Rand Paul: False Positive?

Dems are trashing Paul for not quarantining before the test results:

https://www.yahoo.com/news/sen-kyrsten-sinema-calls-rand-025700153.html

But he was tested due to general high levels of contacts, not for symptoms or a specific contact with a known infected.

Statistically, I believe we are still at an infection level where a positive result on an asymptomatic person who has not had substantial exposure to an infected is more likely a false positive than a true positive.

Meanwhile, Paul is out mowing his lawn while yelling : “Come get me now!” at his neighbor.

9+
avataravataravataravataravataravataravataravataravatar

19 thoughts on “Rand Paul: False Positive?”

  1. ctlaw:
    Statistically, I believe we are still at an infection level where a positive result on an asymptomatic person who has not had substantial exposure to an infected is more likely a false positive than a true positive.

    Any idea what the false positive rate is for the current test? No one seems to talk about that. I know from one of the daily briefings that the test being put forth by WHO had a false positive rate in the 40-50% range according to Dr. Brix: useless.

    0

  2. drlorentz:
    Any idea what the false positive rate is for the current test? No one seems to talk about that. I know from one of the daily briefings that the test being put forth by WHO had a false positive rate in the 40-50% range according to Dr. Brix: useless.

    It all depends on the denominator. Is it total tests or total positives? Is the group being tested limited to the symptomatic?

    If you are testing symptomatic people, then 50% of positives being false is useful if that’s a low overall number.

    If one in 10 symptomatic people actually have the disease and the test flags him plus one false positive, it’s still useful. But if you give the test to 1 million asymptomatics, you will get 100000 false positives and perhaps 5000 true positives. Starts becoming less useful.

    4+
    avataravataravataravatar
  3. ctlaw:

    drlorentz:
    Any idea what the false positive rate is for the current test? No one seems to talk about that. I know from one of the daily briefings that the test being put forth by WHO had a false positive rate in the 40-50% range according to Dr. Brix: useless.

    It all depends on the denominator. Is it total tests or total positives? Is the group being tested limited to the symptomatic?

    If you are testing symptomatic people, then 50% of positives being false is useful if that’s a low overall number.

    A test with 50% false positives is almost never useful. As it is right now, the number of people in the US who test positive is around 10%, and that’s with strict criteria on who gets tested. A 50% false positive rate would erroneously flag as many people as there are true positives. Even “symptomatic” people often end up being negative.

    ctlaw:
    If one in 10 symptomatic people actually have the disease and the test flags him plus one false positive, it’s still useful.

    Incorrect. If the false positive rate is 50%, there will be 4. 5 (statistically) false positives and 1 true positive. This is a disaster. It’s hard to concoct a scenario in which 50% rate is useful. Definition of specificity & false positives.

    I’m still wondering what the false positive rate for the current tests is. I suspect no one knows because you need many cases and a “gold standard” to compute it.

    1+
    avatar
  4. drlorentz:
    A test with 50% false positives is almost never useful. As it is right now, the number of people in the US who test positive is around 10%, and that’s with strict criteria on who gets tested. A 50% false positive rate would erroneously flag as many people as there are true positives. Even “symptomatic” people often end up being negative.

    Note I prefaced my point with “…Is it total tests or total positives?…” I realize we are dealing with press reports which often confuse numbers.

    1+
    avatar
  5. ctlaw:

    drlorentz:
    A test with 50% false positives is almost never useful. As it is right now, the number of people in the US who test positive is around 10%, and that’s with strict criteria on who gets tested. A 50% false positive rate would erroneously flag as many people as there are true positives. Even “symptomatic” people often end up being negative.

    Note I prefaced my point with “…Is it total tests or total positives?…” I realize we are dealing with press reports which often confuse numbers.

    Dr. Brix, not the press, said 50% false positives for the WHO test, as I explained. I heard the words come out of her mouth. She is also the one who has been citing the values for percent positive tests in the US. She calls it “percent positivity.”

    I’ve still not heard anyone report on the false positive rate for the current WuFlu test. The seasonal flu test has a false positive rate of ~10%, if I recall.

    0

  6. Here’s what I’m still trying to figger out:

    are there actually asymptomatic carriers?  Typhoid Marys who will never get sick themselves?

    or are there just PREsymptomatic individuals, who will in a day or so develop  at least mild symptoms?

    how do we know that anyone is getting the virus from either of the above categories, meaning, at what point does an infected individual shed enough of the virus  for contagion?

    All I can say is, I HOPE the scientific community is better informed than we are, because right now, I guarantee that you can find stuff on the internet to conclusively support any theory that may occur to you.

    2+
    avataravatar
  7. Hypatia:
    are there actually asymptomatic carriers?  Typhoid Marys who will never get sick themselves?

    According to the paper on the Diamond Princess cases, yes there are truly asymptomatic carriers. This was a good test case because everyone was tested and everyone was monitored for at least one month.

    Don’t almost all diseases have true asymptomatic cases? That seems to be the rule, not the exception. About 70% of polio (a virus) cases were asymptomatic.

    1+
    avatar
  8. drlorentz:

    Hypatia:
    are there actually asymptomatic carriers?  Typhoid Marys who will never get sick themselves?

    According to the paper on the Diamond Princess cases, yes there are truly asymptomatic carriers. This was a good test case because everyone was tested and everyone was monitored for at least one month.

    then, as I have asked before: since these people’s immune systems arent fighting the virus (because if they were, they’d be  sick, right?)—then when,  if ever,  will they NOT be carriers?*

    Don’t almost all diseases have true asymptomatic cases? That seems to be the rule, not the exception. About 70% of polio (a virus) cases were asymptomatic.

    i have NO idea, s’why I’m asking here.

    *PS: never mind, I looked it up.  The answer is never..  Typhoid Mary had  to be imprisoned on some li’l island near NY for the rest of her life.

    1+
    avatar
  9. Hypatia:

    drlorentz:

    Hypatia:
    are there actually asymptomatic carriers?  Typhoid Marys who will never get sick themselves?

    According to the paper on the Diamond Princess cases, yes there are truly asymptomatic carriers. This was a good test case because everyone was tested and everyone was monitored for at least one month.

    then, as I have asked before: since these people’s immune systems arent fighting the virus (because if they were, they’d be  sick, right?)—then when,  if ever,  will they NOT be carriers?*

    Don’t almost all diseases have true asymptomatic cases? That seems to be the rule, not the exception. About 70% of polio (a virus) cases were asymptomatic.

    i have NO idea, s’why I’m asking here.

    *PS: never mind, I looked it up.  The answer is never..  Typhoid Mary had  to be imprisoned on some li’l island near NY for the rest of her life.

    You are confusing asymptomatic with noncontagious. As I already pointed out, about 70% of poliomyelitis cases are asymptomatic. So the answer is not never. The whole problem with asymptomatic carriers is that they are contagious. That’s why they call ’em carriers. If you’ve got the bug, you can probably pass it along even if you don’t have symptoms.

    Sheesh!

    1+
    avatar
  10. drlorentz:

    ctlaw:

    drlorentz:
    Any idea what the false positive rate is for the current test? No one seems to talk about that. I know from one of the daily briefings that the test being put forth by WHO had a false positive rate in the 40-50% range according to Dr. Brix: useless.

    It all depends on the denominator. Is it total tests or total positives? Is the group being tested limited to the symptomatic?

    If you are testing symptomatic people, then 50% of positives being false is useful if that’s a low overall number.

    A test with 50% false positives is almost never useful. As it is right now, the number of people in the US who test positive is around 10%, and that’s with strict criteria on who gets tested. A 50% false positive rate would erroneously flag as many people as there are true positives. Even “symptomatic” people often end up being negative.

    ctlaw:
    If one in 10 symptomatic people actually have the disease and the test flags him plus one false positive, it’s still useful.

    Incorrect. If the false positive rate is 50%, there will be 4. 5 (statistically) false positives and 1 true positive. This is a disaster. It’s hard to concoct a scenario in which 50% rate is useful. Definition of specificity & false positives.

    I’m still wondering what the false positive rate for the current tests is. I suspect no one knows because you need many cases and a “gold standard” to compute it.

    ‘Splain this.

    False positives are tests that falsely show positive results. They are a percentage of total positive tests. So a 50% false positive rate would mean of all the positive results, 50% were falsely positive. So CT has a point in that IF you are testing a subgroup with criteria for the disease, then you have to have some idea of how big the actual attack rate is. Take an arbitrary number of 100 symptomatic patients. IF the attack rate is 20% and the false positive rate is 50%, then of those 100 patients you would expect 40 to test positive. That IS useful in that you now know the other 60 symptomatic patients (disregarding false negatives) do NOT have the disease. The 40 will now require further study.

    This is pretty much how cardiac triage runs. 2 negative troponins separated by at least 3-4 hours, 2 negative EKG’s likewise separated, no cardiac history and no family history of cardiac disease pretty much mean you go home. They do NOT mean you have no cardiac disease; they only mean the odds of you having a significant cardiac attack are slim, but you may have muscle at threat. Show a positive in that group, especially a troponin, and you end up being admitted, even though you may rule out as having significant coronary disease.

    Take the same criteria and make them a general screening test and you get different results. European Fighter Command decided to run routine stress tests on ALL their fighter pilots. They got a significant number of positive treadmills. Yet the cath, which ALWAYS follow a positive treadmill in the Air Farce, were negative in most of the cases. It’s why the AF has the single largest collection of negative coronary caths in the world; they don’t pay attention to the difference between screening and evaluation for symptoms.

    2+
    avataravatar
  11. Devereaux:
    False positives are tests that falsely show positive results. They are a percentage of total positive tests.

    Nope. I’ll stop you right there and refer you to the definition of the false positive rate (FPR):

    false positive rate = (false positives)/(real negatives)

    I’m not going to read the rest because you are using a nonstandard definition of the term, presumably not the one being used by the pros. Your denominator of “total positive tests” is incorrect. With a little algebra, you can also come up with this:

    FPR = 1 – TNR,

    where TNR is the true negative rate, which is the true negatives divided by all the negatives. Distinguish “true negatives” from “real negatives.” The former is the number of negative test outcomes, while the latter is the number of negatives found by a perfect test.

    A 50% false positive rate is particularly terrible when testing a population that is predominantly negative because half of those negatives will be erroneously categorized as positive. This is the situation that obtains in the COVID-19 pandemic: a tiny fraction of the population is actually infected. Even those who qualify under the relatively stringent requirements for testing (symptoms, travel & contact history) have low positivity according to Dr. Brix (as of a couple of days ago).

    Random testing of the entire population would inflate the numbers of cases tremendously even if the FPR were only 10%. As I’ve explained on other threads, that result would be extrapolated to the entire population of 330M Americans to conclude that 33M people were infected. Horrible.

    Did I ‘splain it good?

    0

  12. drlorentz:

    Hypatia:

    drlorentz:

    Hypatia:
    are there actually asymptomatic carriers?  Typhoid Marys who will never get sick themselves?

    According to the paper on the Diamond Princess cases, yes there are truly asymptomatic carriers. This was a good test case because everyone was tested and everyone was monitored for at least one month.

    then, as I have asked before: since these people’s immune systems arent fighting the virus (because if they were, they’d be  sick, right?)—then when,  if ever,  will they NOT be carriers?*

    Don’t almost all diseases have true asymptomatic cases? That seems to be the rule, not the exception. About 70% of polio (a virus) cases were asymptomatic.

    i have NO idea, s’why I’m asking here.

    *PS: never mind, I looked it up.  The answer is never..  Typhoid Mary had  to be imprisoned on some li’l island near NY for the rest of her life.

    You are confusing asymptomatic with noncontagious. As I already pointed out, about 70% of poliomyelitis cases are asymptomatic. So the answer is not never. The whole problem with asymptomatic carriers is that they are contagious. That’s why they call ’em carriers. If you’ve got the bug, you can probably pass it along even if you don’t have symptoms.

    Sheesh!

    I think, dear  sir, if you prayerfully reread my comments and questions, you will be guided to see, God willing, what I have been asking is when, if ever, asymptomatic carriers become non-contagious.  So,  no, I am NOT confusing “asymptomatic” with “non contagious”

    In Typhoid Mary’s case (and she is the best-known of asymptomatic carriers)  apparently never, till she died.  Her contemporaries believed she was still contagious even though asymptomatic.  Apparently the unfortunate Ms. Mallon obstinately wanted to go on eating, so she wanted to go on plying the only trade she knew: cooking and food preparation.  As a result she was forcibly pluckedOUT of  society and imprisoned.  (I express no opinion as to the necessity of this long-ago  public health decision. )

    Now, ancora una volta   (as they’d say in Italy if anyone were still alive):

    if “never” is not the answer, and if you know:

    At what point do asymptomatic carriers become noncontagious?

    (oh and “sheesh!” yerself, y’old rattlesnake! 😘)

    0

  13. Hypatia:
    I think, dear  sir, if you prayerfully reread my comments and questions, you will be guided to see, God willing, what I have been asking is when, if ever, asymptomatic carriers become non-contagious.  So,  no, I am NOT confusing “asymptomatic” with “non contagious”

    I’m sorry that I misunderstood your comment. I was confused by the proximity of your P.S. to a different question. My mistake.

    I’m still confused about what you are asking because after quoting my question “Don’t almost all diseases have true asymptomatic cases?” you wrote “i have NO idea, s’why I’m asking here.” So are you asking if diseases commonly have true asymptomatic cases? I think the answer is pretty clearly yes. Perhaps that’s not the question you meant to ask.

    0

  14. drlorentz:

    Hypatia:
    I think, dear  sir, if you prayerfully reread my comments and questions, you will be guided to see, God willing, what I have been asking is when, if ever, asymptomatic carriers become non-contagious.  So,  no, I am NOT confusing “asymptomatic” with “non contagious”

    I’m sorry that I misunderstood your comment. I was confused by the proximity of your P.S. to a different question. My mistake.

    very gracious, thank you.

    I’m still confused about what you are asking because after quoting my question “Don’t almost all diseases have true asymptomatic cases?” you wrote “i have NO idea, s’why I’m asking here.” So are you asking if diseases commonly have true asymptomatic cases? I think the answer is pretty clearly yes. Perhaps that’s not the question you meant to ask.

    Here’s what’s bothering me: we’re now being told that a lot of apparently healthy young people are carriers.  If that’s true, what are we looking at?  Will the young  have to avoid the old FOREVER?  How will we ever know at what point it is safe to go to  grandma’s 85th?

    0

  15. 10 Cents:
    I think it depends on the disease in this article whether a person is asymptomatic carrier.

    I think there’s some confusion going around about the meaning of the term “asymptomatic carrier”.  The term is well-defined, meaning somebody who is infected with the virus, is shedding the virus in a manner which can infect others, yet manifests no symptoms of the disease and may be unaware they are infected.

    Many diseases cause a wide range of symptoms in those infected: some of those infected may show no symptoms at all (“asymptomatic”) and the the presence of the disease can only be established by a biochemical assay.  But the normal course of the disease in most infected people, whether they are symptomatic or not, is that the infection will run its course over a period of time until the host body’s immune system is trained to resist it, after which it will be killed off in that host, who will henceforth not be a carrier and unable to infect others.

    There is nothing about a host’s being asymptomatic which means they will be a carrier forever.  The usual case is that they will remain an (unknowing) carrier for about as long as somebody with symptoms, after which they will no longer be a carrier.

    Thus asymptomatic carriers are dangerous in that they may continue to go to work and associate with people in crowds, unwittingly spreading the disease, but not that they will remain infectious for all time.

    Now, there may be cases of certain diseases where this phenomenon exists, and if the immunity that develops in a carrier only lasts a limited time or is restricted to one specific strain and useless against mutants, such as regularly appear in the regular influenza, somebody who has recovered from the disease may become subsequently re-infected, but again that can happen regardless of whether the infected person develops symptoms.

    0

  16. Okay, @johnwalker, you are the first person who has addressed my question.  You say someone doesn’t remain an asymptomatic carrier any  longer than they remain a symptomatic carrier .  Thank you!

    0

Leave a Reply