Wednesday, April 1, 2020

Supermodels, Dangerous Curves and Experts - Part 2

Starting at the end, the polling site 538 has done an excellent job in applying some probability - which is their area of expertise - to these wide ranges of predictions by epidemiological experts.

So let's look at what four experts in infectious disease have to say specifically about the models upon which we're making such huge political and economic decisions.
  1. Dr. Deborah Birx, Clinical Immunology, US Army
  2. Dr. Paul Auwaerter, Clinical Director, Infectious Diseases, Johns Hopkins
  3. Dr Eran Bendavid and Dr Jay Bhattacharya, Standford professors of medicine
  4. John P.A. Ioannidis — Stanford professor of epidemiology and population health
1, Because it's worth showing Dr Birx again as she talks about the Imperial College model.

  • So in the model, to get the numbers of infected people predicted, from which other projections of hospitalisations and death tolls are derived, you have to have either:
  • A large group of asymptomatic people who have never presented for any test. That's possible but to determine that in fact, much testing is going on, Yet - "In no country have we seen an attack rate of more than one in a thousand."
  • Or a transmission rate that's very different from what is being seen on the ground.
  • But the predictions of such models don’t match the reality of what they’re seeing on the ground in Italy, South Korea and China.
  • If you did the divisions according to the models, Italy should have 400,000 deaths. They’re not even close to that.
  • "Models are models. There's enough data now of the real experience with the coronavirus on the ground to make these predictions much more sound."
  • "When people start talking about 20 percent of a population getting infected, it is very scary but we don't have data that matches that based on the experience."
  • "There's no reality on the ground where we can see that 60 to 70 percent of Americans are going to get infected in the next eight to 12 weeks".
2.  Dr. Paul Auwaerter talks about the infection / transmission rate
If you have a COVID-19 patient in your household, your risk of developing the infection is about 10%….If you were casually exposed to the virus in the workplace (e.g., you were not locked up in conference room for six hours with someone who was infected [like a hospital]), your chance of infection is about 0.5%
3. Dr Eran Bendavid and Dr Jay Bhattacharya

a) Testing vs infected population
“…First, the test used to identify cases doesn’t catch people who were infected and recovered. Second, testing rates were woefully low for a long time and typically reserved for the severely ill. Together, these facts imply that the confirmed cases are likely orders of magnitude less than the true number of infections. Epidemiological modelers haven’t adequately adapted their estimates to account for these factors. 
b)  The number of US people infected and the US mortality rate
The epidemic started in China sometime in November or December. The first confirmed U.S. cases included a person who traveled from Wuhan on Jan. 15, and it is likely that the virus entered before that: Tens of thousands of people traveled from Wuhan to the U.S. in December. Existing evidence suggests that the virus is highly transmissible and that the number of infections doubles roughly every three days. 
An epidemic seed on Jan. 1 implies that by March 9 about six million people in the U.S. would have been infected. As of March 23, according to the Centers for Disease Control and Prevention, there were 499 Covid-19 deaths in the U.S. If our surmise of six million cases is accurate, that’s a mortality rate of 0.01%, assuming a two week lag between infection and death. This is one-tenth of the flu mortality rate of 0.1%…”
This point is crucial and is much debated. One of the Trump taskforce has argued in the NEJM that the final mortality rate is likely to be between 0.1 and 1.0% - but has also said that the virus could be ten times more deadly than the flu, if it approaches the upper bound of his estimate. Both his arguments could be correct. It should also be noted here that Dr Birx above does not see - from the actual evidence to date - that there could be 6 million infected people in the US as of March 9)

c) The Italian and Chinese actual mortality rates.
“…Fear of Covid-19 is based on its high estimated case fatality rate - 2% to 4% of people with confirmed Covid-19 have died, according to the World Health Organization and others. So if 100 million Americans ultimately get the disease, two million to four million could die. We believe that estimate is deeply flawed. The true fatality rate is the portion of those infected who die, not the deaths from identified positive cases 
“…On or around Jan. 31, several countries sent planes to evacuate citizens from Wuhan, China. When those planes landed, the passengers were tested for Covid-19 and quarantined. After 14 days, the percentage who tested positive was 0.9%. If this was the prevalence in the greater Wuhan area on Jan. 31, then, with a population of about 20 million, greater Wuhan had 178,000 infections, about 30-fold more than the number of reported cases. The fatality rate, then, would be at least 10-fold lower than estimates based on reported cases. 
Next, the northeastern Italian town of Vò, near the provincial capital of Padua. On March 6, all 3,300 people of Vò were tested, and 90 were positive, a prevalence of 2.7%. Applying that prevalence to the whole province (population 955,000), which had 198 reported cases, suggests there were actually 26,000 infections at that time. That’s more than 130-fold the number of actual reported cases. Since Italy’s case fatality rate of 8% is estimated using the confirmed cases, the real fatality rate could in fact be closer to 0.06%…”
4. John P.A. Ioannidis

This was written on March 17 and while testing in many countries has ramped up massively since then his point still stands:
The data collected so far on how many people are infected and how the epidemic is evolving are utterly unreliable. Given the limited testing to date, some deaths and probably the vast majority of infections due to SARS-CoV-2 are being missed. We don’t know if we are failing to capture infections by a factor of three or 300. 
This evidence fiasco creates tremendous uncertainty about the risk of dying from Covid-19. Reported case fatality rates, like the official 3.4% rate from the World Health Organization, cause horror — and are meaningless. Patients who have been tested for SARS-CoV-2 are disproportionately those with severe symptoms and bad outcomes. As most health systems have limited testing capacity, selection bias may even worsen in the near future.
It's quite obvious that whatever decision governments make are going to be rolling the dice. South Korea, Taiwan and Singapore threw resources at testing and tracking which is risky but seems to have worked. Holland has gone for the herd immunity approach: given their enthusiasm for euthanasia that should not come as a surprise. Sweden also seems to be taking a relaxed approach. Given its Federal nature the USA has different approaches being taken by different states. New Zealand and other nations have gone for a fairly hardline lockdown that may burn to the ground a substantial part of the economy.
“Draconian countermeasures have been adopted in many countries. If the pandemic dissipates — either on its own or because of these measures — short-term extreme social distancing and lockdowns may be bearable,”
“How long, though, should measures like these be continued if the pandemic churns across the globe unabated? How can policymakers tell if they are doing more good than harm?”
Well the answer is that they can't. They may never even be able to point to the number of lives saved, and even if they do there will be lives lost because of the lockdown. The phrase "Excess Deaths", calculated at the end of 2020, may be the best measurement.


RosscoWlg said...

Yep Excess deaths..interesting point.

In the UK they estimate, of the death toll, 50,000 per year are due to poor accommodation. That is lacking in adequate insulation or inadequate heating and cooling.

Of course this toll is concentrated in the over 70's

Hardly see a comment in the UK about this, just background data to the press, boring, nothing to see, lets move on!!

Adolf Fiinkensein said...

It's time to get fit and healthy people younger than sixty back to work. Good to see Stephen Joyce is asking questions about how this can be done. Twyford and the defrocked Presbyterian wouldn't have a clue.

And/orsum said...

And just to remind ourselves cure isn't authoritarianism and that the Chinese Communist have lied. Their dodgy numbers and actions have influenced policy.
And if you're still enamoured of Chinese health practices, then recall that they are very bad global citizens because: via their medicines, nearly causing extinction of tigers (for their penises and bones); rhinos for their horns; saiga antelope (cheaper than rhino); turtles and tortoises (also for meat); bears milked for their bile; and sharks ~just taking the fins, to show off wealth

Dave Taylor said...

Enjoy your pieces Tom...and as a long time lurker (with time on their hands) am posting out of frustration that if anything has reached epidemic proportions, it is media-driven idiocy.

Take NZ as of MOH numbers today (

-23,481 tests conducted (....and remember these are predominantly people presenting with symptoms or 'at risk', so not representative of the general population)

- of those tested, 708 positive or possible (3%)

- of those positive, 15 hospitalised (2.1% of positive tests, 0.06% of those 'at risk')

- of those hospitalised, 1 death (7% of hospitalisations, 0.01% of those positive, 0.004% of those 'at risk')

So it seems we are under martial law facing an economic meltdown for a mini-van accident??

Or how about the big, scary US numbers with 188,530 COVID cases and 3889 deaths so far? Well, maybe for some perspective, compare with stats for the 2019-20 flu season just past...

- 36 to 51 million cases
- 370 to 670,000 hospitalisations
- 20 to 55,000 deaths

Obviously just my observations (and like the proverbial) but do wonder what will happen in 6 months when the general population is like "...whoah, what the fuck just happened. Why have I got no job, house, business, retirement funds etc. I thought you said....."

Tom Hunter said...

I'm reminded of the old quote:

The living will envy the dead

Tom Hunter said...

@Dave Taylor

Well be a lurker no longer. As you can see most of the commentators here are a a range of Centre-Left to Far Left so some balance is needed.

Dave the Tailor said...
This comment has been removed by a blog administrator.
Seth Rich said...
This comment has been removed by a blog administrator.
Anonymous said...

@Dave Taylor. if you don't do your sums why are you commenting and dishing out fake news?

80% of flu cases are not recorded, it is not a notifiable disease. This makes your figures, er made up.

Covid-19 is ten times more lethal normal flu. Dr Fauci, who Tom has used for reference also says so..

Victor Meldrew

Tom Hunter said...

Covid-19 is ten times more lethal normal flu. Dr Fauci,

Only if it's at the top of the range he calculated two months ago - 0.1% to 1.0% - and he has also said that he has never seen a pandemic model that actually reached or even got close to its worst-case scenario.

it is not a notifiable disease.
The CDC records it in the US every year. There's a difference between "notifiable", which is a word used around pandemics like this and "reporting" of ordinary, annual flus and other such diseases.

Anonymous said...

Rosco/Tom. The number of people who die in UK from cold homes is about 15,000 per annum NOT your fake 55,000

A post ago you informed the word that I was making stuff up about a 16 year old girl who died in France. Pity I can't show you the photo, your heart would break.

A 13 year old boy has just died in the UK. The virus is indiscriminate and has no regard for wealth, status or now age.

Petri Dish

Anonymous said...

@ Tom..I meant 80% of cases never see a doctor so remain unknown. Also in most of the world doctors are not required to report seasonal flu just as they are not required to report the common cold... wheres Dave Taylor?? Put him on the line.

Victor Meldrew

Anonymous said...

All this hosepipng of uncheckable info is designed to to confuse and disorientate so if you really want to know if lock down is working follow these three countries on worldometer.

UK, France, Netherlands.

UK and France similar populations and demographics..France has been in lockdown about two weeks before the UK got serious. the UK running at about 50% of Frances infections, newly infected and death rate.

If lockdown works the rate of the newly infected in France should start to level off and fall very soon whereas the UK will continue to rise for two to three weeks then fall. The small problem of the UK public not taking it seriously for a week may come into play.

The Netherlands will show whether the herd immunity approach works without an eyewatering death toll. Just draw your own little graph and check each day. Then it means we can do without all this speculation and modeling.

Petri Dish

Tom Hunter said...

hosepipng of uncheckable info....

...designed to to..

confuse and disorientate

Boom tish!

dEVLIN'S dEEP bLUE sEA said...
This comment has been removed by a blog administrator.
Tom Hunter said...

Ooooo - hurty words.

Dave Taylor said...

Whoa one comment and I am already chuckling at the munters you face Tom

@Dave Taylor. if you don't do your sums why are you commenting and dishing out fake news?

The fuck??...anonymous muppet. I nearly can't be arsed, but given I have just got back from half a box of Wednesday beers with the neighbour (social distancing observed of course) I will rise to the bait.

The Ministry of Health is fake now...who knew??

Or maybe the one summarising the CDC data? If so, here are the numbers from the original source:

...that's the Centers for Disease Control and Prevention....a big sciency governmenty thing in the US that studies various illnesses. So, go on fill your boots anonymous medical know-all guy....obviously someone collects the numbers after all.

Does it make you sad to be wrong?...or rather, is the problem some underlying issue you have with the whole reading, comprehension, maths combo. Either way, don't worry, you don't have to suffer alone ( Sorry I haven't got the equivalent Kiev (or is it Shanghai?) contact

Anonymous said...

Wow..hit a nerve there, you sound like a very cross Tom, he's a yank as well but you probably knew that.

This was pre covid. Only about 17% of people see a doctor when they have seasonal flu and this was a small survey and the 17% should not be considered accurate.

"According to Hayward, few people visit the doctor when they have the flu, and even when they do, doctors rarely diagnose them as having influenza."

The CDC's figures are "estimates' and should be taken as accurate. Like Covid-19 flu kills the weakest and those with underlying health issues including advanced age. This is still happening alongside Covid 19.

So if you don't know how many people have flu you can't make any predictions. Doctors in the UK and NZ are not required to report flu although in view of what is happening today that will have changed.

Just a pointer in NZ we don't call it a "box of beer"....good try.

Victor Meldrew

Dave Taylor said...

Little anonymous man....

CDCs figures are derived from "Laboratory confirmed flu activity as reported by clinical laboratories....."

Oh and BTW, is "we" you and your mum?, around here 12 = doz, 15/18 = box, 24 = slab

Tom Hunter said...

@Dave Taylor

The reason for the "we" is that "Victor Meldrew" is simply one of a multitude of names being used by some wanker here as he skips between sock puppets.

Naturally that means he projects his insanity on to others, which is why he called you "Tom". He'll call you a sock puppet as well. I'm afraid this conspiracy theory stuff stems from one half-insane poster we had here months ago that we mistakenly invited to be an author for a while.