We are developing the social individualist meta-context for the future. From the very serious to the extremely frivolous... lets see what is on the mind of the Samizdata people.

Samizdata, derived from Samizdat /n. - a system of clandestine publication of banned literature in the USSR [Russ.,= self-publishing house]

Germany and Italian COVID-19 data – and why they’re so different

Is there something about being Germany which protects the body against coronavirus Covid-19? Probably not, I would guess. In which case why do the latest figures from the Robert Koch Institute show that the country has a case fatality rate (CFR) of 0.3 per cent, while the World Health Organisation (WHO) figures from Italy seem to show a CFR of 9 per cent? To say there is a vast gulf between those figures is an understatement. If nine per cent of people who catch Covid-19 are going to die from it we are facing a calamity beyond parallel in the modern world. If only 0.3 per cent of people who catch it die from it, this pandemic may yet turn out to be no worse than seasonal flu, which as I have explained here before is estimated by the US Centers for Disease Control to kill between 291,000 and 646,000 people a year without the world really noticing. According to John Hopkins University, which is collating fatalities data, 15,308 have died to date.

Ross Clark, Spectator (behind paywall).

A couple more:

Germany is almost certainly behind Italy in this epidemic. But the main difference between Germany and Italy lies in those countries’ respective attitudes towards testing. Germany has carried out far more enthusiastic testing of the general population – there does not seem to be a central figure for this, but the German Doctors’ Association has estimated that 200,000 people across the country have been tested. In Britain, it is 64,000 people. On the other hand, German hospitals do not routinely test for the presence of coronavirus in patients who are dying or who have died of other diseases. Italy, by contrast, is performing posthumous coronavirus tests on patients whose deaths might otherwise have been attributed to other causes.

It stands to reason that the more people who are tested, the more accurate a picture we will have of the mortality rate, the transmission rate and other metrics which will determine the eventual path of this pandemic. To underline the uncertainties behind the data from which policy is currently being made, the Royal Society of Hygiene and Tropical Medicine the other day estimated the number of people in Britain who already have or have had Covid-19 at between 6,000 and 23 million. That is a pretty broad spread with hugely different implications. If only 6,000 have the disease in Britain, socially-distancing the population or locking down society might have a purpose. If 23 million have the disease, it is pointless – it already has ripped its way through the population but without killing more than a tiny percentage.

What we really need is a huge effort to test a large randomised sample of the population to see how widespread the infection is. Hopefully, that will soon happen. But in the meantime, I am minded to think that the more accurate picture of Covid-19 comes from the country which has conducted the most tests: Germany.

The more I read, the more it seems to me that randomised testing, as Clark writes, is crucial, not least in reducing hysteria and the effects that hysteria is having, and will continue to have, on our society and business. The costs, in terms of stress, the destruction of businesses and so much effort, needs to be weighed.

I have come across a few comments on social media, of the passive/aggressive sort, that “it is so amusing to see all these experts all over social media on epidemics” – the implication being that the mass public should shut up, “stay in your lane” and let the men/women in white coats take charge. The problem with this however is what happens if the experts disagree, or if their policy advice causes so much damage, including to the liberties and welfare of the citizenry, that a democratically elected politician has to make a judgement call? We don’t, rightly, outsource vast coercive powers to everyone who claims to be an expert on something. That’s not how it is supposed to work.

28 comments to Germany and Italian COVID-19 data – and why they’re so different

  • Mr Ecks

    The test is not really that good, having–I understand at least –a up to a 47% false positives rate.

  • Snorri Godhi

    23M currently or previously infected in the UK seems unlikely, since positive tests in Italy are more than 10 times those in the UK, and there cannot be 230M infections in Italy, given a population of 60M!

    Having said that, i already thought that Italy must be grossly under-estimating the number of asymptomatic cases. From the 2nd quote, it seems that they might also be over-estimating the number of victims of the virus. OTOH the Germans might be under-estimating their number of victims!

    Perhaps the best strategy would be to test everybody in the worst affected areas of Italy: that might provide the best estimate of the mortality rate.

    Much has been made of the fact that Italy has an aging population and almost all the victims had prior conditions, but is the number of old people with prior conditions really that much lower in Germany? Obviously not, or else people would be moving to Germany when they retire.

    It is also possible, of course, that potatoes confer a degree of immunity, while eating pasta is a risk factor. Better pasta than bat soup, however.

  • Snorri Godhi

    PS: I have also seen an article that says that Italy might be under-estimating the number of victims, because even excluding the ‘official’ victims of the virus, mortality is higher than usual for this time of the year.

  • Nullius in Verba

    The problem with randomised testing is that – unless it has already spread out of control – it would take a huge sample size to get any sort of accurate figure, and you would have to test everyone very week, since being clear one week doesn’t mean you’ll be clear the next. If about 1% of the population now has it, you would have to test 100 people to have an reasonable chance of finding just one, and 10,000 people to be able to get even a rough and noisy indication with 100 people, and this is certainly not going to find more than a tiny fraction of all the people who have it. So I gather they’ve decided to save their money until later when the data will be more useful.

    I agree that the testing being done in the UK isn’t giving us very useful data. We’re not even testing all the people with symptoms sat at home, and I get the impression that there is a growing backlog, since the number of confirmed cases in the UK is going up at a rate of 10-fold per 10 days, but the number of deaths is going up 10-fold per 7 days. Unless the mortality rate is gradually changing, and I can’t think of any reason why it would yet, the data from testing is very likely duff and getting duffer.

    So, let’s just look at the number of deaths. That’s at about 300 at the moment, and has been going up 10-fold a week. If we suppose the virus kills about 1% of those that it infects, then there must have been about 30,000 people infected at the same time the dead were infected. Since it takes about a week to show symptoms, a week for the symptoms to turn serious, and one to two weeks for people to die, that’s about 3-4 weeks ago. And with a 10-fold increase per week, that would mean there’s another thousand-fold increase in the pipeline already.

    On the other hand, if the mortality rate is higher – let’s say 5% just to make the numbers easy, then the 300 deaths so far means there were only 6,000 infections three weeks ago. That’s an order of magnitude better, but hardly comforting in the long run!

    Conversely, if the mortality is much lower than 1%, then pretty much everyone has got it and is in the pipeline now, it’s too late for any quarantine, but many more are going to survive.

    The experts don’t know. They’re scrambling to analyse the data, and it’s clear that they’re figuring out a lot of things late. And if it turns out they acted too late, then we’re talking half a million dead in the UK over a period of a few weeks just to start. (1% of the whole population is about 6 million.) It could even be worse than that. Or then again, it might be better. Maybe the mortality rate is lower than we think. Maybe we acted in time.

    But fundamentally, the politicians are driven by the knowledge that people care more about people than they do the economy, and they’d rather live with a recession than with the emotional and political impact of half a million dead or more, and the politicians are doing what most people would want and expect in trying to save lives first. They might change their minds. If it turns out they screwed up by waiting too long, or jumping too early, the politicians might yet pay a price. But the biggest problem with this is that nobody can see what’s coming. Not the experts, not the politicians, and especially not the general public! The mathematics of exponential rise, combined with a three-week lag, mean that what you’re seeing now tells you maybe only a thousandth of what’s coming. What you see now is potentially very misleading. We are all blindly rolling down the train track in the fog, and there might very well be a ravine out there. It takes 3 weeks to stop the train. Does anyone wanna put the brakes on?

  • Stonyground

    The problem with trusting the men in white coats is that their predictions of calamity have an unenviable track record of failure.

  • Ferox

    As of right now, Italy has had 5476 deaths. Germany has had 115.

    Those are absolute numbers, not rates. Therefore they can’t be explained (solely) by disparities in testing rates. Morbidity aside, Italy has had 50 times as many deaths as Germany, despite having a 25% smaller population.

    Will Germany shortly follow this track? If not, perhaps that points to a qualitative difference in health care, or some other systemic difference. But it cannot be just an artifact of the statistics.

  • Rudolph Hucker

    Anyone got a link to the Royal Society of Hygiene and Tropical Medicine estimates?

  • Rudolph Hucker

    Here’s a link to the ONS stats:

    https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/excesswintermortalityinenglandandwales/2017to2018provisionaland2016to2017final

    Headline point:

    In the 2017 to 2018 winter period, there were an estimated 50,100 excess winter deaths in England and Wales.

  • Nullius in Verba

    “Will Germany shortly follow this track?”

    Here’s some data. Triangles show cumulative number of deaths. Have a look at the slopes during the early stages. Italy started earlier, so it’s higher, but is starting to level off.

  • Ferox

    Nullius, that graph appears to combine the death and confirmed case data into one number.

    The problem, as I understand it, is that Italy and Germany have different testing rates, so that while deaths are deaths, those confirmed case numbers are not strictly comparable.

    It would be interesting to see that graph without the confirmed case data embedded in it (i.e. only deaths) to compare the growth curves for Germany and Italy.

    EDIT: Ha, I can’t read. I see the death data there in the graph. Yes, it looks like a similar curve, just three weeks behind.

    Let’s hope that doesn’t come to pass – for Germany, and for the UK and United States as well.

  • Itellyounothing

    Niv

    “(1% of the whole population is about 6 million.)”

    I think you are talking about the UK and that’s more like 12% or 13%.

    Also, experts have guessed zillions dead for every tiny whiff of a pandemic for 20 plus years because there is more money and power in it than guess it will be no biggy.

    Chill, the stupid will end around mid April cause we can’t keep the country shut any longer than that.

    Also flights from seriously infected countries continue to land at Heathrow.

    More people will die in the Media / Government induced Panic at this rate than of this pandemic.

    Friend of mine’s phone has been buzzing like made with hot news flashes containing yet more speculation.

    Anyone violating social distancing or coughing a bit ‘sus’ is going to be hung unless the mob is soothed a bit and not just with Eucalyptus bog roll…..

  • Nullius in Verba

    “Nullius, that graph appears to combine the death and confirmed case data into one number.”

    Sorry. Deaths are triangles, confirmed cases are circles, both are plotted separately. I should have said.

    “I think you are talking about the UK and that’s more like 12% or 13%.”

    Oops! You’re right!

  • Kevin B

    I’m not an epidemiologist but since everyone gets to speculate on the internet, I’ll have a go at sussing out the difference between German and Italian death rates.

    Viruses mutate quite quickly and they tend to become less lethal, (at least that’s my interpretation). Italy has joined the Silk Road deal with China and had a very high rate of Chinese immigration, particularly to the North so they got the early version of COVID-19 of the type that devastated Wuhan whereas Germany got the later, mutated, version which is less deadly.

    This theory may also be being followed by BoJo and his advisors since I can’t think of any other reason that flights from China and Italy are still landing at Heathrow every day.

    Mind you, he has just closed all the shops. Oh, and threesomes are out.

  • Rudolph Hucker

    I can’t think of any other reason that flights from China and Italy are still landing at Heathrow every day.

    Are they?
    The only reason I can think of is to help build up the herd immunity.

  • Fraser Orr

    Kevin B
    Viruses mutate quite quickly and they tend to become less lethal, (at least that’s my interpretation)

    You might consider the reason for this: lethal viruses, by definition, kill their host, where as sub-lethal let the host walk around and infect others. So consequently the more lethal strains tend to die out with their host and the less lethal ones tend to propagate.

    Viruses don’t care about killing people, they only care about propagating.

    The greenies must love this, surely there is a meme about sustainability in there somewhere.

  • Eric

    On the other hand, German hospitals do not routinely test for the presence of coronavirus in patients who are dying or who have died of other diseases. Italy, by contrast, is performing posthumous coronavirus tests on patients whose deaths might otherwise have been attributed to other causes.

    But in the meantime, I am minded to think that the more accurate picture of Covid-19 comes from the country which has conducted the most tests: Germany.

    I don’t find this very comforting. The implication is the Germans have a better idea how many people are infected, but the Italians are the ones with a better handle on mortality. Remember, this is the same German government which hushed up the Cologne New Year’s assaults.

  • staghounds

    We only outsource vast coercive powers to some of the people who claim to be experts on something. We decide which ones with a popularity contest among non experts. That is how it is supposed to work.

  • Fred the Fourth

    Experts. Yes. The expert bureaucrats at US CDC at work:
    https://mobile.twitter.com/alecstapp/status/1241027712233410560

  • Alsadius

    The obvious reason why flights from infected countries continue is that a) there’s no such thing as a non-infected country (domestic flights are as dangerous as international), and b) there are British citizens in China, Italy, etc. who ought to be able to return home. Not everything is a conspiracy.

  • bobby b

    “Not everything is a conspiracy.”

    They told you to say that.

  • Dyspeptic Curmudgeon

    Nullius: ” then we’re talking half a million dead in the UK over a period of a few weeks just to start. (1% of the whole population is about 6 million.)”

    You had it correct in the first sentence. Maybe you meant 5% of the whole population is 3 million.

    Your analysis is clear and useful. Regarding the statistics floating around, I have seen reports (on the internet so it *must* be true!) that Italy has been testing all deceased for coronavirus and includes all positive test results as a death caused by the virus, without regard for co-morbidity or actual cause of death. To be clear, someone who dies in a car crash but later tested positive, is a covid-caused fatality! Since this a single point in time test, it also includes false positive results, the rate of which cannot be determined by or even inferred from this methodology.

    Germany, on the other hand, is probably only including those deaths which can be considered as actually caused by or attributed to the virus.

    The statistical gap between these positions allows the passage of large trucks carrying toilet paper…. or something. I am reminded of:

    “The government are very keen on amassing statistics. They collect them, add them, raise them to the Nth power, take the cube root and prepare wonderful diagrams. But you must never forget that every one of those figures comes in the first instance from the village watchman who puts down what he damn pleases.” Sir Josiah Stamp, later 1st Baron Stamp

  • lucklucky

    Where is the data to reach conclusions?

    Italian deaths, average age: 79-80 years old
    Italian infected, average age: 63 years old.

  • Duncan S

    From my, now infrequent, visits to news websites, I understand that Nadine Dorries has returned to the House of Commons. Nadine tested positive for Covid-19 (her symptoms weren’t severe, so how did she jump the queue for testing: rhetorical question), spent a couple of weeks at home, and is now back at work. How many ‘ordinary’ people throughout the UK, told by the NHS ‘help-line’ to self-isolate with symptoms and with no access to a confirmatory test, are now actually recovered from covid-19 infection and no risk to others?

    The headline figures of ‘cases’, ‘deaths’ and ‘recoveries’ are based on those won the post-code lottery for testing (MPs and medics) or those who were so sick they presented for medical intervention and testing. In the case of the latter, they were probably already on the stairway to heaven.

    Unless, and until, mass antibody testing takes place, then all the headline figures about ‘cases’, ‘deaths’ and ‘recoveries’ are meaningless and simply fodder for scare-mongering click-bait headlines.

  • Rob

    the Royal Society of Hygiene and Tropical Medicine the other day estimated the number of people in Britain who already have or have had Covid-19 at between 6,000 and 23 million

    What a completely useless estimate.

  • Duncan S

    “The data on Covid-19 differs wildly from country to country. Look at the figures for Italy and Germany. At the time of writing, Italy has 69,176 recorded cases and 6,820 deaths, a rate of 9.9 per cent. Germany has 32,986 cases and 157 deaths, a rate of 0.5 per cent. Do we think that the strain of virus is so different in these nearby countries as to virtually represent different diseases? Or that the populations are so different in their susceptibility to the virus that the death rate can vary more than twentyfold? If not, we ought to suspect systematic error, that the Covid-19 data we are seeing from different countries is not directly comparable.

    Look at other rates: Spain 7.1 per cent, US 1.3 per cent, Switzerland 1.3 per cent, France 4.3 per cent, South Korea 1.3 per cent, Iran 7.8 per cent. We may very well be comparing apples with oranges. Recording cases where there was a positive test for the virus is a very different thing to recording the virus as the main cause of death.

    Early evidence from Iceland, a country with a very strong organisation for wide testing within the population, suggests that as many as 50 per cent of infections are almost completely asymptomatic. Most of the rest are relatively minor. In fact, Iceland’s figures, 648 cases and two attributed deaths, give a death rate of 0.3 per cent. As population testing becomes more widespread elsewhere in the world, we will find a greater and greater proportion of cases where infections have already occurred and caused only mild effects. In fact, as time goes on, this will become generally truer too, because most infections tend to decrease in virulence as an epidemic progresses.

    One pretty clear indicator is death. If a new infection is causing many extra people to die (as opposed to an infection present in people who would have died anyway) then it will cause an increase in the overall death rate. But we have yet to see any statistical evidence for excess deaths, in any part of the world.”

    extract from The evidence on Covid 19 is not as clear as we think Written by a recently-retired Professor of Pathology and NHS consultant pathologist.

  • Nullius in Verba

    “At the time of writing, Italy has 69,176 recorded cases and 6,820 deaths, a rate of 9.9 per cent.”

    And again, the problem is people keep on ignoring the time differences. The deaths now have to be compared to the confirmed, symptomatic cases a week or more ago, and the infections of two or three weeks before that.

    “If a new infection is causing many extra people to die (as opposed to an infection present in people who would have died anyway) then it will cause an increase in the overall death rate. But we have yet to see any statistical evidence for excess deaths, in any part of the world.”

    Four weeks ago, when these people caught it, probably less than one in a hundred or one in a thousand people had the infection, so any impact on death rates is attenuated a hundred-fold or a thousand-fold.

    I agree that the evidence is not clear, but we wouldn’t expect the current case fatality ratio to tell us the final mortality rate, and we wouldn’t necessarily expect to see any impact on the statistics yet. Those are the wrong places to look.

    By the time you find out what you’re actually facing, it’s already a month too late to do anything about it. It may be less serious than we thought. It might be more serious. We don’t know. We will find out in the next few weeks.

  • Snorri Godhi

    And again, the problem is people keep on ignoring the time differences. The deaths now have to be compared to the confirmed, symptomatic cases a week or more ago, and the infections of two or three weeks before that.

    … which last number is not known anyway; and the confirmed cases of a week ago do not necessarily reflect the number of people infected at the time.

    The only safe assumptions are:

    1. The rate of increase of infections of about 3 weeks ago was about the same as the rate of increase of deaths today.

    2. After making adjustments for the age profile of the population and for the speed of increase of deaths, the number of deaths per million inhabitants from the Wuhan virus (heh!) will be about the same as the numbers in the worst affected parts of Italy — and people there have not yet stopped dying.

    3. Point 2 is only valid if all the deaths occur before an antiviral becomes available. This, to me, is the most important reason to “”flatten”” the “”curve””.

    “But we have yet to see any statistical evidence for excess deaths, in any part of the world.”

    Perhaps Professor Lee is not looking hard enough. Perhaps he is not looking at all.

  • Snorri Godhi

    PS: by “safe assumptions” i did not mean to imply that there can be no reasonable doubt about them: i only meant that any reasonable course of action, both for the government and for us individually, should be based on those assumptions.

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>