What are some nerve-wracking statements

Analysis - How many are there today? Why the daily BAG figures say little - but we still need them

How many are there today? Why the daily BAG figures say little - but we still need them

The problem with the case and other numbers. And why we cannot do without statistics.

Today the BAG reports 234 new Covid-19 cases. These are the positive tests that are reported to the office.

It is often criticized that this number does not say much. I guess that's true. But it certainly concerns something we would like to know (more on that below). And it is an absolute figure and a concrete number. How many have the virus? We also add up the numbers. It is of course not completely in the air. Although certain relational information is more meaningful.

And now the relativizations:

  1. Topicality: Because the reports can be delayed, the figures do not always relate to the last 24 hours. The cantons sometimes have different numbers.
  2. Significance: Critics complain that there are people among them who show no or only very mild symptoms. So it is not actually about sick people. But since people can pass the virus on without symptoms, the number is not meaningless.
  3. Development: If you want to monitor the development of the pandemic, the proportion of those who tested positive in the total number of tests would be more informative. This number bobbed around 1 percent for a long time, but has risen again to 3 to 4 percent. The number of tests has been fairly constant in the last few days (between almost 6000 and a little over 7000 tests during the weeks).
  4. Representativeness: We do not know how the number of those tested is composed. Are there many people returning from vacation? Many young? Lots of elderly people? Rather cautious? Depending on the composition, the positivity rate can fluctuate.

Better indicators

In order to capture how serious the situation is, the development of the number of people who died from Covid-19 and the number of patients who had to go to hospital would be more meaningful. These numbers have remained small in recent days, although the number of those who tested positive increased slightly.

  1. Delay: When the number of dead and seriously ill increases, it is actually too late. The aim of the pandemic measures would be to avoid these cases.
  2. Dark figure: How dangerous is the virus? There are two numbers for this: The IFR (infection fatality ratio): How many of those infected died (this includes an unreported number of cases that were not recognized; one assumes the people in the population who have antibodies)? Or the CFR (case fatality ratio): How many of the identified cases have died? This number is safer but less relevant. Studies come to an IFR for Covid-19 in the range of 0.6%. In the case of seasonal flu, one speaks of an IFR between 0.1 and 0.2 percent.
  3. Mortality: What can be taken for granted is that Sars-CoV-2 increases mortality in old age. That is also to be expected: in old age it is easier to die of a virus. But the virus doesn't just affect the elderly. There are also deaths and severe courses among younger people. However, it is also criticized which deaths are included in the statistics. People do not die of Sars-CoV-2, but of organ failure, which the virus may cause. Only an autopsy could provide reliable information. The figures, which are now being corrected in the UK and the USA, take into account the fact that all cases which had a positive Covid test in the last 30 days before death were recorded as Covid deaths, even if they were, for example, in died in an accident.

Beware of numbers - but we have nothing better

Action is doing something that is based on an analysis of the situation and happens with a goal in mind. The goal at Covid-19 is clear: We want to be in control. We know we cannot eradicate or turn off the virus, but we want to keep the consequences manageable. In particular, we must not overload the health system.

In the analysis, the matter is actually clear: the most precise information about the situation is provided by numbers. Numbers are easy to compare with one another, and numbers can be used for calculations (extrapolating a future course, for example).

So far, so clear. Now come the problems. It is not true that numbers cannot be discussed. Before data becomes numbers, there are a few things that need to be clarified. For example the categorization criteria: what is counted and what is not? Absolute numbers should always be used with caution. Numbers are relatively more reliable. Covid-19, but also other situations that do not develop in a completely linear manner, pose the question: What do we want to know and what can we know? The difference is not small.

Epidemics do not develop linearly. Not exactly chaotic either, but it's difficult to keep the many factors that play a role clearly apart. Statistical analysis works best when the numbers are high and as few factors as possible play a role. But that is precisely the situation that you absolutely do not want. You want small numbers, you want people to behave "correctly" - and thus "disrupt" the spread of the virus. Both of these factors also make it difficult to assess the effectiveness and usefulness of individual measures.