Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
Medicine News

Showing Cold Symptoms, Elon Musk Tests Positive - and Negative - for Covid-19 (seattletimes.com) 216

"Elon Musk predicted in March, at the beginning of the coronavirus pandemic, that there would be 'probably close to zero new cases' in the United States by the end of April," reports the Washington Post.

"Now, the Tesla chief executive is trying to figure out whether he has a virus that has killed at least 242,000 Americans." The billionaire said early Friday that he's experiencing cold-like symptoms, but that four rapid tests have produced two positives and two negatives — an experience that left him questioning the process.

"Something extremely bogus is going on," Musk, 49, tweeted. "Was tested for covid four times today. Two tests came back negative, two came back positive. Same machine, same test, same nurse..."

UPDATE: Thursday Musk tweeted he was also in the process of getting "PCR tests" — plural — "from separate labs," with results to be delivered within 24 hours. On Saturday, Musk then tweeted the news that he "most likely" has a moderate case of Covid-19.

Musk had said he had "symptoms of a typical cold," tweeting that for the past few days he's had "slight fever" and "mild sniffles" — and that he'd been taking NyQuil.
This discussion has been archived. No new comments can be posted.

Showing Cold Symptoms, Elon Musk Tests Positive - and Negative - for Covid-19

Comments Filter:
  • Duh (Score:3, Insightful)

    by Aighearach ( 97333 ) on Friday November 13, 2020 @10:34PM (#60722336)

    I really didn't realize he was this stupid. I thought he was just a comedian.

    But this isn't even funny.

    Like, dude, don't you have to study statistics and probability to be an engineer... or a business person?

    Rapid tests are rapid because of a tradeoff in precision and accuracy.

    • Rapid tests for other viral infections are highly specific; the likelihood of false positive or negative is very low. Why the issues with rapid antibody ELISA tests aren't being resolved or covered by media is deeply disturbing to those who actually know what normal front line and confirmatory testing looks like and how standards of lab diagnostics have deteriorated amid the coronapocalypse. This is not normal and it undermines confidence in lab diagnostics.
      • Re: Duh (Score:4, Insightful)

        by Aighearach ( 97333 ) on Saturday November 14, 2020 @05:35AM (#60723152)

        No, it is disturbing for idiots who haven't noticed all year when people get a test, then they get tested again, they're always also waiting on the results of additional tests.

        That's the way it works with a novel virus. The tests aren't that good at first. So you take them multiple times. All year. Everybody has been needing multiple tests.

        Confidence in lab diagnostics is for people who understand numbers. For everybody else it is magic, and all you need to know is this: The doctors you have are the ones you have to trust. Your magical thinking is deep. But no, you don't get to choose your timeline. This timeline has these various tests.

      • by DRJlaw ( 946416 )

        Why the issues with rapid antibody ELISA tests aren't being resolved or covered by media is deeply disturbing to those who actually know what normal front line and confirmatory testing looks like and how standards of lab diagnostics have deteriorated amid the coronapocalypse.

        As one of those who actually know, this is bullshit. The rapid tests that you're comparing this to weren't developed and deployed in less than 6 months. They were painstakingly developed over years and years in order to optimize true

    • by c-A-d ( 77980 )

      That's a really bad tradeoff for either a 50% false positive rate or a 50% false negative rate.

      • by AmiMoJo ( 196126 )

        Assuming he is telling the truth, and assuming the tests were properly administered. The rapid tests are about 75% accurate if administered by a trained healthcare professional, or about 50% accurate if administered by a trained member of the public or military.

        Anyway he must know by know, assuming he did the sensible thing and got a more reliable test immediately which takes several hours to produce a result.

      • 1) We know he has it, we don't have to pretend to wonder at the answer
        2) No, it isn't a "bad tradeoff" it is a rapid test, and repeated tests confirmed detection of the virus.
        3) Not finding what you were checking for is not at all the same as finding what you were checking for. This test is not designed to tell you that you don't have the virus. It either detected it, or didn't detect it. That is all. It is perfectly normal and expected for some people to have a level of the virus that would be detected 50

    • PCR test time, doofus.

    • What about this is stupid? He got got sick and got a test (already smarter than a lot of dumb fucks out there who just go to work like normal). He went and got the recommended test, disagreed with the result and is now awaiting a PCR test. If you consider that stupid you may as well question most of the population of America.

      Now on the topic of testing as well the rapid test was not generally considered to lack repeatability. It lacked specificity meaning it was likely to produce false negative tests, but i

      • by fazig ( 2909523 )
        You're talking to the guy who said that "influenza != flu" and told me that if I had looked it up on Wikipedia I'd know better. When I did do that and quoted the article, he claimed that the Wikipedia article was wrong.

        Perhaps he's not an expert on infectious diseases.
    • I am kind of surprised at his statements too. I knew this stuff before I got out of HS. Out of UG, I understood the mechanics and statistical reasoning behind it. Forget engineering, how does he understand high level factory yields... especially in auto industry, and not have the basics here?

    • 50% failure in accuracy means the test is no good and a waste of time. Musk is just pointing this out
      • Re: Duh (Score:4, Interesting)

        by GuB-42 ( 2483988 ) on Saturday November 14, 2020 @10:16AM (#60723712)

        If that's what he is pointing out, then he doesn't understand basic statistics.
        To make up if the tests are worth it, one must consider prior probabilities, false positive rate (aka specificity) and false negative rate (aka sensitivity) and run Bayesian analysis. So you need at least these 3 numbers in addition to the test results in order to decide if the results are meaningful.
        Considering how these tests are designed, he is most likely to be positive for real.

        Here is a little story to illustrate that. Imagine a perfect test: you press a button and a little light turns on if you have covid, it works by magic so no mistake is possible. The only problem is the button: 50% of the time, it doesn't register the press and therefore the light don't turn on, and because there is no control light, you don't know if the light is off because you don't have covid or because of a mispress.

        You press the button 4 times.
        If you have covid, you can expect the light to turn on 2 out of 4 times: 2 mispresses and 2 actual positive tests, as it happen in Musk's case. There is absolutely no way it can happen if you don't have covid.
        If you don't have covid, the light will never turn on.

        In reality, it is more complicated than that, we don't have magical tests and false positives are always a possibility. But a 2/4 positive test doesn't mean the test is insignificant.

      • Don't think you can say that. How good the test is at detecting the presence of the virus and how good it is at detecting the absence of the virus are probably different specs.

        Also, probably how much of what the test is testing for (depends on test) is present in the sample, would be a factor too in the test's accuracies.

        But take that with a grain of salt, I'm not qualified, and didn't check my hypotheses.

    • Is the rapid test he used an FDA approved testing method or one of the tests given a waiver? Are the tests a random number generator that randomly say positive or negative? I'd think the accuracy would be closer to 95% at the least.
    • What are you talking about? I was assured by him and Trump that the virus is gone.
    • by MrL0G1C ( 867445 )

      The summary doesn't mention the likelihood of false negatives or false positives. Most tests are tailored to be accurate one of those ways.

      In the case of False positives only happening 0.5% of the time then he's very likely got COVID.
      In the case of False negatives only being wrong 0.5% of the time then he very likely hasn't got COVID.

      A quick scan of the article by SeattleTimes shows nothing useful with regards to this and since I don't know the exact name and perhaps version of the test he took, I can't dra

  • He's Positive (Score:5, Informative)

    by Ksevio ( 865461 ) on Friday November 13, 2020 @10:40PM (#60722370) Homepage

    Was tested for covid four times today. Two tests came back negative, two came back positive. Same machine, same test, same nurse..."

    The false positive rate on those tests is pretty low, but the false negative is pretty high. Chances are he's positive

    • The false positive rate on those tests is pretty low

      Actually you have that totally backwards.

      A BD spokesperson said the company is aware that "a small number of nursing homes in the U.S. are reporting multiple false positive results" from tests run on its machines

      https://www.medtechdive.com/ne... [medtechdive.com]

      • by raymorris ( 2726007 ) on Friday November 13, 2020 @11:09PM (#60722496) Journal

        You point out a news article in which one particular company reportedly had some false positives, which is surprising because in lab tests their test didn't have false positives.

        According to a Harvard study and the CDC, the most common type of tests has a false negative rate up around 20% and a false positive rate of about 3% which the false negatives increasing if the test isn't taken within the first few days of symptoms.

        If the antibody level is low, that'll give a negative - there just isn't enough antibody to register.

        Musk probably had a low level of antibodies on this particular day. Not zero, so one test picked up on it, but low enough that the other test didn't detect it.

        If you have four smoke alarms in your house and two are going off, something is probably burning.

        • According to a Harvard study and the CDC, the most common type of tests has a false negative rate up around 20% and a false positive rate of about 3% which the false negatives increasing if the test isn't taken within the first few days of symptoms.

          Wouldn't you want it the other way around? A high false negative rate allows infectious people to go about in public under the false assurance that they are not sick, at great harm to the public. A high false positive rate OTOH only results in quarantining som

          • Re: (Score:3, Interesting)

            by Kejiro ( 2803123 )

            Wouldn't you want it the other way around? A high false negative rate allows infectious people to go about in public under the false assurance that they are not sick, at great harm to the public. A high false positive rate OTOH only results in quarantining some people who aren't sick as if they are sick, and the harm is limited to those individuals. If you want to err on the side of safety, you want to tune your tests to have as close to a zero false negative rate as possible, even if that makes the false positive rate spike. Preventing one sick person from going about and potentially infecting others prevents more harm, than quarantining several people who are not sick.

            I think it depends on which side you're on. As a health care professional and probably most people, a high false positive is preferable because of what you say. Then we have the unscrupulous business owners and others that just want people to get back to work and their normal lives no matter how many gets sick and dies ;)

            I don't know how those tests work, but I can imagine that a false negative is easier to get than a false positive. I assume that the tests work by testing the presence of something in the b

          • Yeah, in this case probably the other way around might be better. Particularly because a better test, the PCR test, is available. When a better test is available, it's handy to have a screening that can give one of two answers:

            Negative: You don't have the disease
            Potentially positive: The better test should be done

            This test, like most, works by detecting something in the abnormal case (infected). So the actual possible results are:
            A. Infection detected
            B. Nothing detected

            It detects active antibodies.

          • Wouldn't you want it the other way around?

            Yes you would. But people didn't make some trade off when designing the antigen test. The PCR test is significantly more difficult and time consuming. Which brings you to the applicability for basic screening. If you're in quarantine the answer is easy, you want the PCR test with its higher specificity. However if you're random sampling a population not in quarantine do you want an antigen test that gives results straight away and may instantly pick out people for quarantine potentially missing infected peo

        • You point out a news article in which one particular company reportedly had some false positives, which is surprising because in lab tests their test didn't have false positives.

          Yes, the EXACT company in question - BD. If you'd read TFA, you'd see that Elon was talking about the BD test, and the article I linked to above is specific to that machine.

          I mean how much more evidence do you want that the BD has unacceptably high false positives?
          From the FDA:

          Consider any positive result presumptive from tests using the BD SARS-CoV-2 Reagents for the BD Max System. Consider confirming with an alternate authorized test.

          https://www.fda.gov/medical-de... [fda.gov]

    • Well before I really understood a pregnancy test I thought the accuracy was in question. Once I understood the test, only a false negative is possible. You do not produce HCG unless you are pregnant. Likewise I think a test that sees the presence of a coronavirus would be enough to err on the side of caution. A false positive is only likely if he had contracted some other coronavirus. Best to play it safe.

      • by jabuzz ( 182671 )

        I suggest you look into that again. There are a whole range of reasons why there might have elevated HCG levels in a womens urine and she is not actually pregnant. To get you started I suggest reading the following

        https://www.healthline.com/health/pregnancy/false-positive-pregnancy-test

        • You do realize the very article you linked had to reach pretty far to get you 7 reasons. For starter the first 3 ARE pregnancy related. Just because the the egg failed to implant properly does not mean you were without a fertilized egg. In fact only #6 and, possibly, #7 are a valid cause. If you are taking hcg as a therapy you are already under the care of a fertility doctor or OB and would have told you not to take an EPT. Misreading a faint line and thinking its just light in color? Come on, thats not ev

    • With your analysis method, any healthy person that is tested will be positive.
    • Musk is fantastic at debunking meritocracy.
      • Does your shit stink Elon ? âoeUh No.â Probably got COVID-19 . Elon-Since I was a child and can remember? Med - fever, delusions, probably COVID-19.
  • by WorBlux ( 1751716 ) on Friday November 13, 2020 @10:48PM (#60722410)

    is this relevant to anything or even slightly newsworthy?

    • Re: (Score:2, Funny)

      Lord Musk's life may be in danger! If he dies, we all die!

      • by gweihir ( 88907 )

        Lord Musk's life may be in danger! If he dies, we all die!

        Dude, we are _all_ going to die! It usually happens around 80 years after you were born.

      • by AmiMoJo ( 196126 )

        It's worth amplifying the fact that one of the most prominent COVID sceptics has now got it himself. Might make some people who believed him reconsider their decision to ignore it.

        The long term effects could be devastating for someone like Musk who is constantly busy, but he will get the best healthcare money can buy and isn't in any of the major risk categories (as far as we know) so has a good chance of making a full recovery.

        • It's worth amplifying the fact that one of the most prominent COVID sceptics has now got it himself. Might make some people who believed him reconsider their decision to ignore it.

          Trump already got it and nobody changed their mind as to weather it was real or not. Musk isn't even a rounding error compared to that.

      • Lord Musk's life may be in danger! If he dies, we all die!

        We'll all die anyway.
        If Musk dies we all die on Earth. Otherwise a handful will die on Mars instead.

    • It's kind of interesting that even with all the resources in the world (4 tests in a day) it points to some shortfall in the testing technology.

      That said, perhaps he just happened to be near the threshold of detectability today. Clearly a biological process like infection is not instantaneous and on the progression from negative to positive there must be an indeterminate point between.

      • The test sensitivity threshold example is mildly interesting, but this is still mostly reading as a celebrity gossip piece to me.

      • perhaps he just happened to be near the threshold of detectability today.
        If that was the case, he would not have had two positives.

    • It is, but probably not for you ...

    • is this relevant to anything or even slightly newsworthy?

      It would be an issue if there were other cases at SpaceX, given that they are scheduled to launch astronauts to the ISS Sunday (postponed from Saturday due to weather).

    • Musk is positive AND negative.
  • PCR is easy (Score:3, Insightful)

    by backslashdot ( 95548 ) on Friday November 13, 2020 @10:58PM (#60722448)

    Instead of twittering like a president on crack, maybe he could just run the PCR test himself?

    Itâ(TM)s easy .. takes about an hour. Then he can run a gel, confirm that he got specific amplification (aka a nice band, not a smear, of the expected length of DNA). Then, as additional proof he can excise that band from the gel and sequence it either conveniently on a USB nanopore sequencer (yes nanopore sequencers have an error rate but itâ(TM)s good enough since you donâ(TM)t need exact bp match) or Sanger sequence it. That would have a near zero false positive rate literally in the one in trillions.

    The whole thing wonâ(TM)t take long, high school kids can do it.

    • Re:PCR is easy (Score:5, Interesting)

      by Anonymous Coward on Friday November 13, 2020 @11:32PM (#60722562)
      You got most of that right. Sequencing for Covid (or any coronavirus) has some important caveats though.

      First, they're RNA viruses, so you need to do reverse transcription PCR. You could say "reverse transcription PCR is just one enzyme different from regular ol' PCR", and you'd be mostly right. Hence all the other requirements, including good primers and tuning of annealing temps, etc apply as well.

      Second, RNA - even viral genome RNA - is inherently less stable than DNA. This is largely because RNAse enzymes are extremely ubiquitous and very very hardy. You can put human RNAses through an autoclave and they'll still chew up loose RNA.

      Third, you have to hit the sweet spot on specificity. We've already observed several variants of Covid19 in the wild (including the famous example of the poor bastard in Las Vegas who has caught two of them at different times). You want to make sure you're loose enough to catch Covid19 variants without being so loose that you're catching standard conoronaviruses that bring so many standard seasonal colds around.

      There's a reason why we're looking for more certified lab techs right now to handle these samples, and why we're looking for better bioinformaticians to set up the processing pipelines for the data that comes from running the PCR tests.

      I'm writing this as someone who holds a PhD in biochemistry. I've done this. I recently had a recruiter reach out to me to see if I'd be willing to help with a local Covid19 diagnostic facility (I politely declined).
      • I use to do work in bioinformatics pipelines. Background in computer science and mathematics with experience in bioforensics. Now I teach rudimentary English in podunk China. I have a relatively easier life with a higher standard of living. It's funny... you never appreciate what you have until it's lost which while brain drain is an issue here in China, the impact in America is underestimated too... Many educated people without attachments simply want to understand more of the world than the average

    • Instead of twittering like a president on crack, maybe he could just run the PCR test himself?

      There's a lot of dumb posts on this story, but I think you won the dumb award of the day for this comment.

      Worried about accuracy? Don't, just do your own test following these easy steps that are normally done by trained lab technicians who still don't produce perfect results!

    • by AmiMoJo ( 196126 )

      This test was requested by NASA because he wanted to attend a SpaceX launch. Now he can't go, and getting his own test might not be acceptable to NASA. They probably have rules and specific tests/contractors that they have validated.

    • So funny. Ironically it's marked interesting.
  • by Zorpheus ( 857617 ) on Saturday November 14, 2020 @12:32AM (#60722710)
    This must be the study which was enough to get this test approved: https://jcm.asm.org/content/ea... [asm.org]

    It has a quite high false negative rate several days after the start of infection, when the virus load is low. Then the PCR test detects a lot more positive results than this. That should not be an issue though, since people are not really infectious with such a low load.

    The number of positive samples in the study is really small though.
  • Must is obviously an extremely smart guy, but some of the things he says are just bizarre. He seems to be prone to believe in conspiracy theories - have to wonder if it's related to his mental health issues (he's suggested he might be bipolar).

    As an engineer/scientist why would you take 4 tests using the EXACT same machine/nurse/test when you obviously regard some component of this setup as faulty?!! OK, so he's finally gone for a different test (PCR vs antibody), but his conspiratorial view of the first 4

  • by JoeRobe ( 207552 ) on Saturday November 14, 2020 @09:42AM (#60723654) Homepage

    When taking my first Covid test I asked the doctor about the antigen test (which is what I was administered). Her explanation is similar to what I read here:

    https://www.nature.com/article... [nature.com]

    https://www.aafp.org/family-ph... [aafp.org]

    My understanding is that the antigen test has a high false negative rate because it isn't as sensitive. However, it is specific so it has a low false positive rate. That is, if it comes back positive, there is high probability that you are infected. If it comes back negative then it's a tossup.

    If that's correct, and Musk has had two positives, it sounds very probable he has Covid. The negatives would then suggest that his antigen levels are near the sensitivity limit of the test.

    Since he's rich he should just take the test hourly and build up some nice time-dependent statistics for us all.

  • What is with people assuming something sinister, conspiratorial, or otherwise suspicious when reality doesn't work out the way they anticipated. Thankfully, scientists don't proceed that way, and Elon is no scientist.

    The rapid tests are notably unreliable, and that is why many states that have travel restrictions require quarantine or a negative PCR test, specifically, to travel there (I had to do this when I traveled to Vermont in September).

  • We used to joke about people posting cat videos and what they had for lunch.

    Now we gotta hear about Famous Guy's every sniffle?

    Or is this supposed to be schadenfreude?

  • by JoeRobe ( 207552 ) on Saturday November 14, 2020 @11:38AM (#60723856) Homepage

    I get that the reflex is to discard the antigen test due to its potential for it, say, a 50% false negative rate. But the rapid (antigen) tests have a low false positive rate.

    So the way I see it, if you have a test that is 50% false negative and 1% false positive, but you can administer the test frequently and on a massive scale, then it's still incredibly useful. If you test 1 million people with it, such a test will let you identify and isolate half of the positive cases. That massively reduces the potential for spread. If you do that every few days, it even more useful.

    The caveat there is that a 1% false positive rate means that of the million people, you will be unnecessarily isolating 10,000 every few days.

To be awake is to be alive. -- Henry David Thoreau, in "Walden"

Working...