An update on a pre-registered result about the coronavirus • Hacker News Board

Re: An update on a pre-registered result about the coronavirus

#31

Post by EE84M3i » Tue, Apr 21, 2020, 2:59 PM UTC

m_mueller wrote: post #27

grosales wrote: post #21

This is just my 2 cents but according to https://www.worldometers.info/coronavirus/#countries Japan has only done 923 tests for 1M of their population. That's much lower than any developed nation. One thing the article claims is the spread of the virus by asymptomatic infected people. Given this low level of testing - it's hard to gauge the true level of infection in the country - although the continuous decrease on…

What I find confusing though is Japan's low death numbers. I was assuming that spikes in mortality couldn't be hidden, but recent comments from inside sources in Japanese hospitals are starting to make me doubt this assumption. Still, there is not a crisis on Lombardia-level going on, which one would expect given the relative inaction of Japanese government. I'm guessing the truth is somewhere in the middle - Japan's…

One theory is that people are dying but it hasn't been discovered yet because of people staying inside. see https://en.wikipedia.org/wiki/Kodokushi

Having just ridden the Yamanote today though, it's unreal how empty it is.

Re: An update on a pre-registered result about the coronavirus

#32

Post by sdffdsfdsfsddsf » Tue, Apr 21, 2020, 3:01 PM UTC

Assuming the ratio of cases with symptoms to cases without symptoms usually remains the same, I must admit I don't immediately understand why it would be important to determine the amount of asymptomatic cases to determine the risk. It seems to me the growth rate of (symptomatic) cases would be the major indicator for that. (Obviously it would be of general interest, but I mean specifically with respect to gauging the risk).

It all seems to hinge on the assumption that with the implemented "strategy A", nobody with pneumonia would be tested for Covid19 unless they have a known connection to a symptomatic case. That of course would be a mistake. But were people really that complacent that it would never occur to them to test for Covid19, even as the rest of the world is talking about nothing else? (Availability of tests is of course a factor here, but should only be a temporary impediment).

I am interested in it, because here in Germany there also is currently a discussion of the virus spreading silently, only to come back in full force in fall if we become complacent. I must admit I don't fully understand the logic behind that, as surely when it spreads silently, it also has a proportion of symptomatic cases in its wake, so that in the end, it doesn't spread silently after all.

Re: An update on a pre-registered result about the coronavirus

#33

Post by abduhl » Tue, Apr 21, 2020, 3:04 PM UTC

barry-cotter wrote: post #30

kspacewalk2 wrote: post #10

Earlier quoted context omitted.

Honestly, who has 1-1.5h to read this without skimming through most of the verbiage to tease out the point? I read the "pre-registered" part. It was wrong on a bunch of points, both numbers and impact. I strongly doubt a major world economy of 126 million people with many world-class epidemiologists and doctors and generally non-disastrous political leadership was 'forced' into anything by this. The hubris of this gu…

> The hubris of this guy is quite off-putting. You’re not a certified expert. Shut up and listen to your betters. If there were actually the reserves of competence you and others like you believe in; if the authorities actually deserved our trust almost everywhere would have done as well as Taiwan or at least Korea. In the world we actually live in the CDC stopped the Seattle Flu Study from testing for COVID-19 and b…

Yes, but do you have any thoughts about the topic we were talking about before your rant?

Re: An update on a pre-registered result about the coronavirus

#34

Post by pavlov » Tue, Apr 21, 2020, 3:06 PM UTC

joebot123 wrote: post #29

simonebrunozzi wrote: post #12

Earlier quoted context omitted.

Perhaps that's on purpose. Perhaps he wants to avoid being sued, or similar things. Patrick has a ton of credibility and I would give him the benefit of the doubt here.

The purpose of the overall evasiveness could be to prevent personal retaliation of some kind - that would be understandable. But why avoid mentioning the word Olympics and instead call it "the sporting event" repeatedly? Especially now when your connection to the white paper is public.

Maybe because the virus and Olympics are both sensitive topics of national pride in a country that has turned increasingly nationalistic.

Japan has a conviction rate of over 99%. If a prosecutor decides you're guilty of something, it's basically a done deal. You don't want to get on the wrong side of authorities in a country like that.

Re: An update on a pre-registered result about the coronavirus

#35

Post by akavi » Tue, Apr 21, 2020, 3:06 PM UTC

ceejayoz wrote: post #20

akavi wrote: post #4

Earlier quoted context omitted.

Yeah, I'm withholding judgement till "the end of April", but given Japan's doubling time is roughly 10 days, you'd expect there to be ~2.5k severe cases right now. Given that Japan's official count is ~11k, is there reason to believe 1/4 of all official cases are "severe"? Does anyone have up-to-date data on "excess deaths" in Japan for the past few months?

> Given that Japan's official count is ~11k, is there reason to believe 1/4 of all official cases are "severe"? Yes. That fits data out thus far. https://www.cdc.gov/mmwr/volumes/69/wr/mm6912e2.htm > Although the majority of reported COVID-19 cases in China were mild (81%) Combine the 19% severe cases number out of China with Japan's aging population (which COVID-19 hits substantially harder), and 1/4 is entirely rea…

Is "mild" vs. "severe" dichotomous? I feel like I've usually seen the trichotomy of "mild" => not needing hospitalization, "moderate" => needing hospitalization, "severe" => needing ICU treatment.

Re: An update on a pre-registered result about the coronavirus

#36

Post by vikramkr » Tue, Apr 21, 2020, 3:07 PM UTC

sdffdsfdsfsddsf wrote: post #32

Assuming the ratio of cases with symptoms to cases without symptoms usually remains the same, I must admit I don't immediately understand why it would be important to determine the amount of asymptomatic cases to determine the risk. It seems to me the growth rate of (symptomatic) cases would be the major indicator for that. (Obviously it would be of general interest, but I mean specifically with respect to gauging th…

The asymptomatic cases can also spread the disease, and a single individual is enough to trigger another outbreak if they slip through the cracks, so for contact tracing and everything you need to know how many asymptomatic cases there might be and where they might be. And the wake might not be spread evenly throughout the population, you could end up with younger individuals being asymptomatic and creating a breeding ground for the virus to spread and mutate in a college town before, even though it seems to be under control, it spreads back to vulnerable populations where everyone is symptomatic. That's one extreme of this but we don't know without random testing how bad it is towards that end

Re: An update on a pre-registered result about the coronavirus

#37

Post by panarky » Tue, Apr 21, 2020, 3:11 PM UTC

kspacewalk2 wrote: post #2

>We project a true count of over 500,000 infections, including more than 5,000 severe cases, and a breakdown in provision of care (“overshoot”) in Nagoya, Osaka, and Tokyo, before the end of April. The current count is ~21,000 with the number of confirmed daily cases declining for 6 days straight, from 741 to 338 yesterday. So... not only wrong on the numbers themselves, but also on the scale, trajectory and impact o…

One might argue that a "true count of infections" is not falsifiable -- people aren't getting tested, causes of deaths aren't properly attributed, officials are juking the stats.

A more falsifiable claim is "a breakdown of provision of care in Nagoya, Osaka, and Tokyo, before the end of April".

If patients are in parking lots or bodies are in refrigerated trucks and mass graves, that evidence is harder to hide.

There are 10 days left in April, but at this point, is there any evidence of "a breakdown in the provision of care"?

Re: An update on a pre-registered result about the coronavirus

#38

Post by smcl » Tue, Apr 21, 2020, 3:12 PM UTC

simonebrunozzi wrote: post #3

Oh my. The Japanese government should feel ashamed to have hidden all of this from the public (edit: I mean the real data and the real gravity of the situation, of course). I, for one, fled the US to come to Japan for a few weeks because I mostly believed these numbers. I hate this. I really hate that politicians (all over the world, in this specific case in Japan) can afford to treat us like this, and usually not fa…

Please tell me this is a joke comment

Re: An update on a pre-registered result about the coronavirus

#39

Post by will4274 » Tue, Apr 21, 2020, 3:12 PM UTC

m_mueller wrote: post #27

grosales wrote: post #21

This is just my 2 cents but according to https://www.worldometers.info/coronavirus/#countries Japan has only done 923 tests for 1M of their population. That's much lower than any developed nation. One thing the article claims is the spread of the virus by asymptomatic infected people. Given this low level of testing - it's hard to gauge the true level of infection in the country - although the continuous decrease on…

What I find confusing though is Japan's low death numbers. I was assuming that spikes in mortality couldn't be hidden, but recent comments from inside sources in Japanese hospitals are starting to make me doubt this assumption. Still, there is not a crisis on Lombardia-level going on, which one would expect given the relative inaction of Japanese government. I'm guessing the truth is somewhere in the middle - Japan's…

I mean, look at the comorbidity data - heart disease, hypertension, diabetes - aren't the Japanese one of the healthiest people in the world?

Re: An update on a pre-registered result about the coronavirus

#40

Post by aikinai » Tue, Apr 21, 2020, 3:13 PM UTC

simonebrunozzi wrote: post #12

9nGQluzmnq3M wrote: post #6

I usually like patio11's writing style, but this is super verbose even by his standards and oddly evasive/defensive to boot. There's a number of important messages here, but they make up perhaps 10% of the content, and I still don't really fully understand the rationale behind the cloak and dagger "hash of report" business.

Perhaps that's on purpose. Perhaps he wants to avoid being sued, or similar things. Patrick has a ton of credibility and I would give him the benefit of the doubt here.

> Patrick has a ton of credibility

Did? I think he just lost all of it with this arrogant stunt. Especially since he himself said,

Either:

1) I am materially wrong about the most consequential thing I've had to have a view on in 15 years. You should probably degrade your estimate of my ability to think through complex problems.

2) We need a data point to counter "Nobody could possibly have seen this coming."

He didn't predict anything "nobody could possibly have seen coming" and he was also mostly wrong, so...