According to worldometers.info Germany has 1/10th of the deaths of countries with comparable number of infections.

Image of current covid-19 statistics as of 2020-03-13 08:50 UTC

Looking at just France and Germany, both have around the same number of cases, but Germany has far less deaths, even counting in the "serious critical".
Such a statistical outlier looks unlikely, which leads me to the two related questions:

  1. Is the number of deaths accurate or were deaths related to covid-19 not reported?
  2. Does Germany use the same definition of "death caused by covid-19" as other comparable countries?
  • 16
    I would speculate that rather than Germany underestimating its number of deaths, it's the other countries underestimating its total number of cases. Probably none of us can answer that now. – gerrit Mar 13 at 8:50
  • 7
    It's generally not very meaningful to compare these statistics directly because they don't use the same methodologies. Different countries have different test coverage and different criteria when a case is judged as "critical". – Philipp Mar 13 at 10:37
  • 2
    In this "current event": Why does this look like a statistical outlier? Does the illness spread like rays of light? Shandong province has more inhabitants and also 'just' '9 dead'? UK, Switzerland, Poland, Sweden: are they that much "more in line" with your expectations? – LаngLаngС Mar 13 at 11:47
  • 1
    Those "few deaths" are the very first ones in Germany, and they were among the first infected in Germany. So two factors that may make the first infections in Germany different play a role: there were a number of cases where covid-19 was imported to Germany that got mostly young and middle-aged people infected: skiers returning from Northern Italy. Within Germany, the big cluster in the West got a "fast-start" at a carnival session. Both activities tend not to attract the high-risk groups with severe co-morbidities and of old age. Contrast that e.g. with France where the cluster in Alsace... – cbeleites unhappy with SX Mar 14 at 1:10
  • 3
    ... and the US apparently had one of their first clusters in a nursing home (?) so right among a high risk group, and they have a comparatively high number of deaths. Conclusion: at the beginning of a local spread, random/accidental factors can change apprearance a lot. And since it may take two weeks or so even for high risk patients to die of this virus, we see this randomness still when the case numbers are maybe 1 or 2 orders of magnitude larger (and thus exhibit less striking randomness) than when those dead were diagnosed. – cbeleites unhappy with SX Mar 14 at 1:42

Browse other questions tagged .