The problem with finding a fatality rate for COVID-19 is that there are people who have caught it but had only mild or no symptoms, who are not included in the number of cases and so skew the calculation. But now, due to extensive testing regimes in some countries, there are places where effectively everyone has been tested, so no mild or asymptomatic cases will escape notice.
The largest so far is Luxembourg, who have reported 644,901 tests for a population of 627,036, i.e. they’ve tested 102% of the population. Presumably this means they’ve tested some people twice and tested some temporary visitors too. But we can be sure that there isn’t a large pool of unnoticed cases.
They have reported 7,205 cases, with 5,848 recovered, 1,237 ongoing and 120 fatalities.* That gives a COVID-19 fatality rate of at least 120/7205 = 1.67% , and more likely 120/(5848+120) = 2.01% .
A similar calculation for Monaco gives a fatality rate of around 3%. The UAE has tested more than half the populace, and have a fatality rate of 0.5%
This suggests that there have been more than a million undiagnosed cases in the UK, and more than 2 million in the US. It also suggests that without containment measures the number of deaths in the UK and the US would have been much much higher.
Yesterday’s data from Texas comes to virtually the same result!
"Texas (8/8/2020) has an estimated 338,343 recoveries + 134,797 active cases + 8,343 fatalities.
This totals 481,483 CASES (RECOVERED, DEAD, AND WAITING TO SEE).
If you divide total fatalities of 8,343 by total cases, you get a TEXAS death
rate of 1.7% (8343 / 481,483 = 0.0173). "
The mortality rate of the average Flu year is 0.1%; this means Covid-19 is approximately 16 times
more fatal than the average Flu. The advantage of using Texas data is the presumption that if the
numbers are being cooked by the Republican-led state government, then the numbers would be
more favorable, rather than less favorable.
What you are talking about is the case fatality rate. The number of people who test positive is usually lesser than the number who are actually infected (because the disease is asymptomatic in many cases, people dont realise they are sick and hence dont get tested). The no: of people who are actually infected might be many times those who have tested positive and been recorded as actual cases.
What @Roy is talking about is the actual fatality rate.
WHO has estimated this value at 0.6%… it might be lower or higher in different regions (India seems to have a lower infection fatality rate based on random seropositivity tests done in Mumbai and Delhi).
Luxembourg seems to have an unusually high Infection fatality rate if the data is correct.
"The case fatality rate is the ratio of the number of deaths to the total number of positive tests,
while the
infection fatality rate [aka “Total Fatality Rate”] is the ratio of the number of deaths to the total infected population.
.
.
I was pointing out that using 2 different sets of numbers, we get close agreement.
Which is what we would look for if are hoping to validate Texas’ data set!
What you quoted for texas was the case fatality rate… @Roy was talking about the infection fatality rates.
Two different things.
The above article from WHO might held understand the difference.
Quoting the pertinent part below :
The true severity of a disease can be described by the Infection Fatality Ratio:
Serological testing of a representative random sample of the population to detect evidence of exposure to a pathogen is an important method to estimate the true number of infected individuals [7,8,9]. Many such serological surveys are currently being undertaken worldwide [10], and some have thus far suggested substantial under-ascertainment of cases, with estimates of IFR converging at approximately 0.5 - 1% [10-12].
The CDC estimates the infection fatality rate at 0.65%.
Much more if the CDC is correct. You can have a look at the attached study -
To quote:
The estimated number of infections ranged from 6 to 24 times the number of reported cases; for 7 sites (Connecticut, Florida, Louisiana, Missouri, New York City metro area, Utah, and western Washington State), an estimated greater than 10 times more SARS-CoV-2 infections occurred than the number of reported cases.
It depends on the test used. If they are using a PCR test then they can easily miss an infection. A person is only positive while they are actively infected, so unless everyone is tested by PCR once every 2 weeks or so they are going to definitely miss quite a few infections. The chances of catching a 2 week infection window with one test over a 4 month period is definitely not 100%.
If they are using antibody conversion then they are probably overestimating the number of cases since false positives are more likely with antibody tests, at least from my limited knowledge.
The link to the data is in French and my 2 years of high school French more than 2 decades ago isn’t up to the task. If there is an English translation somewhere it would certainly be interesting to see their testing methodology.
I did read your post. Here are where you are wrong -
You confuse case fatality rates with infection fatality rates. The actual fatality rate for Covid-19 is IFR. What you are actually trying to estimate with the luxembourg example is the infection fatality rate even though you dont realise it.
You pointed to luxembourg as an example and extrapolated the results to UK and US and estimated how many undiagnosed cases there are.
Your estimations are totally different from actual studies done by the CDC.
In short, your post is a bit of a confused mess.
I agree with this conclusion. However, it doesnt follwo from any of the data you mentioned. This conclusion is warranted no matter what the IFR is… because, containment measures -
a. Reduce the Spread of the disease.
b. Ensure medical facilities are not overwhelmed by cases. Leading to lack of care and more people dying.
In fact, this is the conclusion you can derrive from the luxembourg example.32% of the people died outside hospitals. Implying healthcare facilities fell short.
Comparing COVID19 stats to the seasonal flu is also a bit difficult. The vast majority of reported cases of the flu are not verified by any clinical tests, and most data is based on estimates from epidemiologists. The asymptomatic carry rate of standard influenza is unknown (as far as I am aware), and many reported cases may not be influenza to begin with.
Speaking as an armchair epidemiologist, the closest we have to an apples to apples comparison between COVID19 and the seasonal flu is people who reported symptoms.
