[Goose]

As you probably know, there were and are significant issues with the German reports of COVID-19 deaths.

The German data on excess deaths (at least through April 26) did not show a large number of excess deaths that were not attributed to COVID-19.

https://www.nytimes.com/interactive/2020...eaths.html

In fact, Germany reported more deaths due to COVID-19 (5737) than the number of excess deaths (3000), though because the baseline is noisy, calculating how many deaths were "excess" is not an exact science.  In fact, the excess mortality was pretty low (only 3% above the expected baseline). Nonetheless, the data doesn't show a large mortality increase in March and April that was probably due to COVID-19 but was attributed to other causes.

I don't think this explanation accounts for Germany's low mortality.

BC

Looking at the excess deaths is a good way to find the "real" increase due to COVID-19, because it should cancel out any reporting bias (either way). I have read (as I am sure you have as well) articles discussing the apparent German low mortality. As you point out, excess mortality is a "noisy" metric, complicated by the fact that deaths due to other causes drops during SIP. However, during the time in question, SIP wasn't in place so that issue disappears.

The mortality rates of Germany and the USA are similar. However, the Germans have 32% as many cases per 10K population as does the US. That isn't just due to testing. If one only tests people who are already requesting medical care/think they are sick, you are measuring the problem but not affecting it. It is possible that Germans are just that much better at social distancing compared to citizens of the US. I do believe that is a factor. However, I believe the major factor in having fewer cases was the significant early on attempt at containment. While it was not entirely successful. it did reduce the penetration of the disease into the population. It got people who didn't know they were sick/not yet sick out of circulation. The initial higher testing rate in Germany was driven by that attempt.

In short, my "read" is that Germany made a much better attempt at containment that the US did, and that reduced the level of community spread both initially and going foreward. It also delivered more and better data to the national leaders, although initially things were pretty confused everywhere. The reduced number of cases per 10K people delivered a reduced number of deaths per 10K people. The larger number of tests in the early days was an expected side effect of aggressive tracing.

[Goose]

As you probably know, there were and are significant issues with the German reports of COVID-19 deaths.

The German data on excess deaths (at least through April 26) did not show a large number of excess deaths that were not attributed to COVID-19.

https://www.nytimes.com/interactive/2020...eaths.html

In fact, Germany reported more deaths due to COVID-19 (5737) than the number of excess deaths (3000), though because the baseline is noisy, calculating how many deaths were "excess" is not an exact science.  In fact, the excess mortality was pretty low (only 3% above the expected baseline). Nonetheless, the data doesn't show a large mortality increase in March and April that was probably due to COVID-19 but was attributed to other causes.

I don't think this explanation accounts for Germany's low mortality.

BC

Looking at the excess deaths is a good way to find the "real" increase due to COVID-19, because it should cancel out any reporting bias (either way). I have read (as I am sure you have as well) articles discussing the apparent German low mortality. As you point out, excess mortality is a "noisy" metric, complicated by the fact that deaths due to other causes drops during SIP. However, during the time in question, SIP wasn't in place so that issue disappears.

The mortality rates of Germany and the USA are similar. However, the Germans have 32% as many cases per 10K population as does the US. That isn't just due to testing. If one only tests people who are already requesting medical care/think they are sick, you are measuring the problem but not affecting it. It is possible that Germans are just that much better at social distancing compared to citizens of the US. I do believe that is a factor. However, I believe the major factor in having fewer cases was the significant early on attempt at containment. While it was not entirely successful. it did reduce the penetration of the disease into the population. It got people who didn't know they were sick/not yet sick out of circulation. The initial higher testing rate in Germany was driven by that attempt.

In short, my "read" is that Germany made a much better attempt at containment that the US did, and that reduced the level of community spread both initially and going foreward. It also delivered more and better data to the national leaders, although initially things were pretty confused everywhere. The reduced number of cases per 10K people delivered a reduced number of deaths per 10K people. The larger number of tests in the early days was an expected side effect of aggressive tracing.