Munich incidence data glitch

style="float: right; margin-bottom: 10px; font-weight: 600;"Tue 13th Apr, 2021

It is a value that currently guides public life: the seven-day incidence. But time and again there is confusion about the value, which indicates the number of new infections within one week per 100,000 inhabitants. For example, the Bavarian State Office of Public Health (LGL) sometimes reports a different value than the Robert Koch Institute (RKI). This was the case only this Saturday.

But how can this happen? The reporting pathway for a new infection is as follows: Physicians or other persons who are obligated to report Corona cases (for example, directors of nursing homes or schools) or laboratories report confirmed cases to the Munich Health Department. The latter then reports the cases to the LGL, which in turn transmits them to the RKI - electronically, mind you, not by fax. "The reporting channel has been electronic since 2001," says RKI spokeswoman Susanne Glasmacher to AZ.

However, this transfer between the LGL and the RKI is subject to occasional hiccups. "A discrepancy cannot be avoided," says spokeswoman Glasmacher. For one thing, the time periods considered differ. The RKI uses the data status of 0 o'clock in each case. The LGL, on the other hand, has the cutoff time of 8 a.m. The LGL figures are always published in the afternoon at around 2 pm.

However, the differences can also have technical reasons. For example, when servers go on strike. "While in principle it can not be completely ruled out that a correct transmission of the data is prevented for technical reasons," says a spokesman for the LGL in response to an AZ query. If there is such a problem, the LGL adds a corresponding note on its website.

Robert Koch Institute: 270 cases were subsequently reported
The RKI always sorts reports for the respective day on which they were recorded at the health office. So if a case is registered at the health office on Monday, it will also end up at the RKI in the figures for April 12 - but in the case of a transmission breakdown, perhaps not on the following day, but only one or two days later.

On the RKI's Covid 19 dashboard, which provides a daily overview of cases in Germany, these so-called late reports are highlighted in orange. For example, on Friday, April 9, only 184 cases from Munich were initially transmitted to the RKI, with a further 120 cases being added later by follow-up notification. On Saturday, April 10, initially no figures at all were transmitted to the RKI, only on Monday 270 cases were subsequently reported.

Big difference on Saturday Munich incidence
However, this leads to the RKI initially underestimating the seven-day incidence in its calculation - because it initially calculated zero new cases for Saturday, for example. The LGL, in turn, also calculates its incidence with the numbers that have not yet made it to the RKI on a given day. And these were considerable on Saturday - hence the large difference in incidence (94.4/116.4).

So what had gone wrong on Saturday? According to LGL, there was no data transmission problem over the weekend. "In the specific case addressed, all daily data received by the LGL by the required transmission time on the afternoon of April 10 were transmitted to the RKI as usual," an LGL spokesperson said.

Was the city too slow in transmitting?
Data that the local health offices subsequently send to the LGL are not sent to the RKI until the following day. "Thus, in the present case, the transmission of the case numbers subsequently reported by the state capital by the LGL to the RKI took place on the following Sunday morning." In other words, according to the LGL, the Munich health department did not transmit the data quickly enough for it to be forwarded to the RKI on the same day.

Write a comment ...
Post comment