There are early results from antibody measurements indicating that the death ratio from coronavirus would be as low as 0.3%. This is much smaller than 2.5% that I estimated. I will make another estimation, but not from antibody measurements as they may only show how many were attacked by coronavirus but did not actually get the virus and probably did not get immunity.
We can use data from healthcare personnel. They have been tested, so we can trust the number of infected, and the number of deaths corresponding to this number of infected is also a reliable number. There may be more deaths among healthcare personnel, but they are not in the used number of infected. The statistics from three places is quite similar.
In China 3,400 healthcare workers were infected, 13 of them died. That is 0.38%.
In Italy at one time 37 doctors died and 6,200 healthcare personnel were infected. That is 0.5%.
In the USA (in one place in one time) 27 healthcare personnel died and 9,200 were infected. That is 0.3%.
The figure is between 0.3% and 0.5%. As healthcare personnel have more women, the average is about 40 years, and all were healthy, we can estimate that in the general population (with almost the same number of men as women and many ill people) the death rate is a bit higher. It can be p=0.5-0.6%. But if it is so that many healthcare personnel are still ill, the final death toll may be higher.
From New York Health 14. April 2020 we have the data of deaths: (year range number) 0-17 3, 18-44 309, 45-64 1581, 65-74 1683, 75- 3263. This data does not give directly the range 0-30, but I estimate as follows: 25% of the population is in the range 0-21 and has about 9 deaths, 50% of the population is in 22-64 and has about 1881 deaths, 25% is in the range 65- and has 4946 deaths. The death rate in the middle range is p=0.5-0.6%. Then the death rate in 65- is:
(0.50/0.25)(4946/1881)p=5.26p
In the whole population the death rate is
p=0.25*0+0.5p+0.25*5.26p=1.8p=0.9-1.0%
Thus, my earlier estimate of 2.5% is too high. The estimate from antibody measurements of 0.3% is too low. The correct death ratio is about 1%. It is about the measured death rate in Iceland. Yet, the final death toll of the healthcare personnel and Iceland may still go up. It is not completely excluded that the death ratio is over 1%. It hardly can be much under 1%. This indicates that not all who have antibodies are immune to the virus. They may have survived a weak attack of the virus but are still vulnerable.
Notice that Greece apparently manages to stop the epidemic and has the death ratio 5%. If the actual death ratio is 1%, then there would be 5 times as many infected than are counted. This more supports my original estimate 2.5% than 1%, but 1% is possible: if the people without symptoms have not infected other people due to social distancing, then it may just and just be possible that the epidemic dies out even if only 1/5 of infected were noticed. I would still be uncertain of it. It may be 2.5%.
Of course, the death ratio is not an intrinsic property. It depends on how good are the doctors, how good is the equipment, and how ill and who are the patients. The death ratio does not need to be the same in all countries and estimating it from the best countries gives only a lower bound. For most countries it is very possible over 1% and more like 2.5%, and of detected infections 5% or much more.
Finnish health authorities still try to limit the epidemic, not to stop it. I think it is a wrong idea. Then it will start again or keep the economy down. It can be stopped like in Greece will happen. That would be the best for economy and for people.
But how could the presence of antibodies for the coronavirus in blood not mean that you are immune? I am not any medical doctor, but I think it is like this. Assume you plan on traveling somewhere where there is hepatitis. So you go to a doctor and get an injection of gammaglobulin for hepatitits. That is, antibodies, proteins specific for binding to this virus. But you have to get enough of these antibodies. Take too little and they do not protect. If you get infected by a virus and eventually die of it, your body does produce antibodies, but not enough to beat the virus. The virus grows in your body exponentially and you need enough antibodies to stop it. Assuming you got infected by few viruses, your body produced some antibodies and stopped the virus. Then you have some antibodies in your blood. If you get infected with very many viruses, they many not be enough, so you are not immune. If, say, in 10 minutes being exposed to someone who has the virus, you get some viruses, then if you are exposed for a day to the same infected person, you get 144 times as many viruses from the person. Your antibodies may not stop this stronger attack. I mean, that’s how I would think it is. Not a medical doctor though.
4 Comments
You should check out UK Column’s videos. I think they do a swell job at reporting the facts concerning COVID-19.
The mortality rate would have to encompass an endemic area’s infection rate; or try to average out every one of them based on population density and medium age. If you leave out infection rate; then you cannot obtain the mortality rate but rather the death risk.
Right, I originally wrote death rate when I meant death ratio, that is death risk (probability of dying of you are infected). I did correct my spelling error later. The rate of deaths naturally depends on the size of the population. I will later look at those videos. But now I have one project with theoretical physics. Just wrote an book and am waiting what the publisher, who asked for it, says about the first draft. So, just now I will not study the videos. I just wrote this short post as the Finnish health authorities (despite a dissident opinion from their own group) still do not try to stop the epidemic. And I am not even sure if the virus does not hibernate in your body, like herpes, and you never get really rid of it, and these people want herd immunity.
Mortality risk. Yes, I guess medical and insurance people would use the word risk. Or, for the insurance people, if a person dies, I guess it is properly called the mortgage payback risk. When I worked in the military, I would have called it the kill ratio, but as my Alma Mater was mathematics, and there are the birth and death processes, I call it the death ratio. What you fancily call life is just a simple case of a birth and death process. The death ratio is a probability and in traffic theory we would often use the letter p for it, but in the city traffic planning department they would say: Hey, you cannot write it as p. P is a parking lot.
In East-Asia the number of infected people ánd the death ratio seem to have both lower values than in Europe.
Young children in East-Asia are systematically vaccinated with BCG-vaccin against tuberculose at young age, whereas in Europe BCG-vaccination is not everywhere applied anymore. In Belgium for instance children don’t get vaccinate at all with BCG save few exceptions, and the vaccine is not produced anymore in our country.
Is it reasonable to assume a connection between both phenomena ?
It is reasonable to expect that there are connections of this type. Not necessarily with tuberculosis vaccination but with something. It may be the vaccination, it may be something else. But men are more likely to be infected and die than women. It may be explained by women being in better shape in old age, or that men smoke more, or anything. But there is some reason. Then there are clear differences with the mortality risk: in Germany, Australia, Iceland, Israel, all East Asia mortality seems to be lower while in Western Europe in Italy, Spain, France, Benelux, UK it seems much higher. Interesting proposal that it could be the BCG-vaccination. Something there is. I was even thinking if this virus is not exceptionally deadly for Y-DNA R1b/R1a men (probably this is not so). There was some small correlation with the blood type, A was more in risk, O least.