In this brief series of three posts, we will look at the data of the first phase of the SARS-COV-2 pandemic reported in China and Italy.

We use graphic representation from the original sources with some annotations from one of our correspondents.

Share

They are as good as any, but there are some real problems with the data, which we will explain in this and the following post to then finish with a list of things (i.e. research) the Inquiry should spend some of its gigantic budget on.

Let’s leave the problems to one side and look at the official data from China from 8 Dec 2019 to 11 February 2020. 

These data are from Wu et al. (JAMA).

Wu et al. diligently annotated some of the key events in dialogue boxes. 

Look at the number of symptomatic cases by date of onset - the green histograms, compared with the date of diagnosis, the peak and the date of onset of the lockdowns. By the 24th of January 2020, Wuhan and 15 other cities were in lockdown.

Let’s now go to Italy - this time, the observation period is slightly later. The dark blue histograms represent the date of symptom onset, and the vertical red line is the national lockdown (10 March 2020). 

There is lots of other information in this infographic by the Istituto Superiore di Sanità (a sort of Italian version of the NIH), but focus on the blue bars, what are they telling you?

Followers of our John Snow series will recognise the pattern. This is William Farr’s law of epidemics in action. The initial period of exponential increase of cases peaked in a few days and then started plummeting to very low levels.

Remember that symptom onset precedes infection by a few days and succeeds diagnosis and reporting by weeks, so symptom onset is the closest approximation to infection date. 

Now, what do you see?

The Italian data shows the four phases from peak infections, the start of symptoms, to peak diagnosis and then peak reporting of PCR test results, which comes 3 to 4 weeks after the peak of infections.  

This lag between peak infections and peak reporting exceeds the short window when viral circulation is on the rise.  

The peak infection to reporting lag creates a false assumption that any restrictions work. By the time governments had reacted, the viral circulation had peaked and was already on a downward trajectory. 

However, the data tell us more: the deaths continued for months. For example, according to WHO, Italy recorded 139 deaths on 29th June 2020. On the same day, there were 1,289 cases confirmed, a weekly decrease of 32%

So, there was minimal viral circulation, but people kept dying. Why?