I talked at length with my massage therapist about why she continues not getting vaccinated. Her elderly parents are not vaccinated; neither are her friends. I asked if she would get the jabs if health care workers were required to do so, and she said yes of course. She doesn’t seem to have particular concerns about side effects, so it seems to be just a matter of convenience and lack of peer pressure. Weird to me, but it seems either a mandate or a change in behavior of the people around her. I support mandates because I don’t know how else we as a society move the needle toward the coronavirus being a small but manageable risk like other contagious diseases.
I have worked for nearly the past 15 years for healthcare organizations *though I am not a direct care worker). All but about 4 of those years I have been required to get a flu vaccine, with which I have no problem. As said many times above, including by Isaac: the employers get to set conditions for employment. (Isaac -- I don't really get how "vaccine mandates should only be issued if all other options have been exhausted" fits into that.)
But I didn't get a flu vaccine the 4 years my employer didn't require it, in 'protest' for the way the vaccines are marketed. A pharmacy nearby put out a sign several years in a row that started "Top 3 reasons to get a flu vaccine: #1 Reduce your chance of flu-related death; #2 Reduce your chance of flu-related hospitalization; ..." That's straight-up fear mongering, and I refuse to respond to it. It's a *distinct* minority of the population who have *any* measurable risk of flu-related death.
As the data on covid has come in, it has happily become clear that for the vast majority of rich-world denizens with basic access to health care, covid, too, would be just another disease that runs its course. As such, I absolutely support arguments like "Those people who currently refuse the vaccine (without health reasons), and those who encourage others to refuse the vaccine, are causing needless death (whatever their intention)," and I'm glad to see that that's the main tenor of arguments here.
Considering the concern expressed about the recent Delta surge, I think this is pretty damn irresponsible.
Intellectual consistency would demand we at least vaccinate these people as they cross and hold them for the requisite time to allow for the vaccine to provide immunity. Instead, we continue to do stupid things and wonder why the surge continues. . .
thanks for information that the protestors (as opposed to the Capitol insurrectionists) are also being questioned and prosecuted. I've evidently missed this salient point on PBS, MSNBC and CNN. This would answer some Republican congresspersons' complaints that the attention given the insurrection is biased because "all" protests are supposedly crimes.
Are we seeing a "spike" in cases? No, not really. Look closely. Deaths are down. Serious cases are down. We are seeing a spike in positive test results.
The test for covid gives at least 10% false positives. Some say 15% to 20%. As people get vaccinated and the prevalence of a disease goes down, the rate of false positives goes up. Here's why: Suppose the prevalence is 2%, as it was for covid last year. If you test positive, and the test gives 10% false positives, what are the odds that you actually have covid?
If you said, 90%, that is flatly wrong. At 2% prevalence, you must do 50 tests, on average, to get one true positive. Those 50 tests will give five false positives. Your true odds of having it are one in 6, or 18%.
But what happens when the prevalence falls to 1%, as it has for covid? Now you must do 100 tests, on average, to get one true positive. Those hundred tests give ten false positives, so now your odds of actually having it are one in 11, or only 9%.
What looks like a spike in positive test results is actually a rise in false positives caused by the lower rate of prevalence.
Shawn -- but when false positives greatly outnumber true positives, shouldn't the rate of positive test results (false + true) hold relatively steady as prevalence rises and falls? If (as in your example) false positive rate is 10%, then the number of positive test results for 100 tests at 2% prevalence should be 12. At 1% prevalence the number of positive test results for 100 tests should be 11.
So even if the odds are low that an individual with a positive test result actually has covid, a spike in *absolute numbers of positive test results* has to reflect either (1) more testing, or (2) more true positives. A spike in *rate of positive test results* has to reflect either (3) a higher rate of false positives, or (4) a higher rate of true positives.
A rise in false positives *relative to* true positives shouldn't push the total rate of positive test results up, should it?
Do you have a specific set of sources or real data we could work off here?
I tried to make it simple, Lark, but in fact it is somewhat complicated. Testing fails when the rate of false positives is high relative to the prevalence of the disease. You see this, for example, in testing for lung cancer (which many physicians decline to do for that very reason) and in testing for HIV, and even for breast cancer in some age groups. Mammograms give 9% false positives, as some of my friends who had a scary one are well aware, and although the prevalence of breast cancer varies widely among subsets of the population, in general, your odds of actually having breast cancer after a positive mammogram are only about 15%.
It is what statisticians call Type 1 Error (type 2 being false negatives). A little Google research on the terms I have used here will tell you more than you want to know. The bottom line is that most--far more than half--of all reported covid cases are false positive tests on people who are sick for other reasons. They then get misdiagnosed with covid because it pays more than what they actually have.
The impact of false positives is so high because the prevalence of covid is relatively low. And thankfully declining, although that makes testing matters even worse. Your intuitive guess--that a positive test result implies 90% chance of infection--assumes 100% prevalence. The actual prevalence of covid, today, is maybe 1%. So the actual chance that someone has it, after a positive test, is therefore less than 10%.
Shawn -- it sounds like you're explaining -- accurately -- why testing *for an individual* is currently too likely to produce a false positive.
But your initial comment started *and ended* "Are we seeing a "spike" in cases? No, not really. ... We are seeing a spike in positive test results" and "What looks like a spike in [cases] positive test results is actually a rise in false positives caused by the lower rate of prevalence." That is a comment alleging that testing *on the population level* is currently not useful, which is a different point and needs a different set of data to support it.
My core question was: "A rise in false positives *relative to* true positives shouldn't push the total rate of positive test results up, should it?" What's your answer to that? I still don't see why a spike in *rate of* positive test results could NOT reflect also an actual spike in *number of* cases, even if the *rate of* cases (i.e. prevalence) is goes down.
Look at it this way: If there were no covid, at least 10% of the population would still test positive. By testing everyone in the US, you would identify at least 33 million "confirmed" cases, all false positives. (We have actually seen 36 million cases.)
Or else, look at it this way: If for some reason you wanted to show a spike in cases, all you would need to do is more testing. When you do 10% more tests (as we recently have), you get 10% more cases (which we are seeing), even if the true incidence of covid declines (as it probably has).
So, when you read media scaremongering about rising case loads, look first for the type 1 error that they never even mention, much less account for. Then check to see whether they are conflating incidence (people who get the disease) with prevalence (people who have the disease), as they almost always do. By definition, the prevalence of having covid (ie, testing positive) cannot fall below the type 1 error rate of at least 10%, even if the incidence (ie, true new cases) falls to zero.
Shawn -- I'm really not seeing the numbers you're seeing (and I'm responding so late because I didn't make the time earlier to check - sorry about that). What test are you talking about? The new home tests have known very significant false positive rates; the PCR tests, however, MUCH less so. Where are you getting your data about the positivity rates of different tests?
You write - "By definition, the prevalence of having covid (ie, testing positive) cannot fall below the type 1 error rate of at least 10%, even if the incidence (ie, true new cases) falls to zero." Yet CDC's https://covid.cdc.gov/covid-data-tracker/#trends_dailycases_7daytestingpositive clearly shows that for months earlier this year the positivity rate was below 5%. https://covid.cdc.gov/covid-data-tracker/#cases_positivity7day shows that the TOTAL tests recorded by CDC *ever* is 507million to get 38million positive (data posted evening of Aug 19)-- significantly less than 10%. There are currently many states in the U.S. showing positivity rate of less than 8%.
Please share the numbers you're seeing on which you're basing your conclusions, and the sources of those numbers.
I don't want to quibble over numbers, Lark. What matters is cases.
But, what is a case? Someone who truly dies of covid? Someone who dies from something else but with covid? Someone who gets sick and tests positive for covid? Who isn't even sick but tests positive? Who died in a car accident and later tested positive?
We are subjected to a constant media barrage of what I call pandemic porn. Scare tactics, because that is what people want to click on. For the same reason we watch A Quiet Place: because we want to be scared.
Now, Lark, I will tell you a story; one that, this late in comments, only you (and just maybe Isaac) will ever see here. I was a covid denier from the very start. Always thought it was way overblown. Never went masked if I didn't have to. Took zero precautions, and I fly twice a month between my homes in Boston and Cape Coral, FL.
For five or six years, until 2014, I was with a woman who was the true love of my life. Alas, we were both married, although not to one another. Then she left me, without a word--neither then nor ever since. Two weeks ago, I learned that she died last May (on my birthday) at age 69. She was so healthy, although she did smoke. It could only have been covid.
Since then, in the last two weeks, my feelings about covid have totally changed. Now, even I am scared. I miss her so much.
I don't know how you mandate vaccinies that are only authorized for emergency use. I think everyone over 60 and those with dangerous co-morbidities should all get the vaccine as the death rate is greater than a normal flu risk.
I am old enough to remember Fen Phen, Thalidomide, VIOXX & Bextra, Quaaludes, Darvon, DES, PTZ, Baycol. We have no idea on what the near or long term effects may be. I cannot imagine giving this to under 20's to whom the danger is low (similar to regular flu). I think adults between 20 and 60 without dangerous co-morbidities should decide for themselves.
Will companies that mandate a vaccine be liable for any long term problems?
In reading the reader question on the disparity between convictions of Capitol rioters and BLM rioters I was reminded of the trend described in a recent Michael Tracey post. Basically the claim is that while the Capitol rioters can't be charged with any of the crimes associated with domestic terrorism, they will be prosecuted in the context of domestic terrorism, with correspondingly stiffer sentencing? If MT is accurately representing the data it is concerning... https://mtracey.substack.com/p/in-radical-affront-to-civil-liberties?token=eyJ1c2VyX2lkIjoxMjQ1Njg4NiwicG9zdF9pZCI6MzkxMjM3MDAsIl8iOiI4L2JsVyIsImlhdCI6MTYyNzQyNjE4MSwiZXhwIjoxNjI3NDI5NzgxLCJpc3MiOiJwdWItMzAzMTg4Iiwic3ViIjoicG9zdC1yZWFjdGlvbiJ9.o50RATC7oHnRs0CGmS8cG9LP7Dt81MOBT-2Q_OTj2XI
Everyone must read this.
I talked at length with my massage therapist about why she continues not getting vaccinated. Her elderly parents are not vaccinated; neither are her friends. I asked if she would get the jabs if health care workers were required to do so, and she said yes of course. She doesn’t seem to have particular concerns about side effects, so it seems to be just a matter of convenience and lack of peer pressure. Weird to me, but it seems either a mandate or a change in behavior of the people around her. I support mandates because I don’t know how else we as a society move the needle toward the coronavirus being a small but manageable risk like other contagious diseases.
I have worked for nearly the past 15 years for healthcare organizations *though I am not a direct care worker). All but about 4 of those years I have been required to get a flu vaccine, with which I have no problem. As said many times above, including by Isaac: the employers get to set conditions for employment. (Isaac -- I don't really get how "vaccine mandates should only be issued if all other options have been exhausted" fits into that.)
But I didn't get a flu vaccine the 4 years my employer didn't require it, in 'protest' for the way the vaccines are marketed. A pharmacy nearby put out a sign several years in a row that started "Top 3 reasons to get a flu vaccine: #1 Reduce your chance of flu-related death; #2 Reduce your chance of flu-related hospitalization; ..." That's straight-up fear mongering, and I refuse to respond to it. It's a *distinct* minority of the population who have *any* measurable risk of flu-related death.
As the data on covid has come in, it has happily become clear that for the vast majority of rich-world denizens with basic access to health care, covid, too, would be just another disease that runs its course. As such, I absolutely support arguments like "Those people who currently refuse the vaccine (without health reasons), and those who encourage others to refuse the vaccine, are causing needless death (whatever their intention)," and I'm glad to see that that's the main tenor of arguments here.
"𝐀𝐛𝐨𝐮𝐭 𝟓𝟎,𝟎𝟎𝟎 𝐦𝐢𝐠𝐫𝐚𝐧𝐭𝐬 𝐰𝐡𝐨 𝐜𝐫𝐨𝐬𝐬𝐞𝐝 𝐭𝐡𝐞 𝐬𝐨𝐮𝐭𝐡𝐞𝐫𝐧 𝐛𝐨𝐫𝐝𝐞𝐫 𝐢𝐥𝐥𝐞𝐠𝐚𝐥𝐥𝐲 𝐡𝐚𝐯𝐞 𝐛𝐞𝐞𝐧 𝐫𝐞𝐥𝐞𝐚𝐬𝐞𝐝 𝐢𝐧𝐭𝐨 𝐭𝐡𝐞 𝐔.𝐒. 𝐰𝐢𝐭𝐡𝐨𝐮𝐭 𝐚 𝐜𝐨𝐮𝐫𝐭 𝐝𝐚𝐭𝐞."
Considering the concern expressed about the recent Delta surge, I think this is pretty damn irresponsible.
Intellectual consistency would demand we at least vaccinate these people as they cross and hold them for the requisite time to allow for the vaccine to provide immunity. Instead, we continue to do stupid things and wonder why the surge continues. . .
Richard -- today Isaac put on 'quick hits' a Texas action that seems to relate also to your concern https://www.politico.com/news/2021/07/28/abbott-immigrant-travel-501378
thanks for information that the protestors (as opposed to the Capitol insurrectionists) are also being questioned and prosecuted. I've evidently missed this salient point on PBS, MSNBC and CNN. This would answer some Republican congresspersons' complaints that the attention given the insurrection is biased because "all" protests are supposedly crimes.
Are we seeing a "spike" in cases? No, not really. Look closely. Deaths are down. Serious cases are down. We are seeing a spike in positive test results.
The test for covid gives at least 10% false positives. Some say 15% to 20%. As people get vaccinated and the prevalence of a disease goes down, the rate of false positives goes up. Here's why: Suppose the prevalence is 2%, as it was for covid last year. If you test positive, and the test gives 10% false positives, what are the odds that you actually have covid?
If you said, 90%, that is flatly wrong. At 2% prevalence, you must do 50 tests, on average, to get one true positive. Those 50 tests will give five false positives. Your true odds of having it are one in 6, or 18%.
But what happens when the prevalence falls to 1%, as it has for covid? Now you must do 100 tests, on average, to get one true positive. Those hundred tests give ten false positives, so now your odds of actually having it are one in 11, or only 9%.
What looks like a spike in positive test results is actually a rise in false positives caused by the lower rate of prevalence.
Shawn -- but when false positives greatly outnumber true positives, shouldn't the rate of positive test results (false + true) hold relatively steady as prevalence rises and falls? If (as in your example) false positive rate is 10%, then the number of positive test results for 100 tests at 2% prevalence should be 12. At 1% prevalence the number of positive test results for 100 tests should be 11.
So even if the odds are low that an individual with a positive test result actually has covid, a spike in *absolute numbers of positive test results* has to reflect either (1) more testing, or (2) more true positives. A spike in *rate of positive test results* has to reflect either (3) a higher rate of false positives, or (4) a higher rate of true positives.
A rise in false positives *relative to* true positives shouldn't push the total rate of positive test results up, should it?
Do you have a specific set of sources or real data we could work off here?
Thanks,
Lark
I tried to make it simple, Lark, but in fact it is somewhat complicated. Testing fails when the rate of false positives is high relative to the prevalence of the disease. You see this, for example, in testing for lung cancer (which many physicians decline to do for that very reason) and in testing for HIV, and even for breast cancer in some age groups. Mammograms give 9% false positives, as some of my friends who had a scary one are well aware, and although the prevalence of breast cancer varies widely among subsets of the population, in general, your odds of actually having breast cancer after a positive mammogram are only about 15%.
It is what statisticians call Type 1 Error (type 2 being false negatives). A little Google research on the terms I have used here will tell you more than you want to know. The bottom line is that most--far more than half--of all reported covid cases are false positive tests on people who are sick for other reasons. They then get misdiagnosed with covid because it pays more than what they actually have.
The impact of false positives is so high because the prevalence of covid is relatively low. And thankfully declining, although that makes testing matters even worse. Your intuitive guess--that a positive test result implies 90% chance of infection--assumes 100% prevalence. The actual prevalence of covid, today, is maybe 1%. So the actual chance that someone has it, after a positive test, is therefore less than 10%.
Shawn -- it sounds like you're explaining -- accurately -- why testing *for an individual* is currently too likely to produce a false positive.
But your initial comment started *and ended* "Are we seeing a "spike" in cases? No, not really. ... We are seeing a spike in positive test results" and "What looks like a spike in [cases] positive test results is actually a rise in false positives caused by the lower rate of prevalence." That is a comment alleging that testing *on the population level* is currently not useful, which is a different point and needs a different set of data to support it.
My core question was: "A rise in false positives *relative to* true positives shouldn't push the total rate of positive test results up, should it?" What's your answer to that? I still don't see why a spike in *rate of* positive test results could NOT reflect also an actual spike in *number of* cases, even if the *rate of* cases (i.e. prevalence) is goes down.
Thanks!
Lark
Look at it this way: If there were no covid, at least 10% of the population would still test positive. By testing everyone in the US, you would identify at least 33 million "confirmed" cases, all false positives. (We have actually seen 36 million cases.)
Or else, look at it this way: If for some reason you wanted to show a spike in cases, all you would need to do is more testing. When you do 10% more tests (as we recently have), you get 10% more cases (which we are seeing), even if the true incidence of covid declines (as it probably has).
So, when you read media scaremongering about rising case loads, look first for the type 1 error that they never even mention, much less account for. Then check to see whether they are conflating incidence (people who get the disease) with prevalence (people who have the disease), as they almost always do. By definition, the prevalence of having covid (ie, testing positive) cannot fall below the type 1 error rate of at least 10%, even if the incidence (ie, true new cases) falls to zero.
Shawn -- I'm really not seeing the numbers you're seeing (and I'm responding so late because I didn't make the time earlier to check - sorry about that). What test are you talking about? The new home tests have known very significant false positive rates; the PCR tests, however, MUCH less so. Where are you getting your data about the positivity rates of different tests?
You write - "By definition, the prevalence of having covid (ie, testing positive) cannot fall below the type 1 error rate of at least 10%, even if the incidence (ie, true new cases) falls to zero." Yet CDC's https://covid.cdc.gov/covid-data-tracker/#trends_dailycases_7daytestingpositive clearly shows that for months earlier this year the positivity rate was below 5%. https://covid.cdc.gov/covid-data-tracker/#cases_positivity7day shows that the TOTAL tests recorded by CDC *ever* is 507million to get 38million positive (data posted evening of Aug 19)-- significantly less than 10%. There are currently many states in the U.S. showing positivity rate of less than 8%.
Please share the numbers you're seeing on which you're basing your conclusions, and the sources of those numbers.
I don't want to quibble over numbers, Lark. What matters is cases.
But, what is a case? Someone who truly dies of covid? Someone who dies from something else but with covid? Someone who gets sick and tests positive for covid? Who isn't even sick but tests positive? Who died in a car accident and later tested positive?
We are subjected to a constant media barrage of what I call pandemic porn. Scare tactics, because that is what people want to click on. For the same reason we watch A Quiet Place: because we want to be scared.
Now, Lark, I will tell you a story; one that, this late in comments, only you (and just maybe Isaac) will ever see here. I was a covid denier from the very start. Always thought it was way overblown. Never went masked if I didn't have to. Took zero precautions, and I fly twice a month between my homes in Boston and Cape Coral, FL.
For five or six years, until 2014, I was with a woman who was the true love of my life. Alas, we were both married, although not to one another. Then she left me, without a word--neither then nor ever since. Two weeks ago, I learned that she died last May (on my birthday) at age 69. She was so healthy, although she did smoke. It could only have been covid.
Since then, in the last two weeks, my feelings about covid have totally changed. Now, even I am scared. I miss her so much.
I don't know how you mandate vaccinies that are only authorized for emergency use. I think everyone over 60 and those with dangerous co-morbidities should all get the vaccine as the death rate is greater than a normal flu risk.
I am old enough to remember Fen Phen, Thalidomide, VIOXX & Bextra, Quaaludes, Darvon, DES, PTZ, Baycol. We have no idea on what the near or long term effects may be. I cannot imagine giving this to under 20's to whom the danger is low (similar to regular flu). I think adults between 20 and 60 without dangerous co-morbidities should decide for themselves.
Will companies that mandate a vaccine be liable for any long term problems?