The Limited Times

Now you can see non-English news...

New daily cases of Covid-19: why so many deviations from day to day?

2020-08-23T10:55:21.480Z


This indicator communicated every evening by the health authorities is sometimes multiplied by 10 from one day to another of the same week. Decr


493 Monday, 2238 Tuesday, 4771 Thursday, 3602 this Saturday ... The number of new cases of Covid-19 in France, announced by the General Directorate of Health (DGS) in a press release and posted on the dashboard of Public Health France (SPF) every evening, has something to think about in view of these huge differences. Suddenly, a “record” number of new cases is announced on certain days, perhaps a little quickly.

Especially since these figures do not correspond to those, also updated every evening, of the SI-DEP file set up by the government on May 13 and precisely intended to "centralize the results of screening examinations". Even if, in the end, they describe the same trend. Decryption.

Two indicators for the same element

The indicator of "confirmed cases", communicated every evening by the health authorities, is the one that the mainstream media (including Le Parisien) communicate in the wake. It lists all the positive tests the results of which have been "fed back" into the system during the last 24 hours. But that doesn't say when the test was performed. Because the waiting time varies according to the laboratories and the periods: from one day most often, it can climb to a few days (generally no more than three).

🗞 Press release | Situation update #Coronavirus of Friday August 21, 2020
30,503 people have died in France
• 84,829 people have returned home
• 4,745 people are hospitalized
• 379 # COVID19 patients in intensive care

- Ministry of Solidarity and Health (@MinSoliSante) August 21, 2020

This indicator of “confirmed cases” is made up largely of the same data as those forming the SI-DEP file. But this is based on the date of collection. For each day, it indicates the number of people tested positive among all those who crossed the door of a laboratory that day. Due to the delays mentioned above, this file is only updated on D + 3. Thus, this Saturday August 22, we learned that 4110 individuals had been positive among all those who were tested on August 19. Those who were tested on August 20 or 21, and already known to be positive, are not counted and will not appear in the data until August 23 or 24.

This SI-DEP file, which the general public does not or little use, is available on the open access database data.gouv.fr as well as on the Géodes platform (Geo data in public health), of Public Health France. . And this indicator at D + 3 seems the most accurate to give a more precise vision of the epidemic since it corresponds to the number of people actually diagnosed each day.

Here is why we sometimes see two different figures passing by the number of new cases with Public Health France as a source (either its dashboard or Geodes). No wonder, since they correspond to different realities.

Weekly average to smooth out the “weekend effect”

There is a statistical pitfall, common to these two ways of counting the daily number of positive cases: the weekend effect. As many laboratories are closed on Saturdays and even more on Sundays, this is reflected in the numbers.

Newsletter - Most of the news

Every morning, the news seen by Le Parisien

I'm registering

Your email address is collected by Le Parisien to enable you to receive our news and commercial offers. Learn more

Thus, on Tuesday August 18 and Wednesday August 19, the SI-DEP file displayed "only" 729 and 675 positive tests. Logical, since it corresponded to the tests - inevitably fewer than during the week - carried out on the previous Saturday and Sunday. At the same time, the DGS announced 2,238 positive cases in 24 hours on Tuesday and 3,776 on Wednesday. The DGS figure is very low on Monday, and to a lesser extent on Tuesday, due to laboratory closures on weekends.

To remedy this bias in the analysis of the figures, the simplest method is to calculate, for each day, the average daily number of new cases over the past week. And this applies to SI-DEP as well as to the “general public” indicator. In its weekly epidemiological bulletin, Public Health France is also used to reasoning at this rate: the figures are given in week 32, week 33, week 34 (from August 17 to August 23), etc.

Strongly increasing positivity rate

We then see, without possible dispute, that the average daily number of new cases of Covid-19 is increasing sharply in France. On August 18 and for the first time, the SI-DEP counter even exceeded 3000. Sign that "the circulation of the virus is accelerating", as recalled by the Directorate General of Health in its press release this Friday evening.

Especially since, as the number of people tested per week increases at a less sustained rate (the 700,000 bar has just been crossed), the positivity rate increases sharply. It went from less than 2 at the end of July to 3.4 from 12 to 18 August.

VIDEO. Jean Castex: "The situation has deteriorated over the past few days"

Source: leparis

All life articles on 2020-08-23

You may like

Trends 24h

Latest

© Communities 2019 - Privacy

The information on this site is from external sources that are not under our control.
The inclusion of any links does not necessarily imply a recommendation or endorse the views expressed within them.