{no need to break out your tranlator DF}
https://esb.nu/blog/20059695/we-kunn...nen-aan-corona
Apr 17 20
Robin Frenchman
Blog
We can now count on corona
In my previous blog, I indicated that we should start calculating the effects of corona and corona policy.
In summary, it was concluded that we are not yet able to calculate because we lack data. We cannot calculate risks without that data. And since corona is potentially catastrophic, the only way - as long as there is no data yet - is to apply the precautionary principle and shut everything down as much as possible. To save time and then look further.
First numbers
So there was fundamental uncertainty, but it was largely removed yesterday. In a presentation to the House of Representatives, RIVM published the first results of research into antibodies by the Sanquin blood bank. We now have an initial estimate of the number of infections in the Netherlands.
Figure 1 Source: RIVM
If we combine Sanquin's data with figures from RIVM, CBS and the NICE Foundation, we get table 1. Persons under 20 and above 69 were not included in Sanquin's study.
Table 1
First impression of health risks
Based on these data, we can calculate the probability per age group that someone who contracted a corona test, ended up in the hospital, was admitted to intensive care or died. This means that we can now see how great our own individual risk is if we become infected with corona. And employers can, for example, look at the health risks of their own employee population.
The question is justified how reliable these data are. Blood donors are of course not a perfect approach to the Dutch population, and those who feel sick or anxious will skip the blood bank during this period. And it may take a while for antibodies to become visible.
Sanquin's research may therefore underestimate the number of infected people, but on the other hand, this also applies to the data on the number of deaths. There are few people under 70 in nursing homes, so the effect will be limited, but people can also die at home, which we miss.
So both denominators and numerators still have some question marks, both upwards and downwards. If there are more infections than we know on the basis of the Sanquin data, the risk of hospitalization, IC or worse after a corona infection is smaller. At the same time, if there are more deaths than we know, the risk of death is greater. But the chances are just as great that these effects cancel each other out. But at least you can assume that the current data gives a good first impression of the risks.
Health risk per infected
With due observance of these comments, we can calculate the probabilities and thus the risks by dividing the different turnover results by the number of infected persons (table 2).
Table 2
Mind you, these are risks you run once you contract the infection. The overall risk is smaller because not everyone will contract the infection and that risk depends on behavior, environment and timing.
And the risks, of course, are based on the past: these were the odds from the past eight weeks, and we assume for now that those odds remain constant. Those chances may get better the more we know about the treatment, but perhaps also worse if, for example, the IC capacity is exceeded.
Low mortality risk
Then comes the question how should we interpret these opportunities? How high or low are they? To put the risks of being infected with corona into context, we can compare them with the normal mortality risk, the chance that we will die in any year. Then you get table 3.
Table 3
In the third column, I have included the number of 2018 deaths by age category. And in the columns thereafter, the mortality risk from the coronavirus compared to the normal mortality risk in 2018.
So Corona adds extra risk of death. But you shouldn't just add the risks of corona and "normal" together. After all, anyone who dies from corona can no longer die from another illness or from an accident. We can only determine the total mortality probability this year afterwards.
The chance of dying from corona is smaller than the "normal" chance of dying. In that sense, the risks of corona can be called low, especially for people under 60. And under 50, the chance of IC absorption is also very small, and so the potential burden on the ICs from that group is also small.
Underlying diseases
Moreover, the chances depend strongly on underlying diseases. Of the 387 deaths in these age groups of 20 to 70, at least 70 percent had a pre-existing condition, such as diabetes, heart disease, cancer, or other conditions. For those who do not have such a disease
Finally
Whether you find the risks high or low or whether or not acceptable is of course a personal consideration. But I can imagine that people and organizations see this as arguments for controlled reopening, starting with healthy people under 55 who also do not live with vulnerable people.
Postscript (May 4, 2020)
A week after the publication of this article, RIVM announced the results of a second antibody sample during a technical briefing in the House of Representatives. This sample, called Pienter, allows us to redo the calculation with more data.
There are two additions. The calculated risks in the blog above are based on the infection rate at the end of March, with the course of the disease until April 15. With the Pienter sample we can do the calculation again with the infection rate from the beginning of April and the disease course until April 27.
The Pienter survey also measures all ages and not only 20–70 years. As a result, we can now also look at the risks of ages 0–20 and 70–85 years. This only concerns the risks of corona after contamination. Nursing homes are not included in the calculation: we do not know the degree of infection in nursing homes and the number of deceased people.
The results of this second calculation hardly deviate from those of the first calculation (compare table 4 with table 2). The Pienter sample is smaller than that of Sanquin (2,800 vs. 4,000) and spread over more age cohorts, so finding meaning in these deviations is of little use.
That we can now also get a good impression of the risks of children and the elderly is an addition to the insights. The effect of triage in the IC is also visible here. Above 75, the chance of death is greater than the chance of admission to the ICU. In that age category, people are admitted to IC less often. This is partly because they no longer want to do so themselves, and partly because admission to the ICU for many people over the age of 75 appears to be not medically meaningful or is estimated as such.
Tabel 4 Risico’s van corona na infectieBron: RIVM Pienter onderzoek, CBS, Stichting Nice
Sources of the figures:
Infection rate
Infectiegraad
Size of age cohorts
Omvang leeftijdscohorten
Deaths per cohort
Overlijdens per cohort
Hospital admissions per cohort & Positive tests per cohort
Ziekenhuisopnames per cohort & Positieve testen per cohort
