Number]

2 minutes ago, KSera said:

Apologies on the tone. I think my tone tends to come across as overly terse in writing. It’s not intended as a talking down, I just couldn’t tell from your responses if you have read the study, that’s all. You responded to me quickly enough after I posted the link that it didn’t seem like there would’ve been time for you to have read it. 

What I'm saying is actually also independent of dataset. It applies to ANY study that's being run off of health records as opposed to questionnaires or interviews. Most COVID infections do not make it into people's records nowadays. 

Edited November 11, 2022 by Not_a_Number

[KSera]

Just now, Not_a_Number said:

Thanks. 

I've seen that dataset before, that's all. As people pointed out upthread, people keep using it. 

Well, it’s almost 6,000,000 people in this case. There’s a good reason they’re using it. Clearly we want more populations studied, but while it’s a narrower set of the population, it doesn’t suffer from some of the other kinds of selection biases that some other study designs might. It has its value for sure.

[Not_a_Number]

Just now, KSera said:

Well, it’s almost 6,000,000 people in this case. There’s a good reason they’re using it. Clearly we want more populations studied, but while it’s a narrower set of the population, it doesn’t suffer from some of the other kinds of selection biases that some other study designs might. It has its value for sure.

How does it not suffer from selection bias? It's a very specific population. 

[KSera]

6 minutes ago, Not_a_Number said:

How does it not suffer from selection bias? It's a very specific population. 

I mean that it’s all the people in the system. It’s not relying on people to specifically sign up for a study or download an app and report symptoms or any of those type of methods that are being used as well. So the population is a specific subset, but of the population, there seems to be less bias than in one that requires people to self select to enter the study.

[Not_a_Number]

7 minutes ago, KSera said:

I mean that it’s all the people in the system. It’s not relying on people to specifically sign up for a study or download an app and report symptoms or any of those type of methods that are being used as well. So the population is a specific subset, but of the population, there seems to be less bias than in one that requires people to self select to enter the study.

Well, it's not self-selected, but it's still very biased as a data set. But I see what you mean. 

I gotta go work now, but I'll check in tomorrow evening, probably. 

[Ausmumof3]

Another question I have around it being veterans data - are they likely to have a different set of vaccinations historically than others? 

[KSera]

15 minutes ago, Not_a_Number said:

I don't think I'm the one who's having trouble understanding things. 

I’m going to try to reword what I’m trying to say here, and then I’ll drop it. I realize you’re the one with a probability PhD, and I’m not, but nonetheless I think we’re talking past each other on this one. Let me put hypothetical numbers on it and see if I can explain my thinking.

Let’s say we stick with just the single infection group and the reinfection group. Single infection is the control group. Let’s say there are 125 heart attacks in the reinfection group but only 100 in the control group. That will be what the increased risk calculation is based on. But let’s say that 15 of those people in the control group who had a heart attack actually had a Covid reinfection but it never got recorded in their chart because they never tested outside their home if at all. Had they tested, they would have been in the reinfection group and then there would have been 140 heart attacks in the reinfection group and only 85 in the control group. How this works out depends of course on population size and statistical significance and all that, but it’s quite possible that that shift means that had those people been tested and classified appropriately, the actual magnitude of risk would have increased.

[Not_a_Number]

13 minutes ago, KSera said:

I’m going to try to reword what I’m trying to say here, and then I’ll drop it. I realize you’re the one with a probability PhD, and I’m not, but nonetheless I think we’re talking past each other on this one. Let me put hypothetical numbers on it and see if I can explain my thinking.

Let’s say we stick with just the single infection group and the reinfection group. Single infection is the control group. Let’s say there are 125 heart attacks in the reinfection group but only 100 in the control group. That will be what the increased risk calculation is based on. But let’s say that 15 of those people in the control group who had a heart attack actually had a Covid reinfection but it never got recorded in their chart because they never tested outside their home if at all. Had they tested, they would have been in the reinfection group and then there would have been 140 heart attacks in the reinfection group and only 85 in the control group. How this works out depends of course on population size and statistical significance and all that, but it’s quite possible that that shift means that had those people been tested and classified appropriately, the actual magnitude of risk would have increased.

Yes, I understand what you mean. But the point is that you're actually shifting the size of the whole GROUP around, not just the numbers in each group. Let me demonstrate with totally fake numbers. 

We have 100 people who got COVID, and another 100 people who didn't get COVID. 20 people in each group had heart attacks. So, in this very hypothetical example, risks are the same. 

However, the people who had heart attacks later were less healthy and were more likely to need to go to the hospital for COVID and to get an official test. So say that all of the 20 people with COVID who later had heart attacks went to the hospital and have it on their record that they had COVID. But out of the remaining 80 people with COVID, only 40 have it on their health record that they had COVID, because most of them were mild and tested at home. 

So what do we have now? We have 140 people whose record says that they didn't have COVID, and 20 of them had heart attacks. 

And we have 60 people whose record says they had COVID, and 20 of them had heart attacks. 

It LOOKS like people who had COVID are more likely to have heart attacks. Much more likely: 33% of the people with COVID on their record had heart attacks, and only 14% of the people without COVID on their record had heart attacks. 

But that's a mirage, because the people more likely to have heart attacks self-selected into the group that was more likely to have COVID on their records. 

Edited November 11, 2022 by Not_a_Number

[KSera]

5 minutes ago, Not_a_Number said:

Yes, I understand what you mean. But the point is that you're actually shifting the size of the whole GROUP around, not just the numbers in each group. Let me demonstrate with totally fake numbers. 

We have 100 people who got COVID, and another 100 people who didn't get COVID. 20 people in each group had heart attacks. So, in this very hypothetical example, risks are the same. 

However, the people who got heart attacks were less healthy and were more likely to need to go to the hospital for COVID and to get an official test. So say that all of the 20 people with COVID with heart attacks went to the hospital and have it on their record that they had COVID. But out of the remaining 80 people with COVID, only 40 have it on their health record that they had COVID, because most of them were mild and tested at home. 

So what do we have now? We have 140 people whose record says that they didn't have COVID, and 20 of them had heart attacks. 

And we have 60 people whose record says they had COVID, and 20 of them had heart attacks. 

It LOOKS like people who had COVID are more likely to have heart attacks. Much more likely: 33% of the people with COVID on their record had heart attacks, and only 14% of the people without COVID on their record had heart attacks. 

But that's a mirage, because the people more likely to have heart attacks self-selected into the group that was more likely to have COVID on their records. 

Well sure. If it’s not actually the case that Covid increases risk of heart attack, that’s the other way it could go. But given all the other evidence we have on Covid and heart attack risk, including in people with mild infections, the scenario where the risk in this study is underestimated by under testing is equally, if not more, likely. 

[Not_a_Number]

1 minute ago, KSera said:

Well sure. If it’s not actually the case that Covid increases risk of heart attack, that’s the other way it could go. But given all the other evidence we have on Covid and heart attack risk, including in people with mild infections, the scenario where the risk in this study is underestimated by under testing is equally, if not more, likely. 

I don't think so. I think the scenario I outlined, where people who later have heart attacks are likely to self-select into the group with COVID with their records is vastly more likely. 

I am really hoping for a more decisive study on this. The Cedars-Sinai study is awful, too. (DH thought their numbers looked like a spreadsheet error, sigh, and he's the one whose JOB involves working with this kind of data. I didn't feel as negative as he did, but I did think the numbers were really weird.) 

Please tag me if you see new studies on this -- I do want to keep the data in mind as we move forward and think about our risk. 

[popmom]

3 hours ago, Not_a_Number said:

and in the meantime, people need to publish or perish... 

I have zero expertise on these things, but as a “lay person” this has been a huge concern of mine throughout the pandemic. Not so much the studies and analyses—but that so much of it is pushed/spread by social media and/or journalists who are also feeling the “publish or perish”. I don’t know what the solution is. It’s one reason I can’t keep up with these threads. It’s why I stopped checking Twitter. At some point it started becoming the scientific community’s version of “fast fashion”. 
 

Just the very “zero clout”, uneducated opinion of someone who wants to be informed but has basically given up.

[Ausmumof3]

27 minutes ago, popmom said:

I have zero expertise on these things, but as a “lay person” this has been a huge concern of mine throughout the pandemic. Not so much the studies and analyses—but that so much of it is pushed/spread by social media and/or journalists who are also feeling the “publish or perish”. I don’t know what the solution is. It’s one reason I can’t keep up with these threads. It’s why I stopped checking Twitter. At some point it started becoming the scientific community’s version of “fast fashion”. 
 

Just the very “zero clout”, uneducated opinion of someone who wants to be informed but has basically given up.

I am actually finding studies that share mechanisms or proposed mechanisms more convincing than stats comparing etc. like the long covid ones where specific things are detected elevated. My knowledge is still not adequate to truly analyse or understand but many of those studies seem less vulnerable to manipulation.

Enough scientist that are well-respected are still preaching caution and I’m happy to go with that. I do like reading the science when I have time but there’s a lot of wading through when you don’t have the background and sometimes it’s just too much!

[Ausmumof3]

This is the first public acknowledgment of post covid heart attacks and sudden death that I’ve seen from any of our health agencies.

[BeachGal]

3 hours ago, Not_a_Number said:

I don't think so. I think the scenario I outlined, where people who later have heart attacks are likely to self-select into the group with COVID with their records is vastly more likely. 

I am really hoping for a more decisive study on this. The Cedars-Sinai study is awful, too. (DH thought their numbers looked like a spreadsheet error, sigh, and he's the one whose JOB involves working with this kind of data. I didn't feel as negative as he did, but I did think the numbers were really weird.) 

Please tag me if you see new studies on this -- I do want to keep the data in mind as we move forward and think about our risk. 

It’s being discussed on Reddit’s Covid19 forum and posters have referenced a couple of other studies that might be better. Most of the posters there have pretty good expertise it seems. Just mentioning this, NAN, because there might be a better study linked in that thread.

https://www.reddit.com/r/COVID19/comments/yrkie3/acute_and_postacute_sequelae_associated_with/?utm_source=share&utm_medium=ios_app&utm_name=iossmf

 

Edited November 11, 2022 by BeachGal
added link

[Ausmumof3]

What do we know about spread of RSV? I’m assuming masking should help for that as well but is it also more surface spread? I know the R0 is typically lower than covid.

[wathe]

4 hours ago, Ausmumof3 said:

What do we know about spread of RSV? I’m assuming masking should help for that as well but is it also more surface spread? I know the R0 is typically lower than covid.

In healthcare we treat it as "droplet contact", meaning protecting against both respiratory droplets and surface spread.  Conventional wisdom cites that it can live on hard surfaces for hours.  I don't have studies to back that up, but should be a quick google.

Our understanding is shifting with respect to all respiratory viruses, though, and it's very very likely that RSV is also airborne - at least short-range airborne.  RSV-containing aerosols have been found in the air surrounding RSV+ inpatients thats capable of infecting human ciliated epitheleum (Kulkami et al., 2016).  A more recent article outlining our shifting understanding of respiratory transmission , with big name aerosol scientists as co-authors (K. Prather, J. Jiminez, L. Marr).  

There is still much resistance to this shift in the IPAC (Infections Prevention and Control) medical establishment.  Likely because of the practical and regulatory implications of a formal shift to airborne IPAC practices for common resp viruses; current hospital infrastructure simply physically cannot accommodate airborne precautions at scale.

Edited November 11, 2022 by wathe

[BeachGal]

7 hours ago, Ausmumof3 said:

What do we know about spread of RSV? I’m assuming masking should help for that as well but is it also more surface spread? I know the R0 is typically lower than covid.

Pfizer’s RSV vaccine will probably hit the market next year—hopefully. Unfortunately, that doesn't help this year so do what you can to keep the little ones healthy.

https://arstechnica.com/science/2022/11/why-pfizers-rsv-vaccine-success-is-a-big-deal-decades-in-the-making/

[BronzeTurtle]

reading up and down this thread is very helpful when it comes to teh reinfection study. This in particular is what Number was trying to say. There's a lot of linked analysis. And a statement from the study authors about the headlines that resulted.

https://twitter.com/mlipsitch/status/1591076296834289664?s=20&t=yGeRPg83K2p7vFI0yuMkbw

[mommyoffive]

Covid Reinfections and Long Covid - by Eric Topol (substack.com)

 

New purifier may cut airborne Covid by 10x - ScienceBlog.com

 

Stay Up to Date with COVID-19 Vaccines Including Boosters | CDC

 

Results Of New Study Make The Case For Mask Mandates In Schools To Prevent COVID-19 (msn.com)

 

Biden to extend COVID-19 health emergency into April 2023: Report (msn.com)

Edited November 12, 2022 by mommyoffive

[Not_a_Number]

Wow. I see that the math I posted wasn't helpful to people. That's... really disappointing, since you really have to understand that kind of toy example to be able to talk about statistics in an educated way. (And yes, that example works exactly the same way if the risks are elevated. The point isn't that the risks aren't elevated. The point is that the numbers we're getting from this study are probably not telling us the magnitude of the effect, whatever it is.) 

Self-selected samples are completely worthless and will give you meaningless results. If you don't understand how that works, you will be swayed by badly designed studies. Many, many medical studies are badly designed -- that's why there's a replication crisis in the medical sciences.

I'm NOT trying to argue that I know what happens with COVID reinfections. I'm as worried about that as the rest of you. But using badly designed studies to justify what you already think is just plain old confirmation bias. 

Edited November 12, 2022 by Not_a_Number

[Not_a_Number]

9 hours ago, BronzeTurtle said:

reading up and down this thread is very helpful when it comes to teh reinfection study. This in particular is what Number was trying to say. There's a lot of linked analysis. And a statement from the study authors about the headlines that resulted.

https://twitter.com/mlipsitch/status/1591076296834289664?s=20&t=yGeRPg83K2p7vFI0yuMkbw

Thank you. Exactly. 

[Ausmumof3]

3 minutes ago, Not_a_Number said:

Wow. I see that the math I posted wasn't helpful to people. That's... really disappointing, since you really have to understand that kind of toy example to be able to talk about statistics in an educated way. (And yes, that example works exactly the same way if the risks are elevated. The point isn't that the risks aren't elevated. The point is that the numbers we're getting from this study are probably not telling us the magnitude of the effect, whatever it is.) 

Self-selected samples are completely worthless and will give you meaningless results. If you don't understand how that works, you will be swayed by badly designed studies. Many, many medical studies are badly designed -- that's why there's a replication crisis in the medical sciences.

I'm NOT trying to argue that I know what happens with COVID reinfections. I'm as worried about that as the rest of you. But using badly designed studies to justify what you already think is just plain old confirmation bias. 

The study itself did acknowledge its limitations and also attempt to control for factors like age, health status etc. the reporting on it may have ignored this but it’s not the study’s fault.

 

[Not_a_Number]

1 minute ago, Ausmumof3 said:

The study itself did acknowledge its limitations and also attempt to control for factors like age, health status etc. the reporting on it may have ignored this but it’s not the study’s fault.

What I'm saying is completely unrelated to that, though. What I'm saying is the health records are incomplete. @BronzeTurtle's link is exactly right. 

[Ausmumof3]

25 minutes ago, Not_a_Number said:

What I'm saying is completely unrelated to that, though. What I'm saying is the health records are incomplete. @BronzeTurtle's link is exactly right. 

“However, the people who had heart attacks later were less healthy and were more likely to need to go to the hospital for COVID and to get an official test. So say that all of the 20 people with COVID who later had heart attacks went to the hospital and have it on their record that they had COVID. But out of the remaining 80 people with COVID, only 40 have it on their health record that they had COVID, because most of them were mild and tested at home.”

this is what I was responding to. 
 

i do see your point I think but I think the researchers tried to control for that. I’m not sure if they did that well, as I found the study methods a bit hard to understand. 
 

Is it standard to test everyone hospitalised for covid there? 
 

I am a bit wary of zeynap as she wrote a fairly strong opinion piece a while ago about how covid would mutate to be a mild cold. I don’t think she’s a scientist she’s a NYT columnist? Lipsitch is well respected.

[Ausmumof3]

31 minutes ago, Not_a_Number said:

What I'm saying is completely unrelated to that, though. What I'm saying is the health records are incomplete. @BronzeTurtle's link is exactly right. 

I guess aside from Stats why do we expect reinfections to be less or more severe? How does that compare to other infections that we get on repeat? For example, flu. Does flu become less problematic the more times we catch it? I believe some infections do work that way (RSV?) but others it’s more to do with the specific strain or health status at the time. 

[Not_a_Number]

6 minutes ago, Ausmumof3 said:

“However, the people who had heart attacks later were less healthy and were more likely to need to go to the hospital for COVID and to get an official test. So say that all of the 20 people with COVID who later had heart attacks went to the hospital and have it on their record that they had COVID. But out of the remaining 80 people with COVID, only 40 have it on their health record that they had COVID, because most of them were mild and tested at home.”

this is what I was responding to. 
 

i do see your point I think but I think the researchers tried to control for that. I’m not sure if they did that well, as I found the study methods a bit hard to understand. 

You can't control for that. The fact that mild cases are unlikely to make it onto health records isn't something you can do ANYTHING about except actually doing a study where you talk to people, instead of simply using health records. And that's much harder and more expensive, which is why we keep seeing studies on this totally ridiculous dataset. 

When you have corrupted data, you can't analyze it away. 

By the way, I was talking to DH about this, and he said he's also seen the VA datasets in cancer research (he does this for work), and they are completely wonky there, too -- there, the problem isn't the record (cancer makes it onto people's records!) but the fact that this population is very different from the average person for a wide variety of people (such as age as well as exposure to a variety of damaging chemicals as part of modern warfare if they are younger.)  

Edited November 12, 2022 by Not_a_Number

[Not_a_Number]

Just now, Ausmumof3 said:

I guess aside from Stats why do we expect reinfections to be less or more severe? How does that compare to other infections that we get on repeat? For example, flu. Does flu become less problematic the more times we catch it? I believe some infections do work that way (RSV?) but others it’s more to do with the specific strain or health status at the time. 

On average, having your immune system trained is good. Yes, it depends how long it had been and what has happened in the meantime. 

I forget how well this works for flu. The problem is that flu has equally bad records -- mild infections don't make it onto health records. I did read a good study about this with respect to some common cold coronaviruses. 

[Ausmumof3]

11 minutes ago, Not_a_Number said:

On average, having your immune system trained is good. Yes, it depends how long it had been and what has happened in the meantime. 

I forget how well this works for flu. The problem is that flu has equally bad records -- mild infections don't make it onto health records. I did read a good study about this with respect to some common cold coronaviruses. 

Given that older people die of flu at higher rates than younger people?

There was some helpful information further up thread on whether or not exposure to viruses is actually a good thing for the immune system I think? Did you see it? 

Edited November 12, 2022 by Ausmumof3

[Ausmumof3]

[Not_a_Number]

2 minutes ago, Ausmumof3 said:

Given that older people die of flu at higher rates than younger people?

OK? But they'd die much more if they were never exposed and were exposed for the first time when they are old. 

 

2 minutes ago, Ausmumof3 said:

There was some helpful information further up thread on whether or not exposure to viruses is actually a good thing for the immune system I think? Did you see it? 

I don't know if it's a GOOD thing for the immune system, but yes, the immune system learns to handle viruses from exposure. 

There's good evidence that prior COVID infection is protective: 

https://news.weill.cornell.edu/news/2022/06/qatar-omicron-wave-study-shows-slow-decline-of-natural-immunity-rapid-decline-of

The best thing you can do in your life is to never catch a virus. Granted. But if that's not happening, then yes, you're best off catching it when your immune system is functioning as well as possible so it can learn to deal. 

[Ausmumof3]

3 minutes ago, Not_a_Number said:

OK? But they'd die much more if they were never exposed and were exposed for the first time when they are old. 

 

I don't know if it's a GOOD thing for the immune system, but yes, the immune system learns to handle viruses from exposure. 

There's good evidence that prior COVID infection is protective: 

https://news.weill.cornell.edu/news/2022/06/qatar-omicron-wave-study-shows-slow-decline-of-natural-immunity-rapid-decline-of

The best thing you can do in your life is to never catch a virus. Granted. But if that's not happening, then yes, you're best off catching it when your immune system is functioning as well as possible so it can learn to deal. 

If it’s a once and done virus, yes. For viruses that cause repeat infections I think that’s much less clear. And even with the once and done, like chicken pox, there’s shingles etc. I’m not convinced that more infection is better than less infection as a general rule. 

[Not_a_Number]

Just now, Ausmumof3 said:

If it’s a once and done virus, yes. For viruses that cause repeat infections I think that’s much less clear. And even with the once and done, like chicken pox, there’s shingles etc. I’m not convinced that more infection is better than less infection as a general rule. 

Well, herpes viruses are just different -- they aren't really done, they just hide. 

We have repeat infections of lots of respiratory viruses. And yes, exposure will make later infections easier for most of them. I can pull up studies for that (although this is, again, very hard to design well), but it's also just... common sense? 

[Ausmumof3]

26 minutes ago, Not_a_Number said:

Well, herpes viruses are just different -- they aren't really done, they just hide. 

We have repeat infections of lots of respiratory viruses. And yes, exposure will make later infections easier for most of them. I can pull up studies for that (although this is, again, very hard to design well), but it's also just... common sense? 

Well that’s what I thought. I mean that was my impression prior to covid before but it turns out a lot i though I knew was common sense was just plain wrong 😑 

Before covid, the general consensus was airborne is not a thing either.

[wathe]

7 minutes ago, Not_a_Number said:

OK? But they'd die much more if they were never exposed and were exposed for the first time when they are old. 

 

I don't know if it's a GOOD thing for the immune system, but yes, the immune system learns to handle viruses from exposure. 

There's good evidence that prior COVID infection is protective: 

https://news.weill.cornell.edu/news/2022/06/qatar-omicron-wave-study-shows-slow-decline-of-natural-immunity-rapid-decline-of

The best thing you can do in your life is to never catch a virus. Granted. But if that's not happening, then yes, you're best off catching it when your immune system is functioning as well as possible so it can learn to deal. 

Can be protective if you survive the initial infection intact in the first place.  If you don't?

Some viruses, like RSV and parainfluenza virus are  objectively clinically much worse in very young patients.  Having these for the first time when a little older is actually probably safer - once airway and lower respiratory tract anatomy have matured, the clinical syndrome tends to be less severe.

I would argue that you're best off training your immune system via vaccination rather than by getting sick.

We have no idea if our bodies completely clear covid either.   There is some evidence that maybe we don't (quick google finds this paper and this other paper.  I know there are more, but need to go to bed, not scour the internet....)  We don't know whether there will be post-polio syndrome-like sequelae, or covid induced cancers (p53 suppression is a plausible mechanism), or multiple sclerosis-like sequelae or shingles-like sequelae, or something completely new.  We just don't know.

I'm not at all sold that seemingly endless viral respiratory infections are good for kids.  I'm sticking with a masking and avoid strategy for mine.