There is so much we don't know about the virus - and here's another reason to be cautious

[Anonymous]

you all fail basic math again

something that has a prevalence of less than 0001% please driving is more dangerous

I swear yall are ridiculous

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:New syndrome in kids may be linked to covid:

https://www.nytimes.com/2020/05/05/nyregion/coronavirus-new-york-update.html?type=styln-live-updates&label=new%20york%20&index=2&action=click&module=Spotlight&pgtype=Homepage#link-1aca38d0

"Fifteen children, many of whom had the coronavirus, have recently been hospitalized in New York City with a mysterious syndrome that doctors do not yet fully understand but that has also been reported in several European countries, health officials announced on Monday night.

Many of the children, ages 2 to 15, have shown symptoms associated with toxic shock or Kawasaki disease, a rare illness in children that involves inflammation of the blood vessels, including coronary arteries, the city’s health department said..."

Yes, I am a pediatrician, and there is a lot of discussion about this behind the scenes.

Can you tell us more?

I would like to know if they are suggesting that COVID-19 causes Kawasaki disease? Or are they saying that Kawasaki disease is COVID-19? I’m sorry if I seem dense I am trying to understand.

Kawasaki syndrome (and its variants) are a mucocutaneous lymph node syndrome. It's thought to be an immune system response triggered by something (maybe viral or bacterial illness of various types) where the body's own immune system attacks its own small and medium-sized arteries, causing problems with the skin, mucous membranes, and vessel walls.

This may or may not be a significant problem with COVID-19. It's too early to tell, but there are some alarm bells going off that need to be followed up on. I think people in the field are more worried about this than they were with SARS or MERS because it seems like COVID-19 has more effects in adults on the cardiovascular system. That's not by any means conclusive, but it is concerning.

[Anonymous]

Anonymous wrote:you all fail basic math again

something that has a prevalence of less than 0001% please driving is more dangerous

I swear yall are ridiculous

If there is bad potato salad served at a picnic, and if the bacteria responsible is known to take 2-8 hours to cause symptoms, do you conclusively say at hour 3 that only one kid has vomited, so there is nothing at all to worry about? Or do you say that anyone who ate the potato salad MIGHT be showing up with symptoms in the next 5 hours, and geez, maybe it's more than just the first one who's going to be sick?

[Anonymous]

Anonymous wrote:Also, I’m not talking about an antibody response. I’m talking about the antibody testing that was done in NYC to estimate the true number of COVID cases.

I’m doing that to estimate the commonality of Kawasaki among NYC COVID-positive children.

You realize that the testing measures the level of antibody response to the virus, right?

Or what was it you were thinking was being tested?

[Anonymous]

Anonymous wrote:

Anonymous wrote:Also, I’m not talking about an antibody response. I’m talking about the antibody testing that was done in NYC to estimate the true number of COVID cases.

I’m doing that to estimate the commonality of Kawasaki among NYC COVID-positive children.

You realize that the testing measures the level of antibody response to the virus, right?

Or what was it you were thinking was being tested?

You’re still not getting it.

I brought up the antibody testing TO ESTIMATE THE ACTUAL NUMBER OF CASES. That’s it. I’m not interested in the antibody response specifically. I was estimating how common Kawasaki is, percentage-wise, in NYC. You need the actual number of cases to estimate that. Actual number of cases can be derived—roughly—from the results of the antibody testing they did.

[Anonymous]

Anonymous wrote:

Anonymous wrote:you all fail basic math again

something that has a prevalence of less than 0001% please driving is more dangerous

I swear yall are ridiculous

If there is bad potato salad served at a picnic, and if the bacteria responsible is known to take 2-8 hours to cause symptoms, do you conclusively say at hour 3 that only one kid has vomited, so there is nothing at all to worry about? Or do you say that anyone who ate the potato salad MIGHT be showing up with symptoms in the next 5 hours, and geez, maybe it's more than just the first one who's going to be sick?

How do you wake up in the morning and even cook breakfast without thinking the stove will explode or that you’ll set your house on fire? You sound insane.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Also, I’m not talking about an antibody response. I’m talking about the antibody testing that was done in NYC to estimate the true number of COVID cases.

I’m doing that to estimate the commonality of Kawasaki among NYC COVID-positive children.

You realize that the testing measures the level of antibody response to the virus, right?

Or what was it you were thinking was being tested?

You’re still not getting it.

I brought up the antibody testing TO ESTIMATE THE ACTUAL NUMBER OF CASES. That’s it. I’m not interested in the antibody response specifically. I was estimating how common Kawasaki is, percentage-wise, in NYC.

[Are you confusing Kawasaki with COVID-19 here? They aren't the same thing.]

You need the actual number of cases to estimate that. Actual number of cases can be derived—roughly—from the results of the antibody testing they did.

Antibody testing is dependent on the antibody response. That is LITERALLY what it measures -- if you cite the test, you are measuring the response. And kids have a different antibody response than adults do -- e.g., this is why the DTap has higher levels of tetanus toxoid than the TDaP, because kids don't respond the same way to the bacteria.

Tests that pick up prior cases in adults might not pick them up in kids, or they might overrepresent them, or they might underrepresent them. You can't extrapolate from the adult response (measured by the antibody test) to make conclusions about what it says about kids. Kids aren't just short adults.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:you all fail basic math again

something that has a prevalence of less than 0001% please driving is more dangerous

I swear yall are ridiculous

If there is bad potato salad served at a picnic, and if the bacteria responsible is known to take 2-8 hours to cause symptoms, do you conclusively say at hour 3 that only one kid has vomited, so there is nothing at all to worry about? Or do you say that anyone who ate the potato salad MIGHT be showing up with symptoms in the next 5 hours, and geez, maybe it's more than just the first one who's going to be sick?

How do you wake up in the morning and even cook breakfast without thinking the stove will explode or that you’ll set your house on fire? You sound insane.

No, just educated in medicine and interpreting scientific data.

That's not fear. It's not emotions, just education.

[Anonymous]

BTW, I find it strange that being evidence-based and rational gets equated to being overwhelmed by emotions. It's actually the less emotional way to respond to a situation.

I'm not afraid or anxious, despite being in a high risk category myself. I work with swabbing people who are symptomatic in an ER parking lot.

I also care a lot about being clear and accurate in interpreting medical data, and in drawing conclusions from it. I care about my patients.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Also, I’m not talking about an antibody response. I’m talking about the antibody testing that was done in NYC to estimate the true number of COVID cases.

I’m doing that to estimate the commonality of Kawasaki among NYC COVID-positive children.

You realize that the testing measures the level of antibody response to the virus, right?

Or what was it you were thinking was being tested?

You’re still not getting it.

I brought up the antibody testing TO ESTIMATE THE ACTUAL NUMBER OF CASES. That’s it. I’m not interested in the antibody response specifically. I was estimating how common Kawasaki is, percentage-wise, in NYC.

[Are you confusing Kawasaki with COVID-19 here? They aren't the same thing.]

You need the actual number of cases to estimate that. Actual number of cases can be derived—roughly—from the results of the antibody testing they did.

Antibody testing is dependent on the antibody response. That is LITERALLY what it measures -- if you cite the test, you are measuring the response. And kids have a different antibody response than adults do -- e.g., this is why the DTap has higher levels of tetanus toxoid than the TDaP, because kids don't respond the same way to the bacteria.

Tests that pick up prior cases in adults might not pick them up in kids, or they might overrepresent them, or they might underrepresent them. You can't extrapolate from the adult response (measured by the antibody test) to make conclusions about what it says about kids. Kids aren't just short adults.

Ok. I’ll try to explain more clearly.

This was my goal: to estimate the percentage of actual pediatric COVID cases that also show Kawasaki.

How do I get there? I use the antibody testing in NYC because that’s the best way to estimate actual COVID cases. That testing was done with a REPRESENTATIVE SAMPLE OF THE POPULATION.

What is a representative sample? It’s a sub-section of the population that mimics the characteristics of the whole population. That includes age.

What does that mean? It means that because the antibody testing suggested the actual number of cases in NYC was about 5 times higher than the confirmed, I can use that to extrapolate the actual case number in any demographic group I want to look at.

So what did I do with that information? I took the confirmed number of pediatric COVID cases—4,088–and multiplied it by 5 to get about 20,000. Then I computed 15/20,000 to estimate the percentage of actual pediatric COVID cases with Kawasaki.

I hope that’s clearer.

[Anonymous]

Follow up: Obviously I can’t do what I did if the antibody testing wasn’t done with a representative sample. Please let me know if that’s the case. Of course, it would’ve been stupid of them to not get a representative sample.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Also, I’m not talking about an antibody response. I’m talking about the antibody testing that was done in NYC to estimate the true number of COVID cases.

I’m doing that to estimate the commonality of Kawasaki among NYC COVID-positive children.

You realize that the testing measures the level of antibody response to the virus, right?

Or what was it you were thinking was being tested?

You’re still not getting it.

I brought up the antibody testing TO ESTIMATE THE ACTUAL NUMBER OF CASES. That’s it. I’m not interested in the antibody response specifically. I was estimating how common Kawasaki is, percentage-wise, in NYC.

[Are you confusing Kawasaki with COVID-19 here? They aren't the same thing.]

You need the actual number of cases to estimate that. Actual number of cases can be derived—roughly—from the results of the antibody testing they did.

Antibody testing is dependent on the antibody response. That is LITERALLY what it measures -- if you cite the test, you are measuring the response. And kids have a different antibody response than adults do -- e.g., this is why the DTap has higher levels of tetanus toxoid than the TDaP, because kids don't respond the same way to the bacteria.

Tests that pick up prior cases in adults might not pick them up in kids, or they might overrepresent them, or they might underrepresent them. You can't extrapolate from the adult response (measured by the antibody test) to make conclusions about what it says about kids. Kids aren't just short adults.

Ok. I’ll try to explain more clearly.

This was my goal: to estimate the percentage of actual pediatric COVID cases that also show Kawasaki.

How do I get there? I use the antibody testing in NYC because that’s the best way to estimate actual COVID cases. That testing was done with a REPRESENTATIVE SAMPLE OF THE POPULATION.

What is a representative sample? It’s a sub-section of the population that mimics the characteristics of the whole population. That includes age.

What does that mean? It means that because the antibody testing suggested the actual number of cases in NYC was about 5 times higher than the confirmed, I can use that to extrapolate the actual case number in any demographic group I want to look at.

So what did I do with that information? I took the confirmed number of pediatric COVID cases—4,088–and multiplied it by 5 to get about 20,000. Then I computed 15/20,000 to estimate the percentage of actual pediatric COVID cases with Kawasaki.

I hope that’s clearer.

It's not. Can you link your citation?

What I see reference to online is not a "representative sample" stratified by age, but a random sampling from various supermarkets. If you are relying on that to make conclusions about kids, you are way off.

If you're relying on a different study, I'd like to see the data, because I don't have a lot of confidence in how it is being assessed and presented here.

[Anonymous]

Anonymous wrote:Follow up: Obviously I can’t do what I did if the antibody testing wasn’t done with a representative sample. Please let me know if that’s the case. Of course, it would’ve been stupid of them to not get a representative sample.

You ... didn't check before drawing your conclusion?

[Anonymous]

PS: In general, it's a good idea to look at how a study was done before making claims about what it shows.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:you all fail basic math again

something that has a prevalence of less than 0001% please driving is more dangerous

I swear yall are ridiculous

If there is bad potato salad served at a picnic, and if the bacteria responsible is known to take 2-8 hours to cause symptoms, do you conclusively say at hour 3 that only one kid has vomited, so there is nothing at all to worry about? Or do you say that anyone who ate the potato salad MIGHT be showing up with symptoms in the next 5 hours, and geez, maybe it's more than just the first one who's going to be sick?

How do you wake up in the morning and even cook breakfast without thinking the stove will explode or that you’ll set your house on fire? You sound insane.

No, just educated in medicine and interpreting scientific data.

That's not fear. It's not emotions, just education.

Educated in medicine? What does that mean, exactly?