ľ�ϸ���Ӱ��

If it were up to prof. dr. Marcel van Assen, we need to interpret more statistics. "We need to be able to interpret these numbers as they help us decide what course of action to take to mitigate COVID-19". Read his opinion piece here.

Because of my frustration with the news, which mostly consists of confused people talking about COVID-19, I searched for data and statistics on the disease. Fortunately, some interesting websites exist that provide updated data (sometimes several times a day) on the  or the . Tomas Pueyo, in an that is read millions of times per day, analyzed and tried to interpret these numbers. Next to interesting statistics, such as the 17.3 days to go from catching the virus to dying on average, and the 26% of contagions happening before there are symptoms, we need to be able to interpret these numbers as they help us decide what course of action to take to mitigate COVID-19.

What is very clear from the statistics from Tomas Pueyo is how social distancing helps to stop spreading the disease. When analyzing the data on the events in Hubei in Chart 7 it becomes clear that after lock-down the number of new cases per day immediately started to get down. From analyses at the country level, it also becomes clear that some Asian countries/regions that were hit by SARS and took extreme measures early on (China [other regions than Hubei], Hong Kong, Japan, Singapore), “flattened the curve” early and do not show exponential growth in the number of Covid-19 cases. On the other hand, we see Korea, European countries and the USA who did not lock-down early and do show an exponential growth of cases, at least until now.

Typical in the evolution of the epidemic is a delay. Turning backing to the case of Hubei, analyses of data makes clear that when the number of diagnostic cases started to increase dramatically, the number of actual cases was already in the many thousands. There was a delay of about 11-12 days between the dates where the number of daily new true cases was at its peak (just before lock-down) and the date where the number of diagnosed cases was at its peak. Hence one should not look at the present number of daily diagnosed cases or fatalities to judge whether measures actually work(ed). And it implies that we may still expect the worst to come, with respect to diagnostic cases and fatalities, in the Netherlands and other countries, after lock-down around mid-March. Even when all measures are highly effective.

One controversial measure in the Netherlands is that primary schools are closed as of March 16, whereas primary schools in the UK remain open. The initial position of the Dutch government was to keep primary schools open, but mainly political and public pressure changed this position. The economic costs and restrictions of individual freedom of this measure are enormous. But what are its consequences for health-related costs? Let us turn to some available data.

First, no casualties are among children of 0-9 years old. Second, young children are otherwise also less affected by the disease, as is clear from the figure below. A bar in the figure shows, of all people diagnosed with COVID-19, how many of them are in a certain age group. Strikingly, very few children (< 1%, 0-9 years) are among them. The three most likely explanations for this low percentage are that they do not get the disease that easily when having the disease they do not show symptoms (or hardly so) and are therefore not tested, and they recover more quickly which makes them harder to ‘catch’ with the disease. Anyway, as COVID-19 is less likely to be transmitted when there are no or hardly any symptoms, children are less likely to transmit the disease compared to others.

Still, the young ones could transmit the disease to others. The extent to which children transmit the disease to others is not yet clear, and the Dutch government will investigate this in the following weeks. However, it is not clear how closing primary schools will prevent transmission of the disease. Children will still play with each other, and perhaps grandparents or other vulnerable people will take care of them when their parents go for work. On the other hand, economic costs are raised tremendously. Unfortunately, we cannot compare future fatalities and diagnosed cases of the UK and the Netherlands to see the consequences of closing primary schools, as the UK did not restrict social contacts as much as the Netherlands. Hence much more fatalities and diagnosed cases may be expected in the UK anyway.

Turning back to the figure, it seems that people in the age of 20-29 years are mostly responsible for spreading the disease. This is suggested by comparing the distribution of cases of Italy and Korea. In Korea, many suspected cases are tested, whereas in Italy only/mainly cases with serious symptoms are tested. Among mostly serious cases (in Italy), few are from the 20-29 year-group, whereas when also less serious cases are included (in Korea), the modal year-group is 20-29 years. The data in this figure suggest restricting the social relations of these people, as they are likely to have only mild symptoms but are likely to interact with others when studying, at work or in their social life.

What about the future of COVID-19? Can data help us predict this future? What to me seems difficult to decide is when to start picking up regular life again. The danger is to pick up regular life too early, but waiting (too) long brings along costs as well. It will not be easy being in governments’ shoes… Meanwhile, I will just keep tracking statistics and trends on COVID-19.