>> The Covid-19 pandemic presented German society with a situation characterized by great uncertainty, especially in the initial phase, which was unique in the post-war period. Science was given a central role in getting to grips with the uncertainties, especially regarding the health consequences of the pandemic - other consequences, such as psychological or economic ones, were talked about much less often.
In keeping with the slogan “Follow the Science” often heard during the pandemic, science should serve as a guiding star. This is based on the naive idea that science, if practiced carefully enough, will provide reliable knowledge on the basis of which politicians can make decisions. This idea is naive in two respects. First, political conclusions are not derived automatically, even when there is reliable evidence. To give an example: Even if it had been clearly shown that children are just as infectious as adults, school closures would not have been the automatic consequence, as a number of other aspects, including the child's psychological well-being, would have had to be taken into account. Secondly, science can only rarely fully fulfill the expectation of producing reliable knowledge - this thesis will be demonstrated below using the question of whether there was excess mortality during the pandemic.
Deaths from any cause are taken into account when calculating excess mortality. This has the advantage of avoiding problems arising from definitions of what constitutes a Covid-19 death, which vary significantly between different countries. To estimate the level of excess mortality, you only need a few variables that can usually be reliably collected: these are the age structure of the population and the age-specific mortality during and before the pandemic. What may be surprising to many is that estimating excess mortality involves a number of subjective decisions that can legitimately be made differently and that have significant influence on the outcome. So these are clearly not errors.
The exact calculation method for excess mortality can only be briefly outlined here; more details can be found in the literature [1, 2]. The central idea is to compare the actual observed number of deaths during the pandemic with the number of deaths that would have been expected during the pandemic if mortality during the pandemic had been the same as in the prepandemic period. Below we would like to name four factors that can be defined differently for good reasons and have a significant influence on the result:
• The choice of the comparison period has a major influence on the result. For example, if only the year 2019 is chosen as the comparison period, in which there were only a few flu deaths and in which mortality was therefore comparatively low, the excess mortality will be higher than if one chooses the years 2016-2019 as the comparison period. However, the choice of the comparison period is a subjective decision for which there are many options beyond the two mentioned.
• The choice of the period for which excess mortality is calculated can also have a significant influence on the result. This can be illustrated with an example: A highly vulnerable man who would have died in February 2021 without the pandemic, but died in the pandemic in May 2020, is an excess death if you calculate the excess mortality for 2020, but not if you calculate it calculated for the years 2020 and 2021.
• Scientists also have leeway when calculating the mortality rate during the prepandemic phase. There are reasons to choose the average of the mortality rates of the pre-pandemic years, but it can also be justified to continue the generally falling trend in the pre-pandemic years instead of the average. The second approach results in significantly higher excess deaths than the first approach.
• There is no standard as to whether irregularly occurring causes of death – such as heat or flu waves – should be excluded from the evaluation or not. For example, excess mortality is higher if heat deaths during the reference period are excluded.
Epidemiological evaluations in general and calculations of excess mortality in particular are not comparable to mathematics tasks in school for which there is a clear solution. An individual evaluation is usually based on subjective determinations that could have been made differently. Incidentally, the Nobel Prize winner Michael Levitt and the Stanford epidemiologist John Ioannidis chose 66 comparison periods in a study on excess mortality instead of just one subjectively chosen comparison period, and an interval of -4.3 to +4 for excess mortality in Germany in the period 2020 and 2021, 6% reported. This may contradict the expectations of the general public, who expect clear statements from scientists. However, science often has to reject this claim to clarity for reasons of honesty. The public, in turn, must endure this lack of clarity instead of accusing researchers who report no or lower excess mortality than others of trivializing the pandemic.
In order not to be misunderstood: there were also mistakes. Many publications on excess mortality did not take population aging into account. Such an approach is unacceptable because the German population is continuously aging and more deaths were expected in 2020 than in previous years simply because of the significant increase in the number of people over 80 years old. <<