Germany: While Daily Cases Remain Low RKI Monitoring Rising R0

#15,337

COVID-19 - at least compared to influenza - has a relatively long incubation period, often followed by a slow ramp up in symptoms. As a result, people who are infected today may not become sick enough to seek medical care or testing for 3 weeks or more. 

Most of the cases we see reported today are a reflection of infections that were acquired anywhere from 2 to 3 weeks ago, and deaths likely stem from infections that were acquired a month or more  ago. 

This built-in signal delay makes fine-tuning a nation's pandemic response extraordinarily difficult, and daily case counts and deaths potentially misleading.  

Simply put, when we react to today's numbers, we are reacting to events that happened in the 1st week of June.  

Germany, which started out with some of Europe's highest cases counts in early spring, had, by early May - reduced the number of new cases, and the diseases' R0  (pronounced R-naught) - enough to begin relaxing their lockdown and social distancing rules. 

Germany Daily Cases as of Mid-May

One of the key metrics that the Robert Koch Institute (RKI) monitors daily is the R_0 , the yardstick by which disease transmissibility is measured. Essentially, the number of new cases in a susceptible population likely to arise from a single infection.

In the simplest of terms, with an R₀ below 1.0, a virus (as an outbreak) begins to sputter and decelerate. Above 1.0, and the number of new cases increase, giving an outbreak `legs’.

Calculating the  R value in real time is extraordinarily difficult, and subject to both surveillance data errors  and misinterpretation. Notably, relatively small increases in cases, particularly when daily case counts are already low, can overly amplify changes in the  R_0.  

All of which means we have to take these estimates with a large grain of salt.  But they are worth monitoring, as they can sometime provide early signs of changes ahead. 

By early-May, Germany's RKI estimated that social distancing, case isolation, and other NPIs had consistently lowered the nation's  R₀ to under 1.0, and daily cases had dropped by nearly 80% (see chart above). 

But, as stated previously, this was a reflection of what was going on in mid-April. 

Fast forward to June 12th - just 10 days ago - and Germany's numbers looked even better.  Daily cases were routinely under 300, and the  R₀ - while a little higher - remained encouragingly low (n=1.04). 

Starting about 4 days ago, the daily numbers in Germany began to spike again, rising over 2600 cases.  At the same time, RKI estimated the outbreak's  R has jumped from 1.0 on the 18th  to  2.88 on the 21st. 

While a dramatic, and headline grabbing spike, there may be less here than meets the eye. We'll have to wait a week or so to see if this is a temporary blip, or a genuine trend.  

The latest (June 21st) RKI report summarizes the latest outbreaks and explains their impact on their  revised  R estimates.  

Outbreaks

A high 7-day incidence rate was observed in four districts, primarily due to localised outbreaks: The districts of Gütersloh and Warendorf (both North Rhine-Westphalia) and the cities of Magdeburg (Saxony Anhalt) and Neukölln (city district of Berlin).

The increase in the 7-day incidence in the district Guetersloh is due to an outbreak in a meat processing plant. More than one thousand employees tested positive for SARS-CoV-2. The affected plant was temporarily closed at short notice and all employees are being quarantined as well as their household members. In addition, all schools and day-care centres in the district have been closed since 18 th of june until the end of the summer holidays (11 th of August 2020s). The town of Verl, located in the affected district, has enforced a quarantine zone for those areas in which higher numbers of the processing plant's employees are housed. The outbreak in Guetersloh is linked to an outbreak in Warendorf. 

Employees of the meat processing company are residents of neighbouring districts.

In Magdeburg, an outbreak affecting several now closed schools has led to an increase in the 7-day incidence. In Neukölln, a district of Berlin, an outbreak is linked to members of a religious community. So far 85 cases are linked to this outbreak.

Once again there were major outbreaks in retirement- and nursing homes. In the district of Verden residents and nursing staff have been tested positive for SARS-CoV-2.

Further outbreaks were reported among members of religious communities, eg. from Berlin, Hesse and and Mecklenburg-Western Pomerania.

 Estimation of the reproduction number (R)

The presented case numbers do not fully reflect the temporal progression of incident COVID-19-cases, since the time intervals between actual onset of illness and diagnosis, reporting, as well as data transmission to the RKI vary greatly. Therefore, a nowcasting approach is applied to model the true temporal progression of COVID-19 cases according to illness onset. Figure 2 shows the result of this analysis.

The reproduction number, R, is defined as the mean number of people infected by one infected person. R can only be estimated based on statistical analyses such as nowcasting and not directly extracted from the notification system.

The sensitive R-value reported can be estimated by using a 4-day moving average of the number of new cases estimated by nowcasting. This 4-day value reflects the infection situation about one to two weeks ago. This value reacts sensitively to short-term changes in case numbers, such as those caused by individual outbreaks. This can lead to relatively large fluctuations, especially if the total number of new cases is small. The current estimate of the 4-day R-value is 2.88 (95%-prediction interval: 2.16 – 3.73) and is based on electronically notified cases as of 21/06/2020, 12:00 AM.

Similarly, the 7-day R-value is estimated by using a moving 7-day average of the nowcasting curve. This compensates for fluctuations more effectively, as this value represents a slightly later course of infection of about one to a little over two weeks ago.

The 7-day R-value is estimated at 2.03 (95% predictation interval: 1.60 – 2.49) and is based on electronically notified cases as of 21/06/2020, 12:00 AM. In light of the still low daily case numbers, both R-values should be interpreted with caution and in their course

over several days.

Estimates of the reproduction numbers (R-value and 7-day R-value) were between 2 and 3 during the last few days. This is mainly related to local outbreaks which are described above, the outbreak in North Rhine-Westphalia playing a particularly important role in this context . Since case numbers in Germany are generally low, these local outbreaks have a relatively strong influence on the value of the  reproduction number. Further developments need to be monitored closely during the upcoming days, especially in regard to whether case numbers are increasing outside of outbreak contexts.

Risk Assessment by the RKI

General assessment

At the global and the national level, the situation is very dynamic and must be taken seriously. The number of newly reported cases is currently decreasing. The RKI currently assesses the risk to the health of the German population overall as high and as very high for risk groups. This assessment may change at short notice based on new insights.

Infection risk

The risk of infection depends heavily on the regional spread, living conditions and also on individual behaviour.

Disease severity

In most cases, the disease is mild. The probability of progression towards serious disease increases with increasing age and underlying illnesses.

Burden on health system

The burden on the health care system depends on the geographical distribution of cases, health care capacity and initiation of containment measures (isolation, quarantine, physical distancing etc.). The burden is currently low in many regions, but may be high in some locations.