[bruxis]

> New tests that produce produce results in 10-15 minutes are becoming available, but even with the improved tech, under the current policy a test will only be administered in the most extreme circumstances, when people have had a fever for four days or more.

This is an important factor that may indicate this "outlier" is more of a delay, rather than a reality of some exception.

As a resident, I'm quite concerned about the lack of testing.

[thelittleone]

I spent some time in Japan for work and was always amazed how many people had face masks on (even when there is no pandemic) and the prevalent of wearing gloves (driving etc). With my limited time there it seemed to be one of the most proactively sanitary cultures I’ve seen. Also it seemed to me that at business meetings generally people bow instead of shaking hands. And I wonder if this reflects the general approach to person to person contact. Contrast that to Italy for example where kissing and embracing is not uncommon even when meeting someone for the first time.

[ysw0]

I think this is the case. Japanese (and other East Asians) already have a pre-existing mask culture. Wearing a mask is normal over there and is encouraged. They also do not hug, shake hands and have body contact with does that are not family or close friends. Another big factor is that they tend to follow the rules unlike in Italy or America where social distancing suggestions were not taken seriously by a large chunk of the population.

Even with these explanation, I'm still perplexed as trains in Japan are always packed and it's impossible to stay 6-feet away from others when you're travelling in dense cities in Tokyo. You're also constantly touching poles/handles in the train or pushing elevator buttons.

[inciampati]

The social distancing suggestions (laws) in Italy have been taken extremely seriously. It is effectively voluntary, so there are some people who flaunt the rules. But people went from kissing everyone to never getting <2m from a non-household member in a matter of days. I went from passing many thousands of other people on my daily movements to 4-5.

I think people in Italy feel that the quarantine isn't working because it's not strong enough. And people around the world kind of expect Italians not to take it seriously. Actually, it doesn't work except to decrease the exponent on the growth of the epidemic. In China, with a much stronger lockdown, they had the same R0 as we have in Italy now ~1.25 until they started moving people who were symptomatic or had contact with positive COVID cases to hotels and, if serious, field hospitals and ICUs. We have to do the same or it will sweep through the entire population.

[lbeltrame]

> I think people in Italy feel that the quarantine isn't working because it's not strong enough.

That's also because, in my view, those in power have no idea how to read the curves and the scale, and the fact that you have at least 10-14 days in lag before seeing any effect. Also, this is about growth, so the number of cases (and, unfortunately, deaths) will keep on increasing even with a perfect containment until it "burns out". Again, the powers that be, at least in their public statements, seem to have little to no understanding of this fact.

But they also prefer cherry picking those who don't follow the rules (or go with statements that can't be traced to the data itself).

What is needed most now is field hospitals, staff and protection equipment, because the system, again even assuming a perfect containment, will take at least a month to get relief from the pressure. But those, as opposed to quarantine measures, still go through bureaucracy (note for the non-Italian HN crowd: it's really hard to explain because it is something very peculiar of highly burocratized governments, and in some aspects, unique to my country).

Also this won't be able to go perfectly for more than a couple months at best, IMO. Let's not mention economy, for now, but isolation (despite the "reassuring" messages from the media) will start taking its toll sooner or later. We're at week 3 and for me (in Lombardia) I'm already starting to feel that. Not that I'm going to break quarantine, but if the period extends for too long, some may tempted to do so, with potentially bad effects.

[hkmurakami]

Also generally speaking the culture doesn't have too much "family style" eating habits which helps at the margins.

[aikinai]

There’s way more family style eating in Japan than the US at least.

[tasogare]

You can’t find masks anywhere since more than a month. Few people wear them (I would say about 30%). Granted, people have less physical contact.

The real reason for the low number of cases is the lack of tests. I’ve spoken today with multiple Japanese friends and other residents, all think the government is lying and hiding it to save the Olympics. We will see the truth in a month...

[vasilipupkin]

agree, we will know in a month, but even if govt is lying about cases, we should be seeing a deluge of patients in the ICUs and rise in deaths and we aren't seeing that

[graeme]

How would you know they haven’t? If you don’t test a death for the virus, the death doesn’t show up in virus death counts. And if you aren’t in a hospitals, the deaths wouldn’t show yet.

At some point it would become too big to ignore. But is Japan there yet? This is definitely a space to watch over the next 2-3 weeks.

[qes]

You look at death counts overall and for other reasons. The spikes will show up somewhere, if they exist.

[kevin_thibedeau]

The CDC lumps deaths by pneumonia into its influenza estimates. Presumably epidemiologists would do the same here and just miss the distinction from actual influenza deaths.

[Aeolun]

Given where they could have been if they’d followed the same trajectory as Italy, I’d say the strain should be visible if it was there.

[michaelbuckbee]

30% usage is approximately 500x what I've seen in suburbia US.

[mekster]

I see more like 70% wearing in Tokyo.

[nostromo]

Japan has the oldest population of any country with a median age of almost 50 years old.

I can see the virus spreading silently among the young and healthy. But Seattle and Italy have showed us that it’s very clear when the very old start catching it.

Doesn’t this suggest that Japan truly isn’t experiencing a wave of silent infections?

[tomerico]

Extreme testing policies wouldn’t hide exponential growth - if it is growing exponentially, you’d expect the number of people meeting the criteria to also grow exponentially.

[CobrastanJorji]

It wouldn't hide it FOR LONG. In a month, we'll know which hypothesis was correct.

[standardUser]

A month ago Italy was fine and now 5,000 people are dead. A month ago the virus had been in Japan for a month already... and still today fewer than 50 deaths.

Obviously time will tell with all of this. But so far, time has told us that either a) the transmission rate is extremely low in Japan for some reason or b) no one is dying from the virus in the world's oldest country, for some even harder to explain reason.

The world needs to be looking hard at Japan to see what the hell is going on there.

[lovemenot]

I strongly agree. Yet Japan is rarely mentioned along with Taiwan, HK, S. Korea as exemplars of good management.

I think it's because they've bucked the Test, Test, Test advice yet still had a relatively good outcome so far.

It feels wrong to so ignore Japan. If they are doing something right here, the world should try to find out.

[eternauta3k]

It wouldn't hide it at all. Shifting an exponential curve gives you another exponential curve.

[mantap]

This is only true over the reals, not over integers. People are integers.

[DoofusOfDeath]

People are complex, and often irrational.

[Retric]

It’s unlikely to take that long, in the US our number of dead ~= number of infected 2 weeks ago. Part of that is from in insufficient testing, but 30+% daily increases mean 2,600+% growth per month.

[Retric]

Ops, 2,600x+ growth or 260,00+% growth per month.

[michaf]

If the number of tests administered is the limiting factor, and if you only increase the number of tests linearly in time, then case numbers would grow linear as well.

[corpMaverick]

I suppose that in that case you would see an increase of deaths and also the count of respiratory illnesses(covid-19 missclassified).

[rolph]

yes and this would give you an opportunity to estimate how far off the mark you might be.

[unishark]

Only if all tests are positive already. Otherwise you should see an exponentially increasing number of positives if the number of infected is increasing faster than the number of tests.

[dwild]

> if it is growing exponentially, you’d expect the number of people meeting the criteria to also grow exponentially.

I don't know for Japan, but here in Canada, on top of showing symptoms, we need to have taken an international flight, or be in contact with a known carrier. The number of people having taken a flight won't grow... that's for sure, and contact, well it's recursive at that point and depends on whether you were in contact with someone that has taken a flight, has shown symptoms, has taken the test instead of following the quarantine asked, and has been positive. So essentially in our case, the number of people meeting the criteria can only go down actually.

We would expect the number of people requiring hospitalization with symptoms to grow though. I don't know if that's enough to skip the criteria though, here in Quebec our number of death got lowered yesterday because the test came back negative...

It does makes sense though to keep theses criterias, because most case doesn't require hospitalization, and there's a very limited amount of test available. In Quebec it seems like we are now able to carry enough test for people with theses criterias, so with some luck, we will remove the contact criteria soon, and get a better picture of the propagation.

In South Korea 30% of the infected are in the 20-30 range, it's most probably the same everywhere but we just don't know it because they would most probably be asymptomatics.

[rolph]

the number of cases presenting themselves for testing follows the exponential curve. people are displaying symptoms in numbers following an exponential curve.

when the number of symtomatics become very large and saturate the testing process, the reporting will no longer follow the real curve of cases over time, only the number of test that can be performed over time.

[i meant to reply here | moved from previous reply below]

[Hamuko]

Finnish health officials have said that they do not plan on implementing lower barrier testing. They made a few points about testing.

* Those who can weather corona at home do not benefit from testing.

* Performing the test requires the same protective equipment as treating corona patients and there's already shortages on those worldwide.

* Performing the tests requires some of the same people, who could be actually treating corona patients.

[paulddraper]

And yet, there have only been 36 reported deaths in the entire country.

As long as you test symptomatic cases, the death count should at least be reliable, right?

That a third the deaths of South Korea (who has very aggressively contained).

[mikekchar]

Kind of extrapolating from other countries, I think this points to the number of cases as being about 3-4x higher than reported -- if you are trying to do an apples to apples comparison (which is virtually impossible, but...). Also, I think it's instructive to look at the ratio of people tested to the number of positive results. Wikipedia has about 18k people being tested as of 2 days ago (compared to the nearly 70k people tested in the UK). So the ratio of people infected to the ratio of people tested is fairly close to other places. This indicates to me that the lower numbers of people infected are primarily due to lack of testing.

[paulddraper]

I'm not talking about cases, I'm talking about deaths.

There could plausibly be 10x more cases than reported.

But it's much harder to under-report deaths. And those are low.

[lovemenot]

I don't think your conclusion necessarily follows from your premises.

For instance, if there's a significant difference in testing regime and/or calibration of results.

[ppod]

I think that as long as the availablity of tests is finite, this is a complicated process. Probably the goal should be to optimise for true positives: of the people that have the virus, we want to find as many as possible. In that case, if there is a limitation on the number of tests that can be performed each day, it does make sense to give those tests to the people who are most likely to be positive. Unless some people fit the symptom profile so neatly that they can be treated as though they were positive without a test. Can anyone with expertise weigh in along this line of thinking?

[mattkrause]

I think it's more complicated than that and the people decrying "wasted" negative tests are [redacted] idiots.

If you're at home with what appears to be a relatively mild case of COVID, testing you doesn't tell us much right now[0]. Isolate yourself, even from your cohabitants, and scrub everything like the dickens when you feel better. You should be doing this anyway, even if it's "just" the flu.

Testing people that were around you, however, is critical for preventing the virus from spreading. This is especially true for people are asymptomatic or presymptomatic but nevertheless shed the virus.

[0] This might change in the future as we learn more about if and how long the immunity lasts; perhaps you could go back to work. Hopefully sereological testing will fill that gap soon.

[unishark]

It could simply be that the test will not work until several days have passed. To get a test with such a fast result, trade-offs elsewhere probably need to be made.

Even South Kora required symptoms for free testing. Else it cost about $140 as I recall. That would certainly motivate me to wait a bit to watch symptoms. A fever should mean staying at home, no matter what the cause turns out to be.

[Aeolun]

I don’t think having a fever for 4 days or more is unreasonable? If you have other symptoms of COVID maybe it makes sense to get tested, but a normal flu also causes a fever and is much more likely.