area42 - BLOG
COVID-19 Update April 13th 2020, HB Dreis
COVID-19 Growth Rate further crumbling.
With 505,561 actice and 27,421 new cases yesterday (April 12th) in the US the growth rate per day is now down to 5.4% (down from 34% at March 19th). Social distancing being certainly a major contributor to this positive development.
In the graph below the Growth Rate is calculated from the number new cases and the number of active cases at the day shown (only active cases can contribute to the spread).
COVID-19 Update April 11th 2020, HB Dreis
He also elaborates in a great way how this will reduce the spread of the disease so much more efficient because so much less people with COVID-19 who do not know they have it will get in contact with others. Romer considers this as so important that he would not wait for the availability of faster reliable tests but to massively top up the numbers of the currently available tests asap.
COVID-19 Update April 10th 2020, HB Dreis
The major goal of the COVID-19 information provided in this blog is to create hope based on facts.
After the available data on April 8th proved a slow down of the growth rates of COVID-19 in the US (which also holds for many other countries in the world) we like to emphasis on the commonly stressed comparison with the flu seasons. We definitely don't want to dismiss the dramatic dimension of the pandemy or to downplay its consequences on individuals and the health care system. But a closer look can help to better understand future governmental and other public measures and how to deal with the economical impact.
As the long term consequences of the current economical impact can be dramatic as well it's imperative to come to realistic conclusions to gear up for normalization of the economy.
To do so we present some more hope giving numbers.
A new study of the Univerity of Gottingen in Germany concluded that worldwide only about 6% of all COVID-19 cases and only 1.2% in the US are kown.
If this is true this might sound horrifying in the first place but in reality it would be rather good news. Why that?
Before we answer the question a short reminder on how we calculate the death rate and thus the mortality of a virus. The mortality is the percentage calculated from the number of death by the virus divided by the number of total cases, times 100 (to get the percent value). The accuracy of the derived mortality depends on the accuracy of the two numbers, death by COVID-19, and total COVID-19 cases. Both numbers have right now very limited accuracies but the number of death by COVID-19 is better known, even if it is most likely a little too high. The reason for that is that currently everyone who dies having the virus is counted as dying from the virus which is certainly not entirely correct but also not too far off. If the Gottingen numbers are however correct this means that the second number - the total cases - which is used to calculate mortality is too low by a factor of 16 world wide and a factor of 55 in the US. The consequence would be that the calculated mortality would in reality be lower by the same factors.
Taking the currently best COVID-19 mortality guess of the WHO which is 3.4% into account this would bring down the COVID-19 mortality to 0.2%. This would be a correction to put a big exclamation mark behind because it is matching very precisely the mortality of the flu or influenca. And despite of the hardship COVID-19 is causing this would be indeed good news about the mortality rate (of course not so good news about the number of total cases).
Having said that, here is another number worth looking at in the light of this statement. The flu has caused in the same period of time COVID-19 is raging between 29,000 and 59,000 death in the US and COVID-19 has caused as of today about 17,000 in the US.
To close this post and to circle back to the post from March 23rd where we called for masses of fast tests we would like to state that this was a great idea on the one side but on the other side reliable (!) and fast tests are still not available and will take much more time to develop.
There is however an efficient method to test if the numbers of Gottingen are correct and this method is called representative tests. It means that instead of testing only persons who have symptoms, to test random chosen but representative groups of people for COVID-19. Such tests have already been started in some countries but much more are needed. In particularly because the differences between outbreak hot spots and other regions with less cases need to be investigated in this representative way. To calculate the real number of infections from those samples the same software used for extrapolations after elections can be used to produce reliable numbers of the overall invection situation. And we know how precise such extrapolations can be.
COVID-19 Update April 8th 2020, HB Dreis
In the post from March 23rd (below) we predicted possible numbers of COVID-19 cases for April 8th (today).
The predictions were based on different growth rates between 10% and 35% which represented the uncertainty of kowledge about the real growth rate at March 23rd in particulary not knowing the effect of social distancing. Today we can derive the real growth rate in the last 14 days with all its attempts to slow down the spread.
The good news is that with 435,000 cases today the growth rate was only moderate in comparisson of what appeared to be a possible worst case scenario (35%) 14 days ago. The average daily growth rate in this critical period has been 17.6%, a very encouraging development in comparison of what appeared to be possible.
But even more important, that's just the average. The growth rate is definitely declining over time which is even better news in a situation which is very bad news overall.
There are certainly various reasons for the slow down and social distancing is certainly one of them but not the only. Also the observation of outbreak hot spots is contributing combined with the observation that not every person acquires the disease even if being exposed, which results in a kind of exhaustion of the spread in hot spot areas, similar to the immunization barrier in populations for influenza.
New post: Created March 23rd 2020, HB Dreis
COVID-19: What we need are TESTS, TESTS, masses of TESTS!
The growth rate in the US from Saturday Feb 21st to Wednesday 25th was 33%, 30%, 25%, 25% -> Worldometers
Even if this looks like a slow down (we hope it is), the rate is tremendous high.
To illustrate what a sustaining daily growth rate of this levels means, here are projections for the situation in 2 weeks.
Number of cases on April 8th depending on growth rate:
• Daily growth rate 10%: 258,000 cases
• Daily growth rate 20%: 873,000 cases
• Daily growth rate 30%: 2,677,000 cases (likely, based on current growth rate)
• Daily growth rate 35%: 4,500,000 cases (possible, based on current growth rate)
Here we see unfortunately the nature of exponential growth and what will happen if the growth rate keeps its pace. It also makes immediately clear why avoiding contact is the best and only option which exists without a vaccine. So, reducing social contacts makes sense, as it will inevitably reduce the growth rate even if we cannot predict how much. In the long run a vaccine is of course the best we can hope for.
But there is another factor. We also have to protect our economy. Not only to keep the standards of living we are used to but also to protect life.
As the date of availability of a vaccine or other working medical treatment is right now mostly unpredictable, we have to think beyond vaccines and about options to normalize the economical processes much earlier.
One way to achieve this in safe way is: TESTS, TESTS, masses of TESTS! - instandly (in minutes), everywhere and for free
The likelihood of the availability of a fast and reliable test which can be provided in masses soon is higher than that of a vaccine. But the positive impact of mass tests could be a total game changer.
With mass test capabilities we can much better isolate ourselves if we test positive or continue with our daily lives if we test negative. And we could do those tests every day or as many times as required.
Companies could screen employees basically at the doorstep and airports can screen passengers before going through security. We could almost return to normal with easy tests and much earlier than it’s thinkable now.
Such “public” tests can also easily be done entirely anonymous by creating a non-personal ID when dropping a sample. Because the testing will take most likely some minutes, the ID will have three states; not yet processed, positive or negative. A scannable tag is issued to everyone who gave a sample and a display system is showing the results. Those who are tested positive simply go home and isolate themselves. To be on the safe side, each tag gets also scanned when entering a building to double check the result.
At the end of this post a few numbers which give some hope, and which were published on March 5th by Timothy W Russell et al. from the Centre for the Mathematical Modelling of Infectious Diseases, Department of Infectious Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK.
The numbers were computed from the Diamond Princess cruise ship. From the 3,711 passengers and crew members onboard 3,063 were C-reactive protein (CRP) tested. This represents a test coverage of 82% and thus the best coverage ever in a contained group of people since the COVID-19 outbreak. In this respect the numbers of the Diamond Princess provide the best foundation for infection and mortality ratio calculations which currently exist.
For the Diamond Princess the infection rate was only 19%, the percentage of positive tested persons who had no symptoms was almost 50% and the overall mortality rate at an average age of 58 years was 0.5%. For comparison the flu mortality rate is around 0.2%.
If the numbers of the Diamond Princess turn out to be applicable to the outbreaks at land, COVID-19 is still worse because deadlier than the flu, but not by a factor of 10 anymore but by a factor 2.5. Let’s hope.