Confusion about EU COVID-19 risk

It's easy to get alarmed about the risk of catching the coronavirus, SARS-CoV-2, when looking at a map that shows your country or region in dark red. Color-coded COVID-19 maps, highlighting danger zones, are everywhere now. And often it's individual countries, drawing up their own maps, based on their own data.

The European Union (EU) wants a more standardized approach, and the idea is simple: Provide citizens with coordinated maps, published every week, showing the levels of risk for travelers across the EU's 27 member states. 

It uses a so-called traffic light system: Red, orange, green and gray.

As DW reported when the recommendations were announced on October 13, the traffic light system works on the basis of new infections, or incidence, per 100,000 inhabitants in the previous 14 days combined with the number of positive COVID-19 tests. 

The recommendations say you are: 

• green if the 14-day cumulative coronavirus case notification rate is less than 25 per 100,000 people and the test positivity rate of tests for coronavirus infection is less than 4%.
• orange if the 14-day cumulative coronavirus case notification rate is less than 50 per 100,000 people but the test positivity rate is 4% or more, OR if the 14-day cumulative coronavirus case notification rate ranges from 25 to 150 but the test positivity rate of tests for coronavirus infection is less than 4%.
• red if the 14-day cumulative coronavirus case notification rate is 50 or more per 100,000 people and the test positivity rate of infection is 4% or more, OR if the 14- day cumulative coronavirus case notification rate is more than 150."

And then there's gray for regions where there's not enough information, or where the testing rate is 300 or less.

That's the theory

The traffic light system has been agreed by all EU member states, but it's only a recommendation. So, as with most recommendations and guidelines, the effectiveness of the map depends on how well it is enforced.

There is also the issue of how the EU's individual states — including individual regions within those states — collect their data. Until recently, European countries have assessed their COVID-19 threat level on different criteria.

Germany, for instance, introduced a threshold of 50 new infections per 100,000 inhabitants per week, above which social, economic and health restrictions are to be tightened. Germany's state of Bavaria, however, has set a lower threshold of 35 new infections per 100,000 inhabitants per week.

Such variations in the data can be adjusted to meet the EU's new standardized thresholds. And that may indeed be helpful for travelers, trying to assess the level of risk of traveling to a certain country. 

But it's not only that the data has been collected differently in each region, so far, but that it is also potentially perceived differently by the people who live there. And that could still lead to confusion, if not alarm.

Those variations have existed for months at local levels. And that begs the question whether residents in Bavaria, for instance, will stay focused on their threshold of 35 new cases and subsequently perceive a lower level of threat. Or, alternatively, whether citizens in regions that have had higher thresholds will now be alarmed by a newfound "red status" if they regions are suddenly perceived to be riskier than before. 

Arbitrary numbers?

The threshold of 50 new cases per 100,000 inhabitants per seven-day period has itself been criticized.

Luxembourg's Minister of Foreign and European Affairs, Jean Asselborn, has, for instance, described it as a German invention. Others have said it was based on the point at which German hospitals would start to feel the pressure — if, for example, there was a surge in demand for intensive care beds. But that would be different for every other country, region or city. 

Stefan Willich, who heads the Institute for Social Medicine, Epidemiology and Health Economics at Berlin's Charité hospital, meanwhile, has said the threshold may have been "a precise indicator" five months ago but it's only a rough indicator now.  

So, look at that criteria again: For instance, orange for regions reporting less than 50 new cases, and positive tests of over 4% — or an incidence of between 25 and 150 new cases and positive tests below 4%. 

The span between 50 new infections, on the one hand, and 25 and 150, on the other hand, is huge. Mixed with the number of positive tests, either above or below 4%, and the potential for misinterpretation, especially by non-statisticians — that is, the general public at whom the new EU map is aimed — is massive.

And it's not just about the number of new "notified" cases, but also the number of tests that are conducted. Experts continue to raise concern about that, suggesting that the rates of testing for COVID-19 need to be standardized as well.

Data published by ourworldindata.org, a site run by the University of Oxford in the UK, suggests there's still a broad span of testing rates across Europe. 

For October 16, 2020, one chart shows the following testing rates per day per thousand people:

• UK: 3.85 (mostly red on the EU map)
• France: 3.52 (red)
• Spain: 2.24 (mostly red)
• Germany: 1.99 (orange)
• Italy: 2.13 (half red, half orange)
• Belgium: 4.79 (red)
• Denmark: 5.83 (orange)
• Czech Republic: 2.55 (red)
• Poland: 1.05 (red)
• Austria: 2.01 (red)   

It doesn't take a statistician to surmise that the less you test, the fewer positive cases you are likely to find and report. But, then, as you can see, that's not always true either.

Hannah Fuchs and Bernd Riegert contributed to this report.