Corona Dashboard EW16/W36 (weekly update)

John Jacobs & Gijs van Loef

Data 30 06 2020

Mortality per million population allows comparison between countries. Major differences are caused by differences in anti-epidemic strategy, which may be affected by the preparedness of countries to the epidemic.

General remarks by making a comparison between countries
We would like to make two important notes.
1. Countries differ in the way the count their mortality rates. The Dutch regional GGD stated early in the epidemic that the national RIVM did not count all cases, i.e. the Netherlands have relevant under reporting of cases.[1] People dying in nursing homes are rarely counted, since these patients will not be tested, even when suffering from COVID-19 symptoms. The Dutch statistics do not count untested people as COVID-19 victims. The data can be corrected for excess mortality. Because the data on excess mortality is delayed for about three weeks, a fixed correction factor is applied.[2]

2. Populations differ in age distribution, and old people have higher mortality rates due to COVID-19.
Using the double-corrected data, a ranking is made from dark red to dark green. The current ranking is (Table 1):
Dark red:> 1000 deaths per million population (> 1000 deaths / M). Ecuador, Peru, United Kingdom.
Medium red: 500-1000 deaths / M. Belgium, Chile, Brazil, Italy, Spain, Netherlands, USA, Sweden
Light red: 200-500 deaths / M. France, Ireland, Switzerland, Portugal, Canada
Ocher yellow: 100-200 deaths / M. Austria, Denmark, Germany
Yellow. 50-100 deaths / M. Finland, Norway, South Africa
Light green: 20-50 deaths / M. Estonia, Israel, Czech Republic, Poland, Lithuania, Iceland
Medium green: 10-20 deaths / M. Greece
Dark green: <10 deaths / M. South Korea, Singapore, Australia, New Zealand, Japan, Taiwan, Hong Kong

Table 1. Overview of mortality per million people (mortality/M), the raw data, corrected for underreporting [a] and age. Data to determine under reporting is derived from various sources.[b]

[a] https://hartblik.weebly.com/beter-c-verschillende-strategieeumln.html
[b] https://www.economist.com/graphic-detail/2020/04/16/tracking-covid-19-excess-deaths-across-countries , https://www.economist.com/europe/2020/05/09/many-covid-deaths-in-care-homes-are-unrecorded , https://www.ft.com/content/6bd88b7d-3386-4543-b2e9-0d5c6fac846c , https://www.washingtonpost.com/investigations/2020/04/27/covid-19-death-toll-undercounted/?arc404=true

It is striking that the best-performing countries are located near Southeast Asia, where the SARS-1 epidemic started in China 18 years ago. Many countries learned from this epidemic and were prepared for a new epidemic[3] Every virologist could have predicted that a new epidemic would come, as I (i.e. John Jacobs) did in 2003.[4]
Controlling a virus epidemic requires a different strategy than a bacterial epidemic. The default anti-virus strategy is testing, proactively detecting and isolation of infected cases.[5] To do this correctly is double efficient: lower disease mortality rates, and less social and economic damage, as only infected cases should be isolated. It is also widely practiced in case of veterinary epidemics.[6]

The policies of European and North American countries had weak to poor antivirus policy results.
Mortality is higher, sometimes up to 100 times higher than in Asian and Australian countries.
The impact on social well-being and the economy is much greater.
The differences within Europe are also huge when Greece and Lithuania are compared to the United Kingdom and Belgium.

We should learn from this epidemic.
It is very doubtful if we will develop a good and safe vaccine soon, since, vaccination is a hard strategy when antivirus immune responses contribute to mortality. It could be that antibody-mediated immune responses are the cause of mortality.[7] In most European countries only about 5% of the people were infected, so this epidemic could continue to spread for a considerable time with the potential to cause high mortality rates.[8] The need to learn and act from the learnt lessons from this epidemic is paramount.
New virus epidemics will come. Just like the prediction of this epidemic decades ago, we know that a new RNA virus epidemic will arise after AIDS, SARS-1, MERS, Ebola and SARS-2.
Our epidemic policy has a major impact on human behaviour and thus on the virus epidemic.[9] If we are prepared [10], the impact could be similar to the South East Asian and Australian countries that prepared themselves after SARS-1.

[1] RIVM is national institute for health and environment. https://hartblik.weebly.com/addendum-correctie.html
[2] https://hartblik.weebly.com/beter-c-verschillende-strategieeumln.html
[3] https://hartblik.weebly.com/voorkom-epidemie.html
[4] https://hartblik.weebly.com/ontsnappende-virussen.html
[5] https://hartblik.weebly.com/fighting-covid19.html
[6] https://hartblik.weebly.com/betere-bestrijding-van-de-epidemie.html
[7] https://hartblik.weebly.com/foute-afweer.html
[8] https://hartblik.weebly.com/beter-a-impact-covid-19.html
[9] https://hartblik.weebly.com/cijfers-en-data.html
[10] https://hartblik.weebly.com/verloren-door-te-weinig-kennis.html

Raw numbers
The Corona Dashboard 1.0 was launched in Dutch at 18th of April 2020 on our websites. It focused at the EW16, 16 ‘modern’ European countries with a population of at least 4 million people. These countries share political, democratic, social economic and technological development, making them excellent for internal comparison. The list consists of Austria, Belgium, Denmark, Germany, Finland, France, Greece, Ireland, Italy, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland and the United Kingdom. Japan, Republic of Korea (South) and the USA are added to complete the list, therefore EW16+3.

Table 2. Raw data from worldometer from EW16+3 [1]
[1] https://www.worldometers.info/coronavirus/

Reported mortality rates
The 16 European countries have huge differences in the mortality rates they report. Remarkably, the epidemic in Spain is the first one to have higher numbers of mortality, even before Italy. In Belgium the epidemic raised to the highest (uncorrected level), before being stopped (at about 6%) by the national policy.

Figure 1. Mortality rates for EW16+3. Above: linear, below: logaritmic
Country codes are explained in Table 1.

Case mortality rates
Case mortality rates show the differences of cases per country. Age corrections is shown to compensate for differences in population distribution. France has corrected its data of the number of cases on june 3d.
Table 3. Case-mortality rates.
Number of critically ill and recovered persons. The lack of data on how many people have recovered, shows the limitations of the care registration systems in the Netherlands, the United Kingdom, Sweden and Spain.

Link between testing and mortality EW16+3
Flu-like symptoms are no evidence for COVID-19, and not everyone infected by SARS-2 will develop symptoms. Policy making on symptoms fails because it discriminates poorly and would have people repeatedly in quarantine due to colds or other respiratory complaints. People succumbing to COVID-19 usually don’t go to many places anymore. Infected people are contagious before they get sick, and not everyone will develop symptoms. The high number of sick people in the Netherlands is partly related to many infected people continuing to spread the disease unmonitored, some even working in care facilities.

Proactive testing starts with testing everyone that has symptoms. If someone test positively, all the contacts will be tested. All contacts who are negative will be retested after 1 and 2 weeks. All infected people will remain in isolation – e.g. in their own house or in special hotels. In Wuhan, the government delivered groceries at their homes, to avoid the need of going outside.

An active testing policy will test many people who are not infected, because the infection rate in the population is usually (far) below 2%. Passive testing leads to a much higher number of infections per test. The danger of passive testing is that many infected people are hidden in the population. It is estimated worldwide half of the infections are caused by people without symptoms. These infections mainly occur in countries that do not test enough. People without symptoms and without a test result have no urgent reason to isolate themselves from social activities or go for  quarantine. Figure 2 shows the statistical relationship between the maximum percentage of positive tests in a country and the mortality rate in the population from COVID-19. Corrected data yield similar results.

Figure 2. Relation between mortality rate and lack of proactive policy, uncorrected data.
The shape of countries corresponds to their continent (circle Europe (EW-16), square other Europe, cross North America, triangle: Australia, diamond: Southeast Asia). Clear correlation. For reference R^2> 0.5 is considered a very clear connection in this domain. Country codes are explained in Table 1.
[1] https://hartblik.weebly.com/dashboard-i.html

Current status EW16+3

The current status shows how the epidemic develops in various countries. Many European countries have experienced a nonspecific lockdown to defeat the virus’ first wave that has paused the epidemic. This required more time in countries that have had a high peak, such as Belgium (Figure 3). The epidemic seems to be in control in all visualized countries, except for UK, Sweden and USA.

​Given the impact of the lockdown, it is obvious that many countries will let the loosen their restrictions. In Europe this is happening. This could reactivate the epidemic. The start of the epidemic is less noticeable in a cumulative graph than in a daily graph. Due to daily fluctuations, the data in Figure 3 is averaged over several days.

Figure 3. Daily mortality rates for 16 European countries + 3.
Moving average over 3 to 5 (and 7 if missing today) filtered for negative values and extreme high values (often data corrections). Country codes are explained in Table 1.

Figure 3A. Zooms in on the most recent dates of Figure 3.

John Jacobs & Gijs van Loef
hartblik.weebly.com corona dashboard

Dutch version, including remarks from healthcare experts: Corona Dashboard Dutch version EW16/W28