I have mentioned Richard Feynman several times already in this series of posts. I consider him to be the ideal of a scientist – iconoclastic, disagreeable, stubborn, determined, curious to a fault.
The last chapter of Feynman’s great book Surely you must be joking, Mr Feynman is titled Cargo Cult Science and in it Feynman outlines what it means to have what he calls scientific integrity.
Scientific integrity has nothing to do with qualifications and degrees and titles. It has the same relation to institutions that a religious spirit has to do with the church. That is to say, there is no necessary relation. In times when institutions fail, you are more likely to find scientific integrity outside of those institutions than inside them just as you were more likely to find the spirit of Christianity outside of the church at various times throughout its history.
What is scientific integrity according to Feynman?
- The first rule of Science Club is: you must not fool yourself. The second rule of Science Club is:…. (you know the drill). Although Feynman doesn’t say it in so many words, I think of this like a duty of care. Most of us have opinions and we throw them around without a second thought. But, if you’re doing science, it means you should have spent some time and energy trying to prove yourself wrong before you share something. It means saying something like: “I believe X is the case. What evidence is available to me to prove it? What evidence would I accept to disprove the statement? Have I even looked for that evidence or tried to produce it myself?”
- Having tried to prove yourself wrong, you should tell other people what you did and let them try and prove you wrong by reproducing your work
- In telling other people, you should also publish all the things that might be wrong with your idea including ambiguities, experiments that failed, facts that could disprove it or cast doubt on it. You shouldn’t just present the facts that back up your conclusion
- As a scientist talking to a layman, you should explain to the layman in terms they can understand
- When building on previous research, you must first reproduce that research. You must verify its truth yourself. You should not blindly trust other scientists
- You must know the limits of your test technique. You must know what your test can say about the world and what it cannot say. Therefore, you will know what you can say about the world and what you cannot say
Other things could be added to this list but I think it provides a nice framework by which to judge whether proper science is taking place.
In this post we are going to use these criteria to judge the science behind the corona event. In fact, we only need to look at the first few weeks to see that (real) science had almost nothing to do with it. Let’s begin.
31st December 2019: WHO office in China picks up info about an apparent cluster of ‘pneumonia of unknown origin cases’ in Wuhan
As part of infectious disease surveillance programs, the WHO and some other organisations listen in on media reports for anything that sounds like it could be a problem. While doing this surveillance, the WHO picked up a report by some local media in Wuhan. The information is sketchy. It seems rumours of SARS were spreading on social media but actual tangible medical information is lacking. The WHO asked Chinese authorities for more information on 1st January but, before even receiving that information, it issues health alerts that day and the next through its international networks.
Whatever this is, it is not science. The WHO apparently propagated a report from a local Wuhan news outlet and social media rumours without even waiting for authorities to confirm it.
Let’s review the facts as told in the official record. There was a cluster of people (about thirty) with “pneumonia or unknown origin”.
The questions to be asked are: is it unusual to have about thirty cases of pneumonia in China in the middle of winter in a city of 11 million people? And is it unusual to have cases of pneumonia where you don’t know what the cause is?
The answer to both these questions is: no.
It is absolutely not unusual to have lots of pneumonia cases in China and in particular in a big Chinese city where air pollution is very high and many people are lifetime smokers. It is also perfectly normal not to know what the cause of a pneumonia is.
The US CDC says: “…clinicians are not always able to find out which germ caused someone to get sick with pneumonia.”
The American Thoracic Society says: “Pneumonia can be caused by lots of different types of microbes, and no single one is responsible for as many as 10% of pneumonia cases. For most pneumonia patients, the microbe causing the infection is never identified.”
Depending on who you ask, it seems that we only know the real cause of a pneumonia in about 15% of cases.
So, what does “pneumonia of unknown origin” even mean? Why would something that seems to be a perfectly normal situation be cause for concern to the Chinese authorities and the WHO?
Turns out the phrase is specific to China and specific to a surveillance program that is run in China.
China implemented a special reporting system following the SARS episode in 2002 and the phrase “pneumonia of unknown origin” is part of that system. The purpose of the system was for doctors to report directly back to Beijing when something seemed wrong. The whole point of this system was to get around the local politics that go on in China and to give Beijing direct access to information on the ground. But the system didn’t work. Local bureaucrats still intervened and controlled the flow of information. In the case of the corona event, some whistle-blowers tried to go around local authorities in Wuhan. One doctor in particular became famous because he was arrested by police for messaging colleagues about what was happening. He died just a few weeks later. It was the actions of these whistle-blowers that caused the social media rumours and the media reports and presumably also got the China CDC involved.
So, we can see that neither the phrase “pneumonia of unknown origin” nor the number of cases on the ground in Wuhan had any scientific importance. What happened was that the WHO asked for clarification of a Chinese language media report that it had intercepted as part of its own surveillance program. In the meantime, the internal Chinese surveillance program had already been activated and the China CDC was on the ground in Wuhan.
Why hadn’t the China CDC notified the WHO directly? How many other times had the China CDC investigated such cases on the ground without the WHO finding out? How many other occasions have there been in China where local authorities prevented the registration of cases of ‘pneumonia of unknown origin’?
We will never know but bear in the mind that the whistle-blowers in this case were arrested by local authorities and forced to sign confessions. So, I think it’s pretty safe to say that there are lots of such events in China that never see the light of day. Let’s be clear about the situation from a scientific point of view.
All around the world every year most cases of pneumonia are of ‘unknown’ type simply because it’s too expensive to test every case. If even rich countries don’t test widely, you can imagine what happens in poorer countries.
We also know that any cold/flu virus can lead to pneumonia and that there are many cold and flu viruses that we simply don’t know about because nobody has bothered to come up with a test for them. Nobody has bothered because it doesn’t matter. The treatment you get is not based on the specific virus that causes the disease.
What happened in Wuhan has everything to do with internal Chinese politics. Corruption is endemic in China and the local population has about zero trust in the authorities there. This was especially true after the first SARS event where the government was seen to have blatantly covered things up. As a result, the average Chinese is already primed to be looking out for respiratory epidemics and assumes the government will try and cover any up. The authorities in Wuhan were seen to be covering up what could be the next SARS and rumours started flying in media and social media.
So, right from the start we can see that it is not science driving this story but internal Chinese politics. For reasons unknown, the WHO propagated that politics. As a peak scientific body in the world, they were seen to be giving scientific credence to a story before they had any official response from the Chinese authorities. If the Chinese authorities had wanted to handle the situation in Wuhan quietly, the WHO had just made that ten times more difficult.
3rd January 2020: The genome of a “novel” coronavirus is sequenced. It is taken from a patient in Wuhan who had pneumonia. (Important: this news wasn’t officially announced until the 7th)
We have to be very clear what the word ‘novel’ means here. It simply means ‘previously unknown’. It means, no human has identified this virus.
But we can be even more specific because in this case the definition was a genomic analysis conducted according to rules set by The International Committee on Taxonomy of Viruses (ICTV). Thus, ‘novel’ means:
Calculated by mathematical analysis to be genetically dissimilar enough to known viral genomes to be considered a new type for taxonomic purposes.
Let’s put that into perspective. Let’s say there was an International Committee for the Taxonomy of Animals. That Committee decides to start using genomic analysis to define the names for animals. Enthusiastic scientists on The Committee put the genome analysis to work and realise there are actually two types of Siberian Tiger. They give them the names: Siberian Tiger 1 and Siberian Tiger 2.
That’s nice for the scientists but if we want to find out whether, for example, species 1 is more dangerous to humans than species 2, we can only prove that by studying the behaviour of members of the group Siberian Tiger 1 and Siberian Tiger 2. The genome analysis cannot tell us that. It can only help us to identify Tiger 1 and Tiger 2.
Same with viruses. The fact that a ‘novel’ virus has been identified by genome analysis doesn’t tell us anything about how dangerous it is. Of course, the virus has been implicated in some pneumonia cases but that’s what respiratory viruses do. It’s not surprising. If Siberian Tiger 1 attacked somebody while Siberian Tiger 2 didn’t, that doesn’t prove anything about whether the species Siberian Tiger 1 is intrinsically more dangerous than Siberian Tiger 2. You don’t extrapolate from an arbitrary sample. You run proper scientific experiments to prove that theory.
This method of classifying viruses by genetic information alone was introduced by the ICTV in 2012. It is controversial even within virology and many virologists have objected to the new system.
For our discussion, it’s enough to know that the ICTV uses familiar terms we all know from high school biology class eg. Family – Genus – Species. At this point in our story, it was hypothesised that the virus was different enough to be called ‘new’. Later, on February 11, the ICTV would officially declare it to be a strain of the SARS-COV species of coronavirus giving the following taxonomy:
|Sub Genus: sarbecovirus
|Strain: 1, 2
Where strain 1 relates to the SARS event from 2002 and strain 2 is the current corona event.
Because the ‘new’ virus was apparently 87.99% similar to a known bat coronavirus there was a further hypothesis constructed that the virus had somehow jumped across from bats. No attempts were made to prove this hypothesis and, in fact, it is not clear how it is even possible to begin to (empirically) prove it. Naturally, it was this story that made it into the media and has become part of the myth of corona.
Let’s be very clear about this from a scientific point of view. All that has happened by this stage in the corona event is that some scientists in China think they have found a virus that is different enough from known viruses to be taxonomically new. At this point there had been no official publication, no official scientific paper outlining their methods, no peer review to check those methods, no other scientists verifying the work, no official ruling from the ICTV that the virus is taxonomically new, no confirmation from the WHO that the disease was new, no proof and not even the attempt at proving that the virus in question causes illness or what specific sort of illness it causes, no empirical studies at all that this virus caused a unique or particularly deadly kind of illness.
Two days later, the WHO would issue a statement saying basically the same thing I have just said although in fewer words.
5th January 2020: the WHO issues a statement
“The symptoms reported among the patients are common to several respiratory diseases, and pneumonia is common in the winter season; however, the occurrence of 44 cases of pneumonia requiring hospitalization clustered in space and time should be handled prudently.”
It would have been nice if things were handled prudently. But prudence, and any semblance of scientific rigor, were about to get thrown out the window because…
1st-16th January 2020: the genome of the ‘novel’ virus is shared with researchers worldwide and PCR tests are created
The Chinese researchers had found a genome and mathematical analysis done on a computer told them it was ‘new’. On the basis of that result, they uploaded that genome to at least two international repositories. One of those repositories is called GISAID. GISAID’s mission in its own words is:
“…overcoming disincentive hurdles and restrictions, which discourage or prevented [sic] sharing of virological data prior to formal publication.”
“Prior to formal publication” means prior to even the first step in the peer review process so it seems the people who started GISAID were explicitly trying to bypass peer review. Sometimes restrictions and hurdles are there for a good reason. Sometimes it is good not share information when you haven’t spent the time to do basic checks of your own work. Like Chesterton’s Gate, if you don’t know the reason a scientific restriction is there, maybe you shouldn’t get rid of it (that’s a real application of The Precautionary Principle by the way). But getting rid of hurdles is apparently the whole point of GISAID.
We saw earlier that Feynman’s number one principle is: don’t fool yourself. In biological sciences all kinds of gates and hurdles and ‘restrictions’ have been put in place over decades for the primary purpose of ensuring that the scientist didn’t fool themselves. Among these are peer review, control tests, blind tests, placebo tests etc. These restrictions are built into science because they lead to good science. It’s no surprise that the corona event should explicitly feature the removal of such restrictions. (Of course, technically speaking, genome analysis is not empirical science, it is mathematics. But that’s a whole other discussion.)
With the genome data uploaded, virologists and other scientists around the world could now use it to do things like create PCR tests. That’s exactly what they did. The German virologist, Christian Drosten, was the first outside China to create such a test which he submitted to the WHO. The Chinese authorities were already using their own test kits early in January.
On the basis of the genomic information and the PCR test, authorities in Thailand and then Japan were able to test for and find ‘infections’ on 13th January and 15th January respectively. The Chinese government locked down Wuhan on 23rd January based on the results of the PCR tests. Other countries started testing and found ‘infections’. The rest, as they say, is history.
Health systems around the world started testing for ‘infections’ before the usual scientific process had got underway. There had been no formal publication showing the virus was the cause of the pneumonia in Wuhan. All they had at that point was correlation based off a very small sample. No peer review had been done and no experimental results had been published. Neither the WHO nor the ICTV had given official recognition to either the virus or the supposed disease.
I have said in earlier posts that the corona event was the ascendancy of the virologists and the public health bureaucrats at the expense of other actors. This is specifically what I mean. The genome and the PCR tests based on that genome were shared before there was any firm science on how dangerous the virus was.
Remember the first rule of Science Club: don’t fool yourself.
If you ramp up testing you are going to find ‘infections’ and it’s going to look for all the world like you have a problem. As a scientist, you must stop, take a deep breath and try and put things into perspective so you don’t fool yourself. But that didn’t happen. Public health bureaucrats started to find infections and they panicked. The Chinese government panicked so much they shut down an entire city of 11 million people.
Two months later at the end of March, when most countries in the west were already going into lockdown, a real scientific paper was published in the New England Journal of Medicine. Buried in it is this warning:
“Care should be taken in interpreting the speed of growth in cases in January, given an increase in the availability and use of testing kits as time has progressed.”
In other words, the more you test, the more you will find. This warning about naïve testing has been repeated many times during the corona event. Let’s take a moment to understand it.
A Brief Interlude: the more you test the more you find
What that warning in the New England Journal of Medicine meant was: if you test more, you will find more cases.
The WHO have stated the same thing in relation to coronaviruses in general: “As surveillance improves more coronaviruses are likely to be identified.”
The reason many people misunderstand this idea is because they have an incorrect understanding of how much we know about viruses. So, let me repeat what I stated earlier:
We don’t know about most of the viruses that cause respiratory illness.
This is literally true. You can check the influenza surveillance program in your country to confirm it. In Australia, we mostly test for influenza but even when we test for other viruses the result comes back negative in the majority of cases. That is, if you show up to hospital with advanced flu symptoms, if they even bother to test you, the test will probably come back negative. That is the normal state of affairs. And that’s just for the infections that get to hospital. Countless more infections will never get tested because the people who have them are asymptomatic or only mildly ill.
We don’t know about most of these viruses because we simply haven’t had the technology to enable us to find them.
Think about it this way. Suppose there is an area of the night sky that nobody has ever looked at before. Maybe there is some religious prohibition about looking at that part of the night sky. We know there will be stars in that part of the night sky and we know that if we started looking we would find stars. If somebody broke the rule and looked and found a star it would be completely obvious and nothing to write home about. Maybe that person would name the star after themselves and we would think they were an egotistical jerk. Maybe they would win an award for finding the star and then everybody else who wants to win an award will start looking and soon we will have found heaps of ‘new stars’. The stars were always there, we just never looked for them before.
Let’s extend the analogy to include the invention of the telescope. With the telescope you can find new things in the sky. You can run new tests. It is fully to be expected that you are going to find new stuff.
This is the exact situation with respiratory viruses. In the last few decades we have built new technology to identify them better so we are going to find more of them. Lots more of them. Get ready because in the years ahead we are going to find lots of ‘new’ viruses.
This eagerness and this ability to find new viruses is behind not just the corona event but all the well known pandemics of the past twenty years. I have already said in previous posts that it was the PCR test which drove the corona event, but more specifically it was the ability to identify the genome and test for it.
7th and 9th January 2020: “Chinese authorities” declare a “novel” coronavirus
On January 7, 2020, the China CDC released a statement declaring a novel coronavirus had been identified by genome analysis and was behind the pneumonia cases in Wuhan. I haven’t been able to find that document, presumably it is in Chinese. But two days later the WHO acknowledged it and went public with almost the same information.
The crucial information is in the first paragraph:
“Chinese authorities have made a preliminary determination of a novel (or new) coronavirus, identified in a hospitalized person with pneumonia in Wuhan. Chinese investigators conducted gene sequencing of the virus, using an isolate from one positive patient sample.”
Do I even need to point out to the reader the problem from a scientific point of view? It’s right there in the phrase “using an isolate from one positive patient sample”. A single sample.
The WHO acknowledges the weakness of this by pointing out that it is a “preliminary determination” but the very fact that this official document was made public is already a tacit acknowledgement of truth. The world’s leading health body is now spreading the story about a “new” coronavirus. In the days that followed, the WHO stepped up its activities including a “meeting on the novel coronavirus outbreak”.
So, we went from a “preliminary determination” to a “novel coronavirus outbreak” within the space of twenty-four hours based on a single patient sample out of group of about thirty people with pneumonia in a large Chinese city in the middle of winter. All this based on a statement by Chinese “authorities”.
When one of the peak scientific bodies in the world speaks, people listen. Just two weeks later, on 23rd January 2020, a team of scientists publishing in a European journal on infectious disease referenced the WHO’s statement and wrote as follows:
“A novel coronavirus currently termed 2019-nCoV was officially announced as the causative agent by Chinese authorities on 7 January.”
Note the language here: “officially announced”? What does “officially announced” have to do with science? And who made the official announcement? Chinese “authorities”? Who are they? Where is their evidence? Where is the scientific paper where these authorities show their work?
And what about the term “causative agent”. That the supposed virus was the “causative agent” had NOT been proven. Nothing had been published to try and prove it. These scientists weren’t referring to actual research. They weren’t referring to a published paper in a scientific journal. They were referring to unknown “authorities”. This is cargo cult science on steroids. It is the opposite of science.
Once upon a time, the “authorities” said the world was flat. The “authorities” said the sun revolved around the earth. The “authorities” said we should burn witches at the stake. History is full of the “authorities” saying all kinds of things. A scientist must have proof. That is what it means to do science. If you don’t have proof, if you haven’t even seen the proof, you are not doing science. What’s more, if the proof has not been provided, it’s your job is to ask for the proof, not parrot the edicts of “authorities”.
The way science works is that, over time, one paper usually becomes the definitive paper that proved a hypothesis. That becomes the paper that everybody refers to when making a claim like such and such a thing was the “causative agent”. Until that happens, you would reference other papers that purport to prove it and you would say something like Smith et al (2020) claim that such and such a thing was the “causative agent” however Johnson et al (2020) called the claim into question. The phrase “officially announced” simply has nothing to do with science.
If even trained scientists were propagating the official announcement just two weeks later, if the WHO was already treating it as a done deal, is it any wonder the public had assumed the science was settled? In truth, the science had not even begun.
On February 11, the WHO officially announced(!) the new disease: “covid-19”. This disease was named according to the WHO’s “best practice” rules for naming which are as follows.
- That is an infection, syndrome, or disease of humans;
- That has never been recognized before in humans;
- That has potential public health impact; and
- Where no disease name is yet established in common usage
It is clear that principles two and four were not fulfilled in this case. Pneumonia and Influenza Like Illness have been recognised in humans before. They are literally two of the most common diseases that affect humans. We have names for them already. We already know that coronaviruses cause these diseases. In short, we didn’t need a new name.
No surprise then that months later doctors and clinicians couldn’t diagnose covid-19 as a specific disease. It is not a specific disease. Most viral disease isn’t. That’s why we have the vague name Influenza Like Illness with a grab bag of associated symptoms that try to cover the possible outcomes of a viral disease.
At this point we can end our review. I’m sure the reader has gotten the point. To give credit, there were many scientists who qualified their language and didn’t simply parrot the position of authorities and there were some scientists who actively spoke out about what was happening. But most just went along with it. I’m sure the scientists, like the rest of us, had no idea what was about to happen and so they kept their mouths shut. Science normally works slowly and speaking out, whether as an academic scientist or a scientist who works in a corporation or research lab, has consequences. Most of us are not free to speak out in our own line of work and scientists are no different.
Even Richard Feynman couldn’t have stopped what happened. From the first, it was not really a matter of science.
Postcript 1: The gold standard test
Everybody knows how you test whether a virus causes illness. It is part of our culture and it is captured in the phrase “to be a Guinea Pig”.
It is the same process stated by microbiologist Robert Koch and is known as Koch’s postulates.
You take a sample from an infected person. You grow the virus in the lab. You expose some animals, probably mice or Guinea Pigs, to the virus and study the response. When they get sick, you take a sample from the animal and confirm as best you can that the virus was what was causing the illness.
It appears nobody has ever run this test for sars-cov-2. I have looked and not found it. I have come across others who also looked and found nothing. I’m sure it is a difficult test to run especially now that viral research labs resemble military facilities in terms of security.
Koch’s postulates are the only test that conclusively proves that a virus causes an illness. In an ideal world, it should have been priority number one right from the start of the corona event. We should have had as many scientists on it as possible so that we could be sure what the illness was and what the severity was.
On 29th January, researchers in Melbourne, Australia became the first outside of China to recreate the virus from a sample taken from an infected patient (as defined by the PCR test). This meant that it was feasible to do the test according to Koch’s postulates. Having grown the virus, it should have been possible to infect animals with it and see what happened. By Feynman’s principle of reproducing other scientist’s work and not just believing what you were told, this should have occurred. Reproducing work is the most basic part of science.
In this case you would think it would have also been a matter of great political importance. Shouldn’t the Australian government, shouldn’t all other governments in the world, have wanted proper research data to guide public policy making? Data provided by their own scientists rather than relying on overseas “authorities”.
But this was not apparently a priority at The Doherty Institute or any other research lab around the world. The priority was to get to work on a vaccine. Research labs are expensive to run and vaccines can make a lot of money. It may be good business, but it is not good science.
You could argue that this is all really silly. We don’t really need to prove that coronaviruses cause respiratory illness. We already know that. We shouldn’t be surprised to find a pandemic for a respiratory virus. We know there are many every year. We shouldn’t be surprised to find a new coronavirus. We know there are many out there that we don’t know about and as we look more we will find them.
Like with my Siberian Tiger example above, the thing we actually care about is: is this virus more dangerous than others that we know about and, if so, how dangerous is it?
That is the domain of epidemiology. But epidemiology takes time. It wasn’t epidemiology driving the corona event. It was “infections” and the fear of viruses. At any one time, each of us will be “infected” with any number of viruses but now we had a test to prove it and with that test the invisible world of microbiology with all its unfathomable complexity was reduced to a single point of fear.
All posts in this series:-