In January of 2020 we began to write about "coronavirus 2019" due to concern regarding increased cases of pneumonia during a mild flu season, while the Chinese dictatorship was denying there was any problem at all. Just over a week later a key whistleblower in Wuhan, Li Wenliang, turned up dead after being arrested and held prisoner for a month by the communist government for "rumor-mongering."

In hindsight, the evidence was there despite China rushing to suppress it, even telling the World Health Organisation to state that the disease could not be transmitted from human to human...or else. A recent analysis reveals that in English, German, French, Italian, Spanish, Polish, and Dutch Twitter messages, people were already talking about pneumonia even as western journalists were calling the American government racist for saying flights from China should be stopped.

The authors note, for example, that Italy had far more mentions of pneumonia during the first few weeks of 2020 despite having a milder flu period than the same period in 2019. The country went into lockdown a month later. In Italy, Spain, and France, the pneumona tweets were in the regions where the first infections were later officially reported.A similar spike emerged from the data with the term "dry cough".

CLICK IMAGE FOR FULL SIZE. Anomalous evolution of pneumonia-related tweets posted across Europe since December 2019. (a) Cumulative rescaled number of tweets citing pneumonia from 10 December 2019 to 1 March 2020. Inset plot shows the evolution of such tweets posted in Italy from 1 July 2019 to 1 March 2020, and highlights the two winter seasons (shaded bars) used to uncover anomalous spikes of pneumonia-related tweets. (b) Two-sample Kolmogorov–Smirnov test of the difference between cumulative distributions of number of tweets citing pneumonia and posted in the two corresponding winter seasonal periods (2018–2019 and 2019–2020) for each of the 7 European countries. The graph reports the average p values over moving window widths w ϵ [50, 70] computed with daily frequency.

That tells us social media can be used for more than canceling people and expressing outrage if someone says astronomy is science and it doesn't matter what your gender is doing it.

But how to use it for epidemiology is tricky. Epidemiology is statistics so can easily be used inaccurately or even malevolently, as anyone who's read a Harvard School of Public Health food frequency questionnaire screed on miracle vegetables or their scary chemical of the week can attest. 

The authors say that a global surveillance system managed by the UN would help but when has that ever helped? The International Agency for Research on Cancer, IARC, is run by the UN and they did more to damage trust in science prior to this pandemic than almost any group.  They put paid environmental activists on panels to "investigate" chemicals their lawyers were suing companies about. Their botched methodology means they think bacon gives you the same chance of cancer as plutonium and mustard gas.

The World Health Organisation will repeat talking points from China the next time China tells them they must, so the same organization that told America trying to eradicate smallpox was a waste of time are also ill-equipped to do anything but create colorful marketing brochures about starvation.

No, centralized bureaucracy is not the solution for epidemiology any more than it is for anything else. Independent investigation like this, which shows value, will lead to the private sector creating better analysis tools specifically optimized for these kinds of events. And there will be more. This was the third coronavirus pandemic of the last 17 years. They are the new normal.