Over the last two weeks, German virologists tested nearly 80 percent of the population of Gangelt for antibodies that indicate whether they’d been infected by the coronavirus. Around 15 percent had been infected, allowing them to calculate a COVID-19 infection fatality rate of about 0.37 percent. The researchers also concluded that people who recover from the infection are immune to reinfection, at least for a while.
I cannot read German, and I am a little worried about how they calculated this. Suppose we find that as of today 10000 people have had the virus and 37 have died. I don’t think it would be correct to infer that the infection fatality rate is 3.7 percent. It could be that there are 50 people currently in the hospital, and if 40 of them subsequently die, then that would mean we should double our estimate of the infection fatality rate.
If we trust the 0.37 percent number (round it to 4 in 1000), we can work backwards from deaths to the number of people who have had the virus. So here is some speculative arithmetic. If 4 Germans have died for every 1000 who have had the virus, and the reported number of deaths is 2600, then the number of Germans who have the virus is 2600 * 1000/4 = 650,000.
We can play the same game for the U.S. Take 17,000 * 1000/4 and that gives 4.25 million people in the U.S. have had the virus.
2. Tim Skellet has some tweets about the German study. I noted this:
Overall in Germany: “So far, no transmission of the virus in supermarkets, restaurants or hairdressers has been proven”. #SARSCov2 was detected by PCR on “remote controls, washbasins, mobile phones, toilets or door handles” BUT NOT in infectious form.
. . .The entire #COVID19 outbreak in #Heinsberg district is traced back to a couple who attended a local Karneval festival. It all kicked off from there.
Pointer from Tyler Cowen.
3. I also liked another link from Tyler, the Dan Wang piece on life in Beijing during the lockdown there.
Shovel-Thermometer-ready jobs! We will never have to worry about how to quickly employ people during a downturn ever again.One of Skellet’s main conclusions is that surfaces don’t matter much for COVID transmission, and it’s all about airborne transmission in crowds. To the extent one believes that’s true, one should increase one’s estimate of the relative effectiveness of mask-wearing, no?
Also note that it takes an average of maybe 20-25 days after infection for someone to die. So if the death/infection ratio is 0.4% then that means there were 4m infections ~3 weeks ago. Of course we’ve been social distancing so the growth in new cases since then is not easy to guess. But it’s not hard to imagine maybe 15-20m infected by now.
Also note that there is a lag between infection and testing that seems to be around 10-12 days. So, say those infections from 3 weeks ago didn’t show up in testing until 3/31. We had 200k positive tests at that point. So that means there would be about 20x more infected people than our testing indicates. That would be good news because it means a lot of people are not serious enough to even get tested let alone end up in the hospital.
Buried in Skellet’s tweets was the important information- that the ratio of total infected to confirmed cases was about 25-27. Now, they need a larger sample, and one across the entire Heinsberg district to start- but you can extrapolate this to all the hotspots everywhere in Germany, and likely to all the other hot spots, too. Those studies should multiply over the next couple of weeks- in Germany and elsewhere.
In Heinsberg, they had 1400 confirmed cases, 39 deaths, with still 670 active cases. If you extrapolate from the first 730 closed cases, they will end up with about 78 deaths from those 1400 cases, or a case fatality rate of 0.06 (6%). If the infected were undercounted by a factor of 25, then the rate is 0.002 (0.2%).
This wasn’t meant as a direct reply to Josh. I must have hit the wrong reply button.
A death rate of 0.37 percent for Germany puts it in a statistical tie with death rate of 3.2 percent for the Diamond Princess (12 deaths so far/3711 passengers and crew).
I like this post, but perhaps we error when we accept covid-19 death rates, when there are multiple comorbidities.
A six-year old boy is infected with covid-19 and never even knows it, nor his 28-year old mom.
An 85-year÷old man with diabetes and a weak heart dies after being infected with the cold virus.
At the risk of sounding macabre, should covid-19 be given full credit for that death, or should the death rate be prorated in some way?
Manipulating the data is always going to be subjective. We already have stochastic models of this pandemic that are screwed up by the one directional subjective choices made in their creation. At this point, if a dead guy has a positive test for COVID-19, I am willing to accept that as the cause because I also know that the dead who didn’t receive a test, but still had COVID-19 aren’t being found, too. I suspect that the fraction of missed deaths is much smaller than missed infected, but it is still something.
Spikes in total death rates imply much undercounting of cocos deaths.
If the spikes are measured, then they weren’t missed, were they.
If you get 5,000 additional deaths but only 3,000 are attributed to coronavirus, it’s likely that a lot of the remaining 2,000 are coronavirus that didn’t get officially designated (died at home, didn’t get tested).
So what we really want to know is “excess deaths”. I am really looking forward to excess death figures for the first six months of this year–but knowing that they cannot be exact because they are statistical manipulations (e.g., trying to correct for change in population numbers and ages).
NYTimes has an article about excess deaths in NYC last month. Deaths from all causes more then doubled, an addition 5,300 deaths over baseline. There are only 3,300 or so confirmed COVID deaths (I can no longer get to paywalled article, so accept this is close enough on memory).
I doubt deaths from non COVID sources are up last month (fewer auto accidents), so this would tell me that our Covid mortalities are likely way undercounted.
“Excess deaths” will tell us whether we are over-counting or under-counting COVID deaths. (Actually, since we seem to be doing both, it will tell us what the net is.)
Though this may sound ghoulish, I am also curious to see if we find a pattern like, “New York has 22,000 excess deaths for the first six months and 10,000 deaths below baseline for the last six months, for a net “excess” of 12,000 for the year.” That would suggest that many of the people that COVID took would have died soon in any case.
So let’s play with something simple.
Let’s say all you people who hate the lockdown could PAY to offset any externalities that you’re getting and transmitting the virus might cause to an innocent bystander.
OMB gives a value of human life as something like $8M. I’ll use that, but I’m going to make the math really simple so you can change the number if you like.
Let’s say that mortality rate if you infect another is 1%. Let’s further be generous and assume that a particular reckless action only brings a 1% change of your being infected and transmitting it to another. So take $8M * 1% * 1% = $800.
If you were going to be charged $800 to go to a restaurant, would you still do it? How often?
You can substitute a different value for a human life if you want. This formulation makes it easy to compute. $2M yields $200.
I already hear people complaining that its way to high for a life, especially the old.
Beyond the fact that non-old people die from this, I’m leaving off several factors working against you.
1) Hospitalization and ICU rates are much higher than the mortality rate, and I haven’t included that burden in my calculation. Nor the permanent lung scaring. Nor the fact that the “milder” cases still involved getting very sick for a week or two (“mild” Covid cases merely means you weren’t hospitalized).
https://cdn.statcdn.com/Infographic/images/normal/21173.jpeg
2) If the medical system gets overwhelmed, the mortality rate could go higher. In Italy it’s a lot higher than 1%.
3) If the plague gets worse due to your actions its likely to have an even bigger negative effect on the economy than any lockdown.
Like all libertarianism, its always always always about wanting get away with lay ing externalities on other people for ones own benefit without being held accountable.
Another random thought I had.
For some time now, people have been extremely angry at me over my observation that not all live is equally valuable, and that some lives may in fact have zero or negative value. Based on this, I wasn’t keen on state subsidies to low value individuals or making them a part of my body politic through immigration. This stance made me the new Hitler.
Now that the plague is on, the libertarian response seems to be that they just want to know how many grandmas or people with health conditions they need to sacrifice to Moloch to get GDP back. And they actively state that, you know, their lives just aren’t worth that much anyway when you think about it so it’s OK.
Some people think that but no one dares say it. (And those who think it include some libertarians, some technocratic lefties, even some conservatives.)