Lessons from A Pandemic Anniversary

It's not just what we know, but how we know it

I’m now having “pandemic anniversaries”—anniversaries of tweets written a year ago. Reflecting on these anniversaries also offers a chance to talk a little bit about metaepistemology: the theory of theory of knowledge. Why did we know what we knew? Which in turn explains: why did we miss it?

Information doesn’t become the kind of knowledge that helps one make better decisions without interpretation, which includes understanding its context, which in turn requires understanding institutions and people—and January of 2020 was a stark lesson on how that operates, and also how it fails.

There were three pieces of information which, in combination and properly interpreted in context, told us a lot: the WHO repeating China’s cover-up of human-to-human transmission on January 14th; China locking down all of Wuhan on January 20th and admitting the existence of human-to-human transmission; and a paper in New England Journal of Medicine authored by scientists from China, many from the Chinese Center for Disease Control and Prevention in Beijing, and some from Hong Kong.

The first was the sign of a cover-up. The second was a costly signal that told us what reality was. And the third was vital information that followed a turnabout. However, by themselves they couldn’t all be taken at face value. They couldn’t all be treated as false, or treated as true. In fact, information from the same source had dramatically different interpretations within the space of a week. Navigating all this required a theory of theory of knowledge—and an interpretation of how a variety of institutions and people behave—to learn the right information from each development. 

In combination, though, they made it clear that we were likely going to get hit with a pandemic.

First this, on January 14th:

World Health Organization (WHO) @WHO

Preliminary investigations conducted by the Chinese authorities have found no clear evidence of human-to-human transmission of the novel #coronavirus (2019-nCoV) identified in #Wuhan, #China🇨🇳.

This strongly suggested that there was an ongoing cover-up in China because, at that point, there were many many signs that human-to-human transmission was occurring. Just two days before, news had broken of a positive patient in Thailand, and one day after that, on January 13th,the Ministry of Health in Thailand had informed the world that a traveler from Wuhan who had never visited the Seafood Market where the epidemic originated had tested positive. (The WHO even issued an advisory about it the same day). Taiwanese officials and Hong Kong infectious disease experts—people with direct contacts in the region and SARS experience—were already on the record as saying there had to be sustained human-to-human transmission. It was based on all we were hearing from Wuhan, including the terrible reality that healthcare providers were getting infected from patients. Of course, the WHO cannot manufacture information out of nothing. It depends on member states to cooperate. However, whatever the exact source, an ongoing cover-up and delay in response, especially when facing an exponential process, is a five-alarm fire.

Let’s call this the Principle of “You Can’t Finesse the Steep Part of an Exponential,” after a Dylan H. Morris quote included in a previous article of mine trying to warn about the more transmissible variants.

The second turning point for me was when China abruptly locked down Wuhan, a city of 11  million. 

zeynep tufekci @zeynep

Whoa. Wuhan, a city of 10+ million, is effectively under quarantine due to coronavirus 2019-nCoV (announced below by Chinese governing party paper.) Swift, even drastic action is good for containing epidemics but has to be paired with transparency and trust in government.

China Daily @ChinaDaily

Public #transportation such as bus, subway, ferry and long-distance bus in Wuhan will be temporarily closed since 10am Thursday. All flights and trains departed from #Wuhan will be temporarily cancelled to reduce risk of spread of the new virus, local govt says. #coronavirus https://t.co/gtzIlFszaf

zeynep tufekci @zeynep

So yeah, there are a lot of tensions with China—and some good reasons for it. On the other hand, from the economy to the climate to pandemics, it really is one interconnected system. We desperately need governments that can manage all this, or we’re all going down.

Until then, I’d been warily watching the whole process unfold., But I saw that as a turning point based on the principle that, especially with competent authoritarians in charge, one should pay attention to what people in power do, not just what they say. That’s especially true when such actions are costly—in this case, very costly. Closing down a whole city is not something the Chinese government would do lightly at all. It was a very very costly move, and information to the rest of the world that the pathogen was much more dangerous than we had been led to believe. The shutdown necessarily came with the admission that, indeed, there was sustained human-to-human transmission. For the Chinese government to do this turnabout, and at this scale, meant that things were very, very grave in Wuhan, and at risk of spreading to the rest of China in a similar manner. This is closely related to the idea of signalling theory and costly signals from evolutionary biology and economics—concepts I use to understand digitally-fueled social movements in my book. (Something I’ll also talk about in later newsletters, as it also pertains to the current brouhaha over Gamestop). 

Seeing the Wuhan lock-down was the first time I talked publicly about the pandemic, because it was then incontrovertible that we were risking one. 

The closing also suggested that the government of China now understood the gravity of the situation, which in turn meant that the Chinese government’s incentives had just changed. Now, at least in the beginning, it would likely do whatever it could to not just stop the epidemic from taking over the rest of China, it would try to prevent it from becoming a global pandemic—that it would be blamed for. (At this point, neither outcome had happened yet, and the best outcome for China would be that neither China, nor the rest of the world, succumbed to it.)

Let’s call this the “Principle of Always Pay Attention to Costly Action.” 

Finally, the third and final turning point for me was this paper in the New England Journal of Medicine. Take a look at this thread I posted in response to an article that was published in the NEJM. This was perhaps the first time I felt compelled to step into this space, directly explaining an epidemiological paper to the public, because I felt like we weren’t properly interpreting what we were being told. The paper, though it was published so early in the game, was in fact the nail-in-the-coffin suggesting that we weren’t going to be able to avoid a pandemic unless we acted very, very quickly and dramatically and if we got lucky, too. 

zeynep tufekci @zeynep

Significant #nCoV2019 paper *just* out at @NEJM. Bad news, which we kinda knew, is there is sustained human-to-human transmission. Goodish news: R0 estimate is 2.2, lower than SARS. But won't be easier to contain because 2020 is more connected than 2003.

zeynep tufekci @zeynep

@NEJM The @NEJM paper finds mean incubation to be 5.2 days, 95% under 12 days. Also human-to-human transmission seems to be through *close* contact. We know that most deaths are elderly or people with significant comorbidities. With an urgent enormous effort, this could be contained.

zeynep tufekci @zeynep

@NEJM Yes all this is preliminary, and there is still under-diagnosis going on due to test availability etc. And these %@^$#!s mutate. That said, for all of us, washing hands often and not touching one's face and self-isolating if feverish are the sensible steps—as in any flu season.

zeynep tufekci @zeynep

@NEJM So far, #nCoV2019 has some features more favorable than SARS (no super-spreaders identified yet etc.) BUT @NEJM paper says some have atypical presentation: gastro symptoms or mild clinical presentation. (SARS came with high fever). Translation: we need maybe millions of tests.😬

Why is this paper significant, and why should we believe its accuracy despite being a source from China’s CDC? 

Exactly because of the locking down of Wuhan. 

Before the lockdown, there was great reason to suspect their official statements. Afterward, however, they had finally unleashed their actual scientists to warn us, because as explained above, now their incentives were aligned with preventing the pandemic, and warn us they did. 

Further, the Criterion of Embarrassment, something historians use all the time, greatly increased my confidence in this paper—and everything the Chinese CDC would publish in the next month or so. This is the idea that something that embarrasses or puts the speaker in a difficult position is more likely to be true. From this paper, we learned that most of the cases had never been to the seafood market and that human-to-human transmission had been occurring since December. This was clearly supremely embarrassing for Chinese authorities and also very dangerous. This confirmed the cover-up, and now both China and the world were risking a pandemic. So everything in this paper and all their warnings now had a great deal of paper.

From this paper, we learned that the mean incubation period was about five days (still true). We learned that the elderly were at much greater risk (still true). 

But we learned something else very important: The paper explained that some cases either had very mild or atypical presentations. They did not reliably come with the high-fever of SARS, which, very, very luckily for the world, coincided with the infectious period of SARS. 

In addition, Chinese officials were already telling us that the disease was spreading from patients without symptoms. 

Here is a CNN article from that week. 

China's health minister Ma Xiaowei made a startling statement Sunday about the Wuhan coronavirus: He said people can spread it before they become symptomatic.

Ma didn't explain why he thinks the virus can be spread before someone has symptoms. If the Chinese health minister is right -- and there are those who doubt him -- that means the five confirmed cases in the United States might have been infectious while traveling from Wuhan to Arizona, California, Illinois and Washington state, even if they had no symptoms at the time.

Some in the US wanted to send a team to China to confirm—a reasonable step if at all possible, but the question facing us wasn’t what to do if we had perfect information, but how to interpret the information we had, i.e. metaepistemology.

Fauci said that CDC disease detectives would need to see precisely how Chinese health authorities have gathered their data and how they came to their conclusion.

"To my knowledge, we have not seen the precise minute, granular data and how they collected it," he said. "We need to get to the real bottom line of how they collected their data and see if it's valid."

"The Chinese have good people. I don't want to impugn their capabilities," Fauci added. "But when it's something as important as this, our people who are trained epidemiologists need to go over their data and the best way to do that is go there and see how they're collecting it."

Other experts in the US weren’t just asking for more confirmation, but were downright openly skeptical of the possibility of transmission during the presymptomatic incubation period.

In a fast-spreading, evolving outbreak like this one, information often changes.

Some experts are skeptical because of the lack of data from China.

"I seriously doubt that the Chinese public officials have any data supporting this statement," said Michael Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota. "I know of no evidence in 17 years of working with coronaviruses -- SARS and MERS -- where anyone has been found to be infectious during their incubation period."

(Before the end of the month, from Germany, in a correspondence in the New England Journal of Medicine, we got yet another confirmation of something the Chinese officials had been telling us: that infectiousness period preceded symptom onset.)

I was paying a lot of attention to this question at the time because, while not being a medical professional, it wasn’t hard to realize almost everything hinged on it. For years, I had taught about SARS, not just to teach the sociology of pandemics, but also to teach about how one can think about history and social dynamics in a more comprehensive manner. 

SARS was useful to teach about how structural—reoccurring, with big social dynamics—and how contingent history can be—at the same time. Epidemics and pandemics recurring features of human societies, but which pathogen will take off? Which one will be confined to a small region? Which one will we be able to beat back? Here, the study of history tells us that big, structural dynamics may determine broad currents, but to also pay attention to the specifics upon which things can turn, sometimes unexpectedly. 

I had always emphasized to my students how pandemics were inevitable, eventually, but also that we had been able to contain SARS exactly because it did not spread during the incubation period, and also because high-fever was a reliable symptom—we could point temperature guns at people at airports and find and isolate the sick before they infected further people. If not for that, I’d tell them we’d probably have had a pandemic in 2003. It was a near-miss. A near-miss, interpreted correctly, is a great teacher.

Now, this was round two. The Chinese officials were telling us the truth, because it was now in their interest to do so. Their scientists were doing their best to warn us, because as scientists that is what  they want to do, and they had finally been allowed to share this with us. Everything we needed to know to act was right there in front of us, but it required not just knowledge, but a theory of knowledge to turn it into actionable, timely information.

So here we are.