How to play ball game seven

Corona: what is the seven-day incidence?

The so-called seven-day incidence shows how many new infections have happened in the last 7 days. In order for the data to be comparable, the number of new infections per 100,000 inhabitants is calculated back. It is an important key figure that corona measures are also based on.

Since the number of infections per 100,000 inhabitants is always considered, large cities can be compared with small rural districts. Or different federal states. A simple example. Let's assume that 20 people were newly infected in one day. In a city with 100,000 inhabitants the incidence is then exactly 20. But if the city has 200,000 inhabitants, then the incidence is only 10. So you can compare the two cities and see where there is “less going on”.

Everything should be said with this. Unfortunately, you will always find different incidence information for the same location.

How do you calculate the seven-day incidence?

First the hint: calculate got to of course you don't do them yourself, you can find them on various websites, for example our overview of Corona in all German circles or of course the RKI's Corona dashboard. But if you had to do the math yourself, then you do it as follows. The explanation deliberately does not assume anything other than the basic arithmetic operations and is deliberately kept simple.

  1. You have to find the numbers of new infections for the last 7 days. You can find them mostly in the press area of ​​cities, districts and federal states. If not: then you have to search. They are also available on many websites (here with us for the Main-Tauber district). Let's say there were 8, 2, 9, 12, 10, 3, and 4 cases.
  2. Now you have to add up the numbers. With the numbers from now, that would add up to 48.
  3. Now we need the population of the city, district or state. If necessary, you can find them on Wikipedia. Let's just assume there are 129,000 inhabitants.
  4. Now you calculate: Sum of just * 100,000 / number of inhabitants, so in the example
    48 * 100,000 / 129,000 - makes about 37.2.
  5. Finished! 37.2 is now our 7-day incidence per 100,000 population

Depending on how precise the numbers of new infections are and how precise the population is, there is a slight deviation from the "correct" value. But it shouldn't be too high. So don't be alarmed if it doesn't "fit" 100%.

Why are there sometimes different incidence values ​​for the same place?

Warning: from now on it gets a lot more complicated. Unfortunately, it is in the nature of things. Because “of course” it is not “that simple” after all.

There are several reasons for the differences: reporting delays and the method of calculation.

Reporting delays occur again and again because (apparently) many new infections are still reported manually. In any case, you will see a different number of cases if you look at the district, state and federal (RKI). Anyone who closely observes the numbers will notice this. The Main-Tauber district described this nicely in a press release: "The health department collects the newly registered cases of infection in the Main-Tauber district seven days a week and communicates them to the state health department. This reporting channel continues to the Robert Koch Institute, for example. A time delay can occur here, which can result in different numbers from different sources such as the District Office, State Health Office and Robert Koch Institute. Since the district health department is at the beginning of the reporting chain, its numbers are always the most up-to-date.". However, you will not necessarily find yourself in the RKI numbers for the corresponding day. The However, RKI numbers are used when identifying risk areas.

What are the differences in the calculation method?

Roughly speaking, you can calculate according to the registration status or registration date. At the Registration status “only” considers how many cases have been added since the previous day. Usually, however, cases from several days before are reported in the course of a day. Let us assume that 10 new cases are reported on a Thursday. Of these, 2 are from last Tuesday, 6 from Wednesday and 2 actually from Thursday. When calculating according to the reporting status, a total of 10 cases are reported for the day.

The calculation according to Reporting date works differently. There the cases are beaten on the day they became known. In the example above, the numbers for Tuesday are increased by two and those for Wednesday by six. So Thursday has only two cases for the time being. These The number can increase in the following days. If, for example, three cases are known on Friday and seven cases on Sunday, Thursday has a total of eleven cases after the report on Sunday (2 Thu, 3 Fri, 7 Sun). Sounds complicated, but it's actually not that much. Somehow the process isn't illogical either. If all cases were reported directly (electronically) on the day of the positive result, then Thursday would have had 11 cases from the start.

Especially when the reports come in relatively slowly, there may be a clear difference in the incidence according to the registration status and that according to the registration date - completely independent of the aforementioned (further) delays in the reporting channel. Usually the value according to the reporting date is lower than that according to the reporting status - but it could also be the other way around.

The calculation using the example of the Main-Tauber district

The whole thing has consequences, shown here using the specific example of the Main-Tauber district (only then do we have all the necessary data):

  • the District health department calculates obviously according to reporting status and gives the incidence calculated today as the value of the following day
  • the State Health Office Baden-Württemberg calculates obviously according to the reporting date and gives the calculated incidence with the exact time (i.e. in principle today's data also for today)
  • the RKI calculates apparently also according to the reporting date and calculates these values ​​“at midnight”. Seen in this way, today's incidence is published the following day, but clearly stated that it belongs to the previous day. To put it more simply: the RKI dashboard always shows the value of yesterday (how should you know today's, it is only developing)
  • Last but not least: in our own evaluations, like the circle, we calculated according to the reporting status, but referred to the results in the afternoon as the “current” incidence. To avoid confusion, we now refer to this as the “forecast for tomorrow” today. Therefore we have a very high forecast quality ;-) In order to make graphics comparable with RKI, we add the value in the graphics to the current date (almost as if they were calculated at midnight).

Still not complicated enough? Well, then it should be explained why the words "obviously" and "likely" are used above. This is due to the fact that, as far as we know, the exact method of calculation is not published (but somewhat vague for the circle). We have therefore determined on the basis of the results which method the calculations are likely to be made. We just calculated all possible types and then looked over several days to see what fits.

Some real dates

If you are still reading now, we will offer you some real data. They are imported "live". Therefore, the difference may not be noticeable on some days. But that probably happens rarely. Here we go:

Those who deal more intensively with the numbers (and only that one!), may wish for a few more details. But they can also be confusing, so “those who do not need to” are warned against an interpretation. Here are the Corona new infections in the Main-Tauber district in the last seven days according to the district's daily press release:

And here the corresponding case numbers of the RKI according to reporting status (the day they were first reported):

As well as the same number of cases from the RKI according to the reporting date (roughly: the day they were confirmed):

We do not automatically take over data from the State Health Office. So there are none here. But it's probably complicated enough as it is.

So let's leave it alone.

Further information about Covid-19 (Corona, SARS-CoV-2), also in the Main-Tauber district, is available on the overview page for coronavirus in Großrinderfeld and Main-Tauber district.
Please follow me on Facebook or Twitter for the latest updates.Or subscribe to the news via Telegram Messenger.
Information on corona mutations in Baden-Württemberg can be found here.