FDA statistics scandal update

The other day we reported on the director of the FDA who got embarrassed after garbling some statistics at a news conference. At the time, I wrote:

The commissioner of the FDA might well too busy to be carefully reading the individual studies. I assume the fault is with whatever assistant prepared the numbers for him.

Paul Alper shares this update:

Two senior public relations experts advising the Food and Drug Administration have been fired from their positions after President Trump and the head of the F.D.A. exaggerated the proven benefits of a blood plasma treatment for Covid-19.

On Friday, the F.D.A. commissioner, Dr. Stephen M. Hahn, removed Emily Miller as the agency’s chief spokeswoman. The White House had installed her in the post just 11 days earlier. Ms. Miller had previously worked in communications for the re-election campaign of Senator Ted Cruz and as a journalist for the conservative cable network One America News.

Calling One America News “conservative” is an extreme understatement, or a misclassification, depending on how you think about it.

Also this:

Ms. Miller’s termination comes one day after the F.D.A.’s parent agency, the Department of Health and Human Services, terminated the contract of another public relations consultant, Wayne L. Pines, who had advised Dr. Hahn to apologize for misleading comments about the benefits of blood plasma for Covid-19.

The Department of Health and Human Services denied that Mr. Pines’s contract was terminated because of his involvement in the plasma messaging.

It was “100 percent coincidence,” said Brian Harrison, the department’s chief of staff. “H.H.S. has been reviewing and canceling similar contracts, so I had it sent to our lawyers, who recommended termination. This was routine.”

“Senior public relations experts” are fine, but at some point somebody’s gonna check the numbers. Otherwise the FDA’s as bad as Harvard.

71 thoughts on “FDA statistics scandal update

  1. Having worked in the government for a long time, and since I don’t want to get in a political battle, let’s just say some administrations put highly qualified people in top positions and provide pretty direct access to the scientific knowledge in the agency, while others have a combination of either unqualified people at the top or the installation of a bunch of political hacks between the person at the top and the scientific knowledge of the agency, or both, making certain that only the “correct” message gets out. I would say if you could ask me in person, I doubt you would be surprised at which administrations were some of the worse.

    As for the present administration, just look at sharpiegate and draw your own conclusions.

    • From the beginning we knew vitamin C deficiency was probably an issue. By mid March we knew smoking tobacco was protective for some reason, by mid April we knew HBOT corrected the hypoxemia.

      I think every single biomed researcher who didn’t advocate for following up on these leads should be fired. They were all as obvious as anything could ever be.

        • By “horrific” do you mean “bad”? or “very bad” or “really really bad.” ? Or “horrible” or “horrifying”, “horrid”, “horrendous”, or just “scary”. I’m not criticizing your First Amendment Right, I simply wonder what people are trying to communicate when they use this cliched expression. Like it is isn’t enough to be bad, it has to be “horrifying.’ I’m curious about this apparent desire to tell readers how to emotionally respond to whatever the topic happens to be.

        • I personally suffered from multiple years of chronic sinusitus until I found a way to get it under control. It was very very bad. We’ve done 500k tests a day for months on end. Basically tens of millions of people have had a stick rammed into their sinuses. Only on the order of 5% of them have been positive. So basically tens of millions of people have been subject to potential injury that could lead to infection or chronic problems for zero reason other than “Thats the way we’ve always done it”. If that isn’t horrifying to you then probably you haven’t experienced serious adverse effects of sinus problems.

        • But since when has it been obvious that saliva works as well?

          Sometimes it’s just so much easier to critique a particular protocol with the benefit of hindsight. From what I remember in the early days of covid I never heard of this particular critique.

          And assuming that the study data for this came up in the last few months, we all know how many new studies turn out to be utter crap.

          I just think every error in hindsight will per force have an I told you so camp criticising it.

        • “Since when has it been obvious that saliva works as well?”

          “Saliva is more sensitive for SARS-CoV-2 detection in COVID-19 patients than nasopharyngeal swabs”, by A.L.Wyllie et al, of Yale, preprint April 22. (There may be earlier studies.) I remember a Boston hospital getting FDA authorization for saliva testing at around that time, but can’t find the evidence of that. Timeline-wise, the benefit was established too late to be useful for the big first wave in the Northeast, but there seem to be quite a few places offering this type of testing now.

        • This is the medical image that shows the swab procedure:

          https://www.nydailynews.com/resizer/PCDI4x6ePRT0_2tk3AvkTLh6yN8=/800×599/top/arc-anglerfish-arc2-prod-tronc.s3.amazonaws.com/public/KZRRVZFMUVHYZKHGHMTJCWASGA.jpg

          They are doing this on both nostrils. Technique varies widely. I promise you if you do this to 10M people you will create 10s of thousands to hundreds of thousands of injuries, some leading to chronic illness. all for no reason because spitting in a tube is actually better.

        • Michael Mina of Harvard has been advocating for a rapid test for home use. It relies on a saliva sample.

          https://www.nytimes.com/2020/07/03/opinion/coronavirus-tests.html

          All of the approved so-called rapid tests require a visit to a care facility. You can’t administer the test to yourself. The obstacle to getting a home test is that Mina’s proposed test has less sensitivity than PCR tests.

          However, the purpose of Mina’s rapid test is not diagnosis for treatment. It’s population surveillance for mitigation of the risk of community spread. If you take the test frequently when you’re asymptomatic, you catch a lot of potential infection before it heads out the door, a lot more than what we catch now with clinical-grade tests that are less frequent, not timely in the results, and expensive. The focus of the FDA is on individual care. They haven’t wrapped heads around the motivation of public health and the benefit of population surveillance.

          Imagine if you could test yourself at home every day for a buck a test and get the results back in 15 minutes. You take a picture of your negative result. You present that for entry to your office, your gym, a dine-in restaurant, etc. Still practice control measures like reduced occupation, social distancing, masking, etc. Meanwhile, surveillance has detected and averted a meaningful number of potential infections.

          Inexpensive, home-use tests would be, as they say, a game changer. If you feel the same, I prepared an advocacy kit with messages easily copied and pasted to your representatives.

          https://datascience.jeffryes.net/post/rapid-test-advocacy-kit/

          Jai

        • The obstacle to getting a home test is that Mina’s proposed test has less sensitivity than PCR tests.

          The current tests have like 70% false negative rate, it is worse than that?

          Imagine if you could test yourself at home every day for a buck a test and get the results back in 15 minutes. You take a picture of your negative result. You present that for entry to your office, your gym, a dine-in restaurant, etc. Still practice control measures like reduced occupation, social distancing, masking, etc. Meanwhile, surveillance has detected and averted a meaningful number of potential infections.

          What is the point if you have an insensitive test no one trusts, so you act the same regardless of the result?

          Also, now its being claimed 90% of the positive pcr results were effectively false positives (viral load so low transmission and illness is unlikely):

          In three sets of testing data that include cycle thresholds, compiled by officials in Massachusetts, New York and Nevada, up to 90 percent of people testing positive carried barely any virus, a review by The Times found.

          https://www.nytimes.com/2020/08/29/health/coronavirus-testing.html

          From the beginning Ive been saying this… my guess was 70% false negatives and 70% false positives.

        • > What is the point if you have an insensitive test no one trusts, so you act the same regardless of the result?

          Anoneuoid, Mina addresses your question in the article I linked, which I recommend.

        • > Also, now its being claimed 90% of the positive pcr results were effectively false positives (viral load so low transmission and illness is unlikely.

          The “less accurate” tests might be better described as less sensitive. You are less likely to get a positive result for having levels of the virus where you aren’t infectious. There is a window very early in when you might be at a level that is infectious but the test isn’t sensitive enough to kick that up. But repeated tests will compensate for thsf problem. Alsoz repeated tests will minor the impact of false negatives. With each successive test you reduce the likelihood of missing an actual infection.

          Sure, some people might ignore a positive test and behave just as they would otherwise. But others certainly won’t. Nothing is perfect. And companies and schools can require rapid tests for people who wish to enter their facilities. And families can pool their tests.

          I certainly would love the option to conduct a rapid test at home, for myself, for people who wish to visit, for people who wish to have xk tact with my 90 year old mother-in-law, etc. There would be logistical obstaclesesz but I think it’s insane that we aren’t moving on this. It could well advance the cause of limiting spread and opening the economy.

          The interesting question is why it isn’t happening. I have yet to see a really solid counter-argument yet. The closest I’ve seen is that there would be logistical difficulties. We’ll, yeah, nothing is perfect. What else is new?

        • Google “medcram how to fix covid testing Mina” for a good overview. Or “YouTube ucsf grand rounds new approaches to covid 19 Michael Mina or” Atlantic article Michael Mina.”

        • Right now at Northeastern U. They are testing studies and faculty and staff on a massive scale. Students will be tested every few days after an initial round of three tests in their first week. They are employing a dedicated setup to get a rapid turn around on PCR tests.

          Google “Northeastern U dashboard”

          Thus far they have had what seems like good results (very low positive rate). It seems like a viable option for what it’s worth. With those kinds of resources applied, a rapid turn-around PCR test can certainly be effective.

          But it isn’t as yet practical on a more massive scale. Way too expensive and it requires a whole set of dedicated resources. Such a policy is more in the “individual health” domain than the “public health” domain. It isn’t really a public health option that targets contagen control on a massive scale. For that purpose the rapid antigen testing seems a much better option.

          Mina speaks to the whole “individual health policies ” versus “public health policies” juxtaposition. Very interesting imo.

        • I certainly would love the option to conduct a rapid test at home, for myself, for people who wish to visit, for people who wish to have xk tact with my 90 year old mother-in-law, etc. There would be logistical obstaclesesz but I think it’s insane that we aren’t moving on this. It could well advance the cause of limiting spread and opening the economy.

          Why would you trust your mother-in-law’s life to a test that misses more than 70% of the cases? And it isn’t like these false negatives are independent of each other so repeated testing doesnt solve that problem.

        • > Why would you trust your mother-in-law’s life

          I’m not “trustingc her life. I’m doing the best I c an to make informed decisions about the risks. “Trust” has nothing to do with it. I find that to be a rather common misconception these days where people foist “trust” on me where it doesn’t exist.

          > And it isn’t like these false negatives are independent of each other so repeated testing doesnt solve that problem.

          I don’t know anything about statistics, but it seems to me that with each iteration of a test the chances of an infection passing through undetected becomes exceedingly small. At $1 and 15 minutes a pop, I can realistically repeat basically as many as I want until I am comfortable with the probabilities.

          We are all evaluating risks and probabilities all the time. At least with this I have some small measure of control over the probabilities. In fact, I can lower the probabilities far below the results of a single PCR test or even multiple PCR tests – particularly since there is a long turn around time for most PCR tests if have any access to.

          I can’t choose to realistically hermetically seal off my mother-in-law, so I can have a choice to have information when I make decisions rather than just flying blind. Perhaps you prefer total ignorance to partial knowledge?

          And your categorization of “not independent” seems dubious to me. I see no reason, other than if someone falls into the relatively short window of time when someone would be infectious but not be captured with multiple tests 15 minutes apart, that the tests wouldn’t effectively be independent. I could even choose tests from different manufacturers for repeated tests if I were concerned about a participle flaw in a particular test that affected a particular person’s test.

          But again, it’s not realistic to think thstbehree is a perfect solution.

          In that circumstance I’m far less concerned about a false positive. And the PCR tests cause positive results with people that aren’t infectious g in that sense they are effectively “false positives.”

        • I’m not “trustingc her life. I’m doing the best I c an to make informed decisions about the risks. “Trust” has nothing to do with it. I find that to be a rather common misconception these days where people foist “trust” on me where it doesn’t exist.

          So are you going to screen people from seeing her based on the results of these tests?

          I don’t know anything about statistics, but it seems to me that with each iteration of a test the chances of an infection passing through undetected becomes exceedingly small. At $1 and 15 minutes a pop, I can realistically repeat basically as many as I want until I am comfortable with the probabilities.

          It has nothing to do with statistics. You can repeat as many samples as you want from part of the body without active infection and keep getting the same false negative results.

        • Anoneuoid –

          > So are you going to screen people from seeing her based on the results of these tests?

          Once again, it has nothing to do with “trust.” I notice that you didn’t correct for your misconceptiom and you seem to have not taken in what I said.

          Once again, I would use the tests to evaluate probabilities. They could be a factor in my decision-making. The results of the tests would inform my decision-making. It’s information.

          > You can repeat as many samples as you want from part of the body without active infection and keep getting the same false negative results.

          It still informs about the probabilities. Repeated tests change the probabilities. You seem to be stuck on the notion of “trust.” I wouldn’t “trust” the testing process to be foolproof or disposoroce. You seem to be stuck in a mistaken in a belief that I would. I wouldn’t. Repeating the same test doesn’t eliminate uncertainty but it changes the probabilities. And over a period of time an infection would spread – so repeated tests over two or three days would be testing the infection at different stages. And it’s better than nothing.

        • I’m ignoring the quibbling about the word trust, which doesn’t matter at all. Your plan is that if someone tests negative you are going to let them come into contact with your mother in law, correct?

          The false negatives are not due to a random error, they are from sampling from an area without active virus while it is present elsewhere in the body.

        • Anoneuoid –

          > Your plan is that if someone tests negative you are going to let them come into contact with your mother in law, correct?

          You can call it “quibbling” if you’d like but you’re *STILL* stuck on the concept of me trusting a negative result. That isn’t a principle that I’m operating from. I don’t consider any given test result dispositive.

          At any rate, if you aren’t even going to integrate what I’ve already written, there’s no point in responding any further.

        • Here is the point: you are just wasting your time and money and your mother in law is probably even more likely to get covid if you implement your plan with such a test.

          Your misunderstanding of how the false negatives work and resulting overconfidence is probably very representative of what will happen in general. Your plan is actually a perfect example of a reason that will be used to not give the public access to these tests.

        • Anoneuoid –

          Ok then. Let’s recap. First you misconstrue what I”m saying to mean that I would “trust” the tests.

          When I explain to you that I wouldn’t, but that I would use them as a way to inform me about the probabilities of the decisions I have to make, you continue to insist that I will make a binary choice based on the results of a test.

          When I explain to you that would be the same as “trusting” a test, and that again that’s not what I’d be doing, you insist on your original interpretation and then go on to tell me that I’m being over-confident and that my over-confidence could lead to my mother in law’s death.

          Actually, in my opinion it is you that is over-confident because you’re simply not taking the time to examine your error even though it has been explained to you: I would use these tests to help inform me of the probabilities to help with the types of decisions I would have to be making anyway.

          So I will note your belief – that I wouldn’t have better chances of evaluating the probabilities that somewhat or myself might be *infectious,* (let alone infected) because with repeated tests (perhaps over a couple of days), perhaps even finger-prick tests in addition to swabs or saliva tests, are *likely* to still show negative in someone who is not only infected but also *infectious* because the virus hasn’t gotten into their saliva or nasopharynx, and is just hiding out somewhere else in their body (and that they’re still *likely* to infect someone despite not having the virus in their saliva or nasopharynx).

          And then I’ll put that belief of yours in the same category as the beliefs that some people have in the Gates/vaccine/surveillance state conspiracy theory, and the hospital administrators and doctors and nurses deliberately mis-diagnosing people as covid positive so they can make more money conspiracy theory, and file it away.

          But in the meantime, I think it’s your civic duty to contact Mina to correct his misunderstanding as well, as he’s pushing pretty hard for rapid home antigen tests and he might be gaining some momentum and I fear that tens of thousand might unnecessarily die unless you correct his misconception about how to test for the virus.

        • Anoneuoid –

          Ok then. Let’s recap. First you misconstrue what I”m saying to mean that I would “trust” the tests.

          When I explain to you that I wouldn’t, but that I would use them as a way to inform me about the probabilities of the decisions I have to make, you continue to insist that I will make a binary choice based on the results of a test.

          When I explain to you that would be the same as “trusting” a test, and that again that’s not what I’d be doing, you insist on your original interpretation and then go on to tell me that I’m being over-confident and that my over-confidence could lead to my mother in law’s death.

          Actually, in my opinion it is you that is over-confident because you’re simply not taking the time to examine your error even though it has been explained to you: I would use these tests to help inform me of the probabilities to help with the types of decisions I would have to be making anyway.

          So I will note your belief – that I wouldn’t have better chances of evaluating the probabilities that somewhat or myself might be *infectious,* (let alone infected) because with repeated tests (perhaps over a couple of days), perhaps even finger-prick tests in addition to swabs or saliva tests, are *likely* to still show negative in someone who is not only infected but also *infectious* because the virus hasn’t gotten into their saliva or nasopharynx, and is just hiding out somewhere else in their body (and that they’re still *likely* to infect someone despite not having the virus in their saliva or nasopharynx).

          And then I’ll put that belief of yours in the same category as the beliefs that some people have in the Gates/vaccine/surveillance state conspiracy theory, and the hospital administrators and doctors and nurses deliberately mis-diagnosing people as covid positive so they can make more money conspiracy theory, and file it away.

          But in the meantime, I think it’s your civic duty to contact Mina to correct his misunderstanding as well, as he’s pushing pretty hard for rapid home antigen tests and he might be gaining some momentum and I fear that tens of thousand might unnecessarily die unless you correct his misconception about how to test for the virus.

          Yes or no. Do you plan on using this test with (apparently) 70+% false negative rate to screen who will come into contact with your mother law?

        • Leaving aside the merits of the different techniques, I find it amusing that a few months ago you saw no reason not to deploy the army overnight to ram sticks on the sinuses of hundreds of thousands of random people.

        • First of all, the point was to make testing available using mobile resources, so that people who WANTED to contribute to fighting the pandemic could VOLUNTEER their samples at pop up mobile testing locations. I never once suggested forced testing.

          second of all I had no particular opinion about the method of testing. Saliva tests are obviously better and we’ve known that since around mid april. but in March when I made this suggestion we’d take what we could get.

          it’s one thing to use a swab test because noone has tried anything else. it’s another thing to continue using them on tens of MILLIONS of people even though people have tried out alternatives and found they work better and with less injury

        • Ok, but I wonder how many people would have wanted to volunteer if you started by telling them that hundreds to thousands would get injured, some of them acquiring chronic ilness.

        • I think Carlos has a good point.

          One could even argue that it is fortunate we avoided a additional human-subjects disaster by failing to subject a couple million more people to needlessly brutal tests, given that we now know there are safer and less painful alternatives.

        • Sure, so, why did the “experts at testing” even decide to do NP swabs in the first place? I bet this is left over from back in the day when testing involved culturing bacteria so you needed to get some from deep in the NP swab region or it wasn’t viable.

          It’s a case of TTWWADI as Martha says (that’s the way we’ve always done it). The right thing to do was to start off designing the test based on trying 4 or 5 different sample collection methods: nasal swab, cheek swab, saliva sample, NP swab, throat swab, blotting paper, whatever… and then using the one that had the least intrusiveness and least supply requirements while meeting acceptable sensitivity and specificity. All those things could have been done in a week or so back in mid Jan in China. It’s particularly a failure because the swabs themselves became the hold-up for quite a while as the supplies dwindled.

          Also when you swab 10M people you’re likely to get tens of thousands or hundreds of thousands of injuries. When you swab 100k people in the early days of the pandemic and use that information to sensibly set policy across the country, millions of lives would have been less disrupted and interventions would have been better tolerated due to having data that justified them. And, if you did NP swabs you’d probably injure 100 to 1000 people. There’s a big difference between getting fast actionable information in a week at the cost of 100 to 1000 injuries that results in a smart response and dramatically less disruption, vs getting stupid levels of expensive useless information that comes back 10 days too late to affect behavior at the cost of 100k injuries that we’ve known were unnecessary for months.

        • Sure, so, why did the “experts at testing” even decide to do NP swabs in the first place? I

          Pretty sure it was the same as “early intubation”. Anonymous rumors from China repeated by the WHO.

        • Where has this been shown? Does this degrade the microfilters in any way? They’re made from artificial fabrics that you ought to wash at lower temperatures if in clothing.

          The FDA authorizations for mask sterilization that I’ve seen employ sterilization with hydrogen peroxide via sterilization equipment already used in hospitals for other purposes; if steam would work as well, I’m confident these equipment manufacturers would have received EUAs, too?

          UV does work; store your masks in a sunny spot and they’ll be naturally sterile.

        • It’s DRY heat. I’ll try to find the references there are at least two separate groups that studied this. the electric cooker one was relatively recent but the dry heat sterilization technique was first reported months ago… but it doesn’t require fancy expensive equipment and therefore there is no company pushing an FDA approval.

        • Just as an aside I’m pretty sure you can find two groups to support almost any covid related finding.

          Not to say this particular group is wrong but my point is that policy makers and regulators have a difficult job. There are a lot of contradictory findings out there. Specifically I refer to the swab vs saliva finding.

          A regulator cannot change protocols every month based on the directions set by each successive paper. So it’s of course good to change based on results but it’s a fine line between being responsive to data and changing course way too often.

          So I say be a bit more generous about the regulators. It’s far too easy to judge them harshly with the benefit of hindsight.

        • So I say be a bit more generous about the regulators. It’s far too easy to judge them harshly with the benefit of hindsight.

          I disagree. I see that every decision along the way seems to be for the most expensive and dangerous/disruptive option often with little to no evidence for benefit available. Then the cheap, safe options never get the deserved attention.

          Eg, it took 8 months for someone to publish a report of vitamin C levels in covid patients. When someone finally did it the reported 100% were deficient and 95% basically had scurvy levels (undetectable): https://ccforum.biomedcentral.com/articles/10.1186/s13054-020-03249-y

          It was obvious this should have been investigated from the beginning. So many people have suffered and died unnecessarily due to these poor choices.

        • Here’s research from back in April:

          https://news.stonybrook.edu/sb_medicine/dry-heat-ovens-can-effectively-disinfect-n95-masks/

          here’s some more comprehensive stuff from May:

          https://www.sciencedaily.com/releases/2020/05/200505164638.htm

          Here are some people coming up with the idea of using a simple off the shelf electric cooker to provide the heat:

          https://www.sciencedaily.com/releases/2020/08/200806164654.htm

          IN all cases it’s *dry* heat.

          The point is by purchasing just a few electric cookers, and stacking 100 masks in each one each night… You could turn a couple boxes of N95s from a one day supply into a months supply for a hospital.

          We had pretty clear indication of this back in April. But it’s not a way to make millions of dollars by selling expensive chemical sterilization plants, so it wasn’t pushed through the FDA “medical device” approval process.

          For example around april 27 LA county sheriffs dept bought one of those peroxide sterilization systems. They cost $1000/day to run, the machine was probably extremely expensive… But they could have bought 100 Crock-Pots for $2000 and a daily cost of $1 to run instead.

          Government rent is a powerful source of cash.

      • The smoking observation has been followed up in many places, please don’t suggest it hasn’t. Keep up with the science!

        The final summary from an article with plenty of sources: “In summary and based on the available evidence, it seems unlikely that smoking offers some degree of protection from the effects of COVID-19. The ONS report1 shows that smoking rates are the lowest at 7.9% in those 65 years and over and it is possible that the lower incidence of infection among smokers, simply reflects the fact that fewer older people smoke rather than an intrinsic and protective effect among smokers. Until more definitive evidence emerges, it would seem prudent that current healthcare advice to stop smoking is continued rather than patients relying on a potentially spurious belief that somehow smoking will prevent them from becoming infected with COVID-19.”
        https://hospitalhealthcare.com/covid-19/are-smokers-protected-from-covid-19/

        Age is always a factor with Covid-19, it determines the mortality risk, hospitalisation rate, and probably also the asymptomatic rate: and all of that means that Covid-19 infections are more likely to be detected in older people (and there’s the problem with retirement home outbreaks), but that population smokes less than the average, so if you do an age-agnostic study, it looks like smoking protects when it does not. Don’t trust any statement on Covid-19 that does not consider age demographics!

        • The survivorship bias explanation was the first thing checked and dismissed back in February, I can’t believe people are still trotting that out. Sorry, but that article is just incompetent.

        • Anon, I see that you are a proponent of many ‘alternative’ methods, but those are usually very noisy.

          There is nothing inherent in smoking that would help with COVID. It is probably that smokers are used to living their day to day lives operating at oxygen defficiency and can handle low O2 saturation or whatever COVID throws at them better than the rest of us. Thus, they don’t need to be intubated as often as non-smokers do, which leads to lower mortality, etc.

          I hope you don’t propose there is something special in nicotine or smoke that would help stop the virus. Anyway, what would be some public policy decision, if this is shown to be effective? We all take up smoking?

        • Anon, I see that you are a proponent of many ‘alternative’ methods, but those are usually very noisy.

          I just follow the evidence and do cost benefit analysis. Unfortunately that is considered alternative to people who do NHST, which apparently leads people to regularly choose the most expensive, dangerous, and questionably effective interventions.

          There is nothing inherent in smoking that would help with COVID. It is probably that smokers are used to living their day to day lives operating at oxygen defficiency and can handle low O2 saturation or whatever COVID throws at them better than the rest of us.

          There are probably multiple ways smoking is beneficial, you do not get such a strong effect from only one mechanism. I suspect that adaptations to intermittent low oxygen is one. But smokers are also less likely to test positive for antibodies:

          “Intriguingly, We observed an inverse correlation between induction of an antibody response and
          smoking habit. This may indicate either a lower incidence of SARS-CoV-2 infection in smokers, or
          their inability to induce an immune response to the virus. The interpretation of this result remains to
          be understood as the role of smoking in COVID-19 infection is still controversial, as stated by the
          WHO”

          https://www.medrxiv.org/content/10.1101/2020.05.24.20111245v1

          There seems to be both chronic and acute befits of tobacco smoke on high altitude sickness, which is mimicked by covid:

          To begin with, in severe cases, both COVID-19 and HAPE exhibit a decreased ratio of arterial oxygen partial pressure to fractional inspired oxygen (Pao2:FiO2 ratio) with concomitant hypoxia and tachypnea [4,5]. There also appears to be a tendency for low carbon dioxide levels in COVID-19 as the median partial pressure of carbon dioxide (PaCO2) level was 34 mmHg (inter-quartile range: 30-38; normal range: 35-48) in a recent JAMA article describing 138 hospitalized cases [6]. Initial exposure to hypoxia at high altitude leads to an immediate increase in ventilation that blows off large quantities of carbon dioxide, producing hypocapnia as well [7]. Furthermore, blood gases of non-acclimatized mountaineers with severe illness were accompanied by a significant decrease in arterial oxygen due to an increase in alveolar- arterial oxygen difference, although herein arterial PaCO2 did not change significantly [8]. In short, hypoxia and hypocapnia are seen in both conditions, but there is more.

          Radiologic findings of ground-glass opacities are present in up to 86% of patients with COVID- 19 with 76% having bilateral distribution and 33% peripheral [9]. Notably, lung cavitations, discrete pulmonary nodules, pleural effusions, and lymphadenopathy were absent [10]. In addition to this, patchy infiltrates are present [11]. Patients with HAPE also exhibit patchy infiltrates throughout the pulmonary fields, often in an asymmetric pattern and CT findings reveal increased lung markings and ground glass-like changes as well [12-14]. It has been shown that widespread ground-glass opacities are most commonly a manifestation of hydrostatic pulmonary edema and this is a central point to consider going forward [15].

          https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7096066/

          Cigarettes as an aid to climbing Report, November 21 1922 Captain GJ Finch, who took part in the Mount Everest expedition, speaking at a meeting of the Royal Geographical Society, London, last evening on the equipment for high climbing, testified to the comfort of cigarette smoking at very high altitude. He said that he and two other members of the expedition camped at 25,000ft for over 26 hours and all that time they used no oxygen.

          About half an hour after arrival he noticed in a very marked fashion that unless he kept his mind on the question of breathing, making it a voluntary process instead of an involuntary one, he suffered from lack of air. He had 30 cigarettes with him, and as a measure of desperation he lit one. After deeply inhaling the smoke he and his companions found they could take their mind off the question of breathing altogether … The effect of a cigarette lasted at least three hours, and when the supply of cigarettes was exhausted they had recourse to oxygen, which enabled them to have their first sleep at this great altitude.

          https://www.theguardian.com/books/2007/oct/17/sportandleisure.sport

          AMS, according to the Lake Louise score, was significantly lower in smokers; the value was 14.9%, 95% CI (6.8 to 23.0%) in smokers and 29.4%, 95% CI (23.5 to 35.3%) in non-smokers with an adjusted OR of 0.54, 95% CI (0.31 to 0.97) independent of gender, age and maximum altitude reached. […] Probably because of its influence on the blood’s oxygen transport as well as through its effects on vasoconstriction, smoking is a protective factor for the onset of AMS.

          https://www.ncbi.nlm.nih.gov/pubmed/28947454

          I hope you don’t propose there is something special in nicotine or smoke that would help stop the virus.

          There are plenty of papers proposing such things, don’t know why you have a problem with that. In particular it was reported very early on that ACE2 expression is altered in smokers to be in cells found more commonly high up in the respiratory tract vs nearer the lungs. Others have reported reduced ACE2 expression overall in smokers

          We also found that ACE2 gene is expressed in specific cell types related to smoking history and location. In bronchial epithelium, ACE2 is actively expressed in goblet cells of current smokers and club cells of non-smokers. In alveoli, ACE2 is actively expressed in remodelled AT2 cells of former smokers. This may indicate that 2019-nCov infect respiratory tract through different paths in smokers, former smokers and non-smokers, and this may partially lead to different susceptibility, disease severity and treatment outcome.

          https://www.medrxiv.org/content/10.1101/2020.02.05.20020107v3

          Our study showed that cigarette smoke or direct nicotine inhalation inhibits the expression of angiotensin-converting enzyme 2/AT2R in multiple organs and cell types. In the lung, cigarette smoke (6 cigarettes/d, 12 wk) inhibited the expression of both angiotensin-converting enzyme 2 and AT2R.

          https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5955030/

          The literature presented in this review strongly suggests that nicotine alters the homeostasis of the RAS by upregulating the detrimental angiotensin-converting enzyme (ACE)/angiotensin (ANG)-II/ANG II type 1 receptor axis and downregulating the compensatory ACE2/ANG-(1-7)/Mas receptor axis

          https://www.ncbi.nlm.nih.gov/pubmed/30088946

        • Mods: please delete the other copy of this post if it shows up.

          Anon, I see that you are a proponent of many ‘alternative’ methods, but those are usually very noisy.

          I just follow the evidence and do cost benefit analysis. Unfortunately that is considered alternative to people who do NHST, which apparently leads people to regularly choose the most expensive, dangerous, and questionably effective interventions.

          There is nothing inherent in smoking that would help with COVID. It is probably that smokers are used to living their day to day lives operating at oxygen defficiency and can handle low O2 saturation or whatever COVID throws at them better than the rest of us.

          There are probably multiple ways smoking is beneficial, you do not get such a strong effect from only one mechanism. I suspect that adaptations to intermittent low oxygen is one. But smokers are also less likely to test positive for antibodies:

          “Intriguingly, We observed an inverse correlation between induction of an antibody response and
          smoking habit. This may indicate either a lower incidence of SARS-CoV-2 infection in smokers, or
          their inability to induce an immune response to the virus. The interpretation of this result remains to
          be understood as the role of smoking in COVID-19 infection is still controversial, as stated by the
          WHO”

          https://www.medrxiv.org/content/10.1101/2020.05.24.20111245v1

          There seems to be both chronic and acute befits of tobacco smoke on high altitude sickness, which is mimicked by covid:

          To begin with, in severe cases, both COVID-19 and HAPE exhibit a decreased ratio of arterial oxygen partial pressure to fractional inspired oxygen (Pao2:FiO2 ratio) with concomitant hypoxia and tachypnea [4,5]. There also appears to be a tendency for low carbon dioxide levels in COVID-19 as the median partial pressure of carbon dioxide (PaCO2) level was 34 mmHg (inter-quartile range: 30-38; normal range: 35-48) in a recent JAMA article describing 138 hospitalized cases [6]. Initial exposure to hypoxia at high altitude leads to an immediate increase in ventilation that blows off large quantities of carbon dioxide, producing hypocapnia as well [7]. Furthermore, blood gases of non-acclimatized mountaineers with severe illness were accompanied by a significant decrease in arterial oxygen due to an increase in alveolar- arterial oxygen difference, although herein arterial PaCO2 did not change significantly [8]. In short, hypoxia and hypocapnia are seen in both conditions, but there is more.

          Radiologic findings of ground-glass opacities are present in up to 86% of patients with COVID- 19 with 76% having bilateral distribution and 33% peripheral [9]. Notably, lung cavitations, discrete pulmonary nodules, pleural effusions, and lymphadenopathy were absent [10]. In addition to this, patchy infiltrates are present [11]. Patients with HAPE also exhibit patchy infiltrates throughout the pulmonary fields, often in an asymmetric pattern and CT findings reveal increased lung markings and ground glass-like changes as well [12-14]. It has been shown that widespread ground-glass opacities are most commonly a manifestation of hydrostatic pulmonary edema and this is a central point to consider going forward [15].

          https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7096066/

          Cigarettes as an aid to climbing Report, November 21 1922 Captain GJ Finch, who took part in the Mount Everest expedition, speaking at a meeting of the Royal Geographical Society, London, last evening on the equipment for high climbing, testified to the comfort of cigarette smoking at very high altitude. He said that he and two other members of the expedition camped at 25,000ft for over 26 hours and all that time they used no oxygen.

          About half an hour after arrival he noticed in a very marked fashion that unless he kept his mind on the question of breathing, making it a voluntary process instead of an involuntary one, he suffered from lack of air. He had 30 cigarettes with him, and as a measure of desperation he lit one. After deeply inhaling the smoke he and his companions found they could take their mind off the question of breathing altogether … The effect of a cigarette lasted at least three hours, and when the supply of cigarettes was exhausted they had recourse to oxygen, which enabled them to have their first sleep at this great altitude.

          https://www.theguardian.com/books/2007/oct/17/sportandleisure.sport

          AMS, according to the Lake Louise score, was significantly lower in smokers; the value was 14.9%, 95% CI (6.8 to 23.0%) in smokers and 29.4%, 95% CI (23.5 to 35.3%) in non-smokers with an adjusted OR of 0.54, 95% CI (0.31 to 0.97) independent of gender, age and maximum altitude reached. […] Probably because of its influence on the blood’s oxygen transport as well as through its effects on vasoconstriction, smoking is a protective factor for the onset of AMS.

          https://www.ncbi.nlm.nih.gov/pubmed/28947454

          I hope you don’t propose there is something special in nicotine or smoke that would help stop the virus.

          There are plenty of papers proposing such things, don’t know why you have a problem with that. In particular it was reported very early on that ACE2 expression is altered in smokers to be in cells found more commonly high up in the respiratory tract vs nearer the lungs. Others have reported reduced ACE2 expression overall in smokers

          We also found that ACE2 gene is expressed in specific cell types related to smoking history and location. In bronchial epithelium, ACE2 is actively expressed in goblet cells of current smokers and club cells of non-smokers. In alveoli, ACE2 is actively expressed in remodelled AT2 cells of former smokers. This may indicate that 2019-nCov infect respiratory tract through different paths in smokers, former smokers and non-smokers, and this may partially lead to different susceptibility, disease severity and treatment outcome.

          https://www.medrxiv.org/content/10.1101/2020.02.05.20020107v3

          Our study showed that cigarette smoke or direct nicotine inhalation inhibits the expression of angiotensin-converting enzyme 2/AT2R in multiple organs and cell types. In the lung, cigarette smoke (6 cigarettes/d, 12 wk) inhibited the expression of both angiotensin-converting enzyme 2 and AT2R.

          https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5955030/

          The literature presented in this review strongly suggests that nicotine alters the homeostasis of the RAS by upregulating the detrimental angiotensin-converting enzyme (ACE)/angiotensin (ANG)-II/ANG II type 1 receptor axis and downregulating the compensatory ACE2/ANG-(1-7)/Mas receptor axis

          https://www.ncbi.nlm.nih.gov/pubmed/30088946

        • Exactly, I think the idea that smoking might have a beneficial effect but we shouldn’t study it because we can’t recommend people start smoking is just bass-ackwards science.

          The reason to study COVID + smoking is because it can illuminate the mechanism by which COVID causes severe disease, and then through that illumination we can then design treatments around controlling that mechanism.

          So, suppose the mechanism is related to reduced ACE2 expression… Can we create drugs that alter ACE2 expression short-term? Like I had heard about a human soluble ACE2 drug. The idea would be basically that you give people this drug and it puts the ACE2 soluble form throughout your blood stream, and this basically binds to the virus’ spike protein and blunts that protein…

          Or what if it’s acclimatization to low blood oxygenation. So then we may not be able to acclimatize people, but we can perhaps spend more time working on re-oxygenating the blood: hyperbaric oxygen therapy, anti-oxidants, and vasodilators or something.

          Whatever the mechanisms are, if we understand them, we understand the illness.

          The assumption in medicine that we don’t care about mechanism and it’s all in the “reduced form” give X or placebo and see if X is better… this is killing people.

          We need *science*

        • There seems to be a principle of strategic lying in public health that’s gaining popularity, whereby “officials” worry that facts may be misinterpreted by a hypothetical average joe who doesn’t understand his own utility function as well as they do, so they suppress them entirely. People might rub their grubby hands over their mouths if they wear masks, and the hordes can’t be trusted to not take all the PPE from the hospitals, so “don’t wear masks, there’s no ‘proven’ benefit.” Or on this, as if the hordes of unwashed masses will peruse JAMA and find that smoking may prevent covid fatalities, then spontaneously forget the decades of lung cancer messaging and start lighting up. I think the typical person is much smarter than the college educated class gives them credit for, and while there exist guys that call poison control to check if Donald Trump cured covid by huffing disinfectant, some people taking things the wrong way is just part of saying anything worth saying.

        • As always, thank you for your references. I don’t have a problem with researching smoking more, and I am aware that medical establishment chooses data to fit already approved agenda.

          I know the issues of NHST, but many sources you post use it and I can only go by point estimates and CIs to judge if there’s any effect.

          My point was that smoking is probably only indirectly related to any potential benefit regarding COVID, and even that is noisy. It’s a latent variable of another latent variable, if anything. We have all been ‘smoking’ here in CA for the past 2 weeks or so, due to fires. Are we going to see the benefits? What about high altitude transmission (Denver, Reno, etc.)?

          If not, than it’s not O2 issue, but something in nicotine smoke specifically.

          I believe it is just that smokers are simply used to O2 deprivations, so it might as well be any smoke/polution/O2 deprivation that helps, if one is acclimated. It would be interesting to get some data from the indigenous population in Nepal or South American Andes to see what’s at stake.
          OTOH, there were cases of free divers (who are probably the best adapted to O2 deficit among us) suffering some severe COVID issues in Italy. Being used to severe lack of oxygen didn’t help, it seems.

          As I suspected, it is too noisy and not enough research is done on those who recovered, or even identifying them via constant mass antibody testing, instead of nose swabs. Looks like smoking status is just another variable that’s collected and in desperation to cling to anything, we start seeing patterns. It has to be properly controlled and studied.

        • Looks like smoking status is just another variable that’s collected and in desperation to cling to anything, we start seeing patterns

          It is like the most obvious thing ever seen in medical research. You have two very similar viruses in 2003 and 2020 both with a large number of “missing smokers”, but nothing like this in between. It is a huge 50-75% reduction in reported cases, consistently reported across different cultures and decades out of tens of thousands of patients that have been looked at.

          Then smoking also has acute and chronic benefits for high altitude sickness, which is mimicked by covid. This doesn’t happen from “seeing patterns”.

          I think statistics have ruined peoples ability to have any scientific intuition at all.

        • Anyway, what would be some public policy decision, if this is shown to be effective? We all take up smoking?

          The smoking thing has been known since SARS1, way back in 2003, but was suppressed by biased do gooder medical researchers. As Daniel said, this should have been studied to figure out why. I put a lot of the current problem on the people who suppressed and ignored this: https://edition.cnn.com/2003/WORLD/asiapcf/east/04/18/china.sars.smoking/

          But yes, having all adults pick up smoking to reduce the transmission rate by 50-75% would have been a superior strategy to what was done. The 20th century was basically a century of smoking and was the century of greatest human technological and social progress in human history (not because of smoking, but it clearly didn’t stop it).

          Widespread smoking is much less socially and economically dangerous than all these lockdowns and health mandates from people who have no clue what they are doing. And there are additives to make the cigarettes taste bad or trigger a bit of nausea or whatever that could make it more like medicine and less addictive.

        • I don’t agree with any of that stuff about smoking. If smoking is protective, let’s find out why, yes. But I absolutely disagree with encouraging smoking. The cost benefit calculation would undoubtedly come out against it as well.

          still the research is important because it elucidates mechanism, and science is about mechanism.

        • The cost benefit calculation would undoubtedly come out against it as well.

          Yes, lots of people seem to think smoking is worse than covid. But we never had this level of hysteria over smoking.

        • That’s just trolling, you’re better than that.

          Not trolling. If it turned out that everyone smoking for a few months could stop covid transmission, that would be a much less costly thing to do than shut down all these businesses, etc.

          But we don’t even need that anyway, since we know that covid patients suffer from severe vitamin c and oxygen deficiencies that can be quickly corrected with vitamin c and hyberbaric oxygen treatment.

  2. Can someone clarify what the scandal actually is? Just reading the linked post from the other day it seems like the “35 out of 100 deaths would have been prevented” statement is the correct interpretation (assuming all those underlying numbers are right).

    Or is the right statement “out of every 100 deaths of people who got the low-antibody plasma 35 could have been saved had they been given the high-antibody plasma”? Either way, doesn’t seem like something malicious. And that 35% seems pretty significant — so why the language of scandal?

    Or is there a problem with the underlying study itself? If so, it would be great to see a little dissection on this blog. Those are always good reads.

    • Thom:

      Apparently he didn’t say that said 35 out of 100 deaths would’ve been prevented. He said that said 35 out of 100 Covid-19 patients would’ve been saved, which is a different thing entirely.

      Also, to continue from the news story:

      But scientists were taken aback by the way the administration framed this data, which appeared to have been calculated based on a small subgroup of hospitalized Covid-19 patients in a Mayo Clinic study: those who were under 80 years old, not on ventilators and received plasma known to contain high levels of virus-fighting antibodies within three days of diagnosis.

      What’s more, many experts — including a scientist who worked on the Mayo Clinic study — were bewildered about where the statistic came from. The number was not mentioned in the official authorization letter issued by the agency, nor was it in a 17-page memo written by F.D.A. scientists. It was not in an analysis conducted by the Mayo Clinic that has been frequently cited by the administration.

      This is the head of the FDA we’re talking about, not someone doing an advertising pitch. That’s the scandal. The FDA is supposed to be giving us the straight story.

      • I don’t understand what the scandal is. My understanding is that he garbled a stat during a meeting and then came out quickly with an unequivocal statement that he had been wrong and that his critics were correct to hold him to the fire. Isn’t that exactly the type of clearheaded humble approach that we should be celebrating?

        His exact statement:
        “I have been criticized for remarks I made Sunday night about the benefits of convalescent plasma. The criticism is entirely justified. What I should have said better is that the data show a relative risk reduction not an absolute risk reduction”

        • I’m with you, Ian. Epi people love all those funky little ways of rearranging ratio stats and the terminology always trips me up, too.

          Going ballistic over his bungling interpretation of the percentage risk reduction unfortunately distracts attention from the real message that it’s a cherry-picked subgroup statistic from studies that overall are not showing any meaningful benefit. That’s what really needs to be emphasized.

        • Are we losing sight of a pretty important finding though? Like if 35% of any group of people that would otherwise die could be saved that seems like a big deal. And the comparison is between high and low antibody plasma. If the comparison were between high antibody plasma vs no plasma the number would presumably be even greater than 35.

          I think maybe we should focus on the study itself. As scientists let’s be interested in seeing if we can draw that conclusion out of the study. Can we put a confidence/credible interval on that 35%? Maybe it really is a promising treatment. Or what iffy assumptions would have to be made to get 35 and can we test or evaluate those? So many real things to think about.

          In general, I think there’s a really worrying trend and the tone of that nyt article really captures it. Like they seem to just want to confirm their readers’ abhorrence of this gov. And “science” is just another expendable soldier to be hurled at the enemy— who cares if it dies in the attack. I hate Trump but science is precious and I want to see it survive as a human activity with a reputation and credibility that’s in a class by itself…

        • Thom:

          There’s fault all around, but it doesn’t help if the FDA hires propagandists. Also, like it or not but I doubt the government will be responding to reasoned arguments. If a scandal in the newspaper is what is required for even the mildest of reforms, then I think the press is doing its job by reporting it.

          To put it another way: All the things you’re taking about in your first paragraph, those are the things that we’re paying our tax dollars to have the FDA work on. What we’re not supposed to be paying for is press conferences with misleading statistics. I think the government is falling down on the job when they do this sort of thing, and I think the press is doing its job to point it out.

        • Andrew said, “Also, like it or not but I doubt the government will be responding to reasoned arguments.”

          Agreed — but the public deserves reasoned arguments.

        • Ian:

          You don’t think it’s a scandal that the director of the FDA hyped a study, exaggerating a risk ratio based on a small subgroup using a statistic that as not in the original report? It was good for him to apologize and possibly fire the person responsible. He apologized because it was a small scandal.

          The point about “he garbled a stat” is that presumably he wasn’t ad libbing; rather, I assume he garbled it because the garbled stat was in the briefing that was given to him, which means that whoever was in charge of preparing that briefing garbled the stat, which means that the FDA—a huge government agency—has some mixture of incompetents and ideologues in at least one influential position.

        • idk, I have not been following it super closely. To me it sounds like exactly the type of thing you’ve been advocating for a long time.

          1. Scientist makes a statistical mistake.
          2. People call out the statistical mistake.
          3. Scientist doesn’t equivocate, blame the critics, dodge the blame, downplay the issue, appeal to authority, or delay addressing it. He apologized, issued a correction and took steps to prevent something like it happening again.

          On the other hand, I think it is fair to call Miller’s appointment a scandal.

  3. More on Emily Miller, late of the FDA can be found at

    https://www.thedailybeast.com/ex-oann-reporter-emily-miller-fired-as-fda-spokeswoman-after-just-11-days-on-the-job:
    ——
    Her tweets have also shown a lack of understanding of the virus itself, including one in which she falsely claimed that there is “zero scientific study to show there is long term health problems for covid patients.” She has also defended Trump’s use of the term “Kung Flu” and Vice President Mike Pence’s decision not to wear a mask during a visit to the Mayo Clinic.

    On May 30, she flippantly wrote: “Remember coronavirus?” The number of coronavirus deaths in the U.S. had recently passed 100,000. It is now more than 180,000 nationwide.
    ————————–

  4. Apparently they’ve changed their minds on firing Miller? She’s just been moved to another position?

    Good thing this administration is a well-oiled machine, or the reaction to the COVID-19 epidemic might not be going that well and we’d have tens of thousands of deaths.

  5. Emily Miller as seen by Politico

    https://www.politico.com/news/2020/08/28/fda-top-spokesperson-leaves-404422

    “In May, the White House called me ask asked [sic] if I would consider a high level communication role at the FDA,” Miller wrote on her public Facebook page on Wednesday night, as scrutiny mounted on the agency and her role. “I knew God was directing my path, and I had to come back to DC to work.”

    She is the author of “Emily Gets Her Gun: But Obama Wants to Take Yours.”

  6. “…the last to fall into disgrace was the historian [and old party member] Walter Frank … who in 1933 became became director of the newly founded Reichsinstitut des Geschicte des Neuen Deutschlands … In the early 1940s Frank had to cede his position and influence to the notorious Alfred Rosenberg whose Der Mythos des 20 Jahrhunderts certainly shows no aspiration whatsoever to “Scholarship”. Frank clearly was mistrusted for no other reason than that he was not a charlatan.” Hannah Arendt

  7. “[Pines’s termination] was ‘100 percent coincidence,’ said Brian Harrison, the department’s chief of staff. ‘H.H.S. has been reviewing and canceling similar contracts, so I had it sent to our lawyers, who recommended termination. This was routine.'”

    Or, at least, statistically indistinguishable from 100 percent coincidence.

Leave a Reply to rm bloom Cancel reply

Your email address will not be published. Required fields are marked *