Resources For Novel Coronavirus Covid-19 | Hacker News Reader

Resources For Novel Coronavirus Covid-19

community.wolfram.com

90 pointstosh37 comments

Anyone that wants to stay up to date with the virus should follow this channel [0]

It's from the semi retired Dr. John Campbell, his videos are low tech but high content. I've come to realize it is all our responsibility to make sure the health care professionals can handle this.

If we assume this is going to be as big as some predict it will be, we basically cannot escape it, but it is possible to lengthen the time in which most infections well arise, lowering the peak and curve, which will in turn greatly reduce the peak load on healthcare professionals.

A 1000 admissions a day might be impossible to handle, but 100 admissions stretched out over 10 days is much more manageable. This is something everyone can help with.

Also, take a lot of Vitamin D, it helps with handling the Covid19 and other respiratory tract infections (both viral and bacterial). [1]

[0] https://www.youtube.com/channel/UCF9IOB2TExg3QIBupFtBDxg [1] https://www.youtube.com/watch?v=W5yVGmfivAk

> take a lot of Vitamin D

In France you can only get pills dosed at a very small 800 IU (international unit, see: http://www.nafwa.org/vitamind.php) without prescription, but in the US, you'll find some bottles dosing a single pill up to 10 000 IU.

The pills look all the same no matter the dose. They are minuscule yellow balls, but they can be very concentrated.

So "take a lot of Vitamin D" doesn't mean "take a lot of pills". Target an amount of IU, and take exactly what you need for that.

Vitamin D has been found toxic if you take more than 60 000 IU/day for 3 month. You'd have to be quite heavy in your dosing to reach this, but I'd rather share the info.

There is no overdose possible if you get it from the natural source: exposition from the sun. And it's healthier. But as a geek you may see the light of your monitor much more than the one from the sun, and D supplements are incredibly cheap for the huge benefit in energy and mood they provide in the winter.

I take 5 000 UI/day (as Cholecalciferol, so D3), and my doctor already thinks it's a lot since the daily recommended dose here is 200 IU.

Be careful about vitamins in the US though because the industry is entirely unregulated and many vitamin companies are pyramid schemes, so you have very little guarantee that what’s on the bottle is in the pills.

Naturemade is the only brand that comes to mind with external lab certification of ingredients and being free from contamination, but I’m sure there are others:

https://www.consumerreports.org/supplements/how-to-choose-su...

"Daily oral dosing of vitamin D3 using 5000 TO 50,000 international units a day in long-term hospitalized patients: Insights from a seven year experience."

https://www.ncbi.nlm.nih.gov/pubmed/30611908

> In summary, long-term supplementation with vitamin D3 in doses ranging from 5000 to 50,000 IUs/day appears to be safe.

seems like SV is going from micro-dosing acid to micro-dosing vitamin D :)

Who said I stopped acid? :)

Ireland just closed all schools and creches until the end of March. Indoor public gatherings limited to 100 people, outdoor to 250. Hope fully this will slow it enough to allow the health services to cope.

How much is a “lot of Vitamin D” for this particular goal?

The "Epidemic Data for COVID-19 (World)" data is showing 28 deaths for Germany, which doesn't reflect the WHO or Robert Koth Institut data, which show only 3...

https://www.who.int/docs/default-source/coronaviruse/situati...

https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus...

It's the same for Norway. This website report 400 cases, 4 deaths and 22 recoveries. These are quite far off from the actual official numbers. The official numbers are released once per day at 18:00, by the Norwegian institute of Public Health [0], and VG (a newspaper) are keeping the numbers updated [1] with reports coming from municipalities and hospitals around the country [1]. The last official numbers (from yesterday) were 489 cases, 0 deaths, 1 recovery. The last numbers from VG are 642 cases, 0 deaths, 1 recovery.

0: https://www.fhi.no/

1: https://www.vg.no/spesial/2020/corona-viruset/

One of the relevant data sources listed is "Mapping 2019-nCoV" https://systems.jhu.edu/research/public-health/ncov/ , and that shows Germany as having 3 deaths and 25 recovered cases. Perhaps these are getting falsely aggregated to give 28?

Ok, that makes this website completely worthless at the moment.

https://www.arcgis.com/apps/opsdashboard/index.html#/bda7594... seems up to date.

While the site and dataset is good, that particular link doesn't work well on mobile Android Chrome. This one works a treat https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.h...

The data source for Wolfram is the John Hopkins University dataset. Here, there are listed 25 recovered cases for Germany. The erroneous number 28 might come from the calculation 25 + 3.

So the dataset is good, the calculation rubbish. Says a lot about the site. Why don't we just stick to the WHO, and the official last cal sources (CDC, Johns Hopkins, Robert Koch, whatever)?

From Europe, that reminds me a lot of Chernobyl time in 1986:

- Everybody speaks about it.

- Nobody has seen it yet (at least where the people I know live).

- The invisible threat seems to be everywhere.

- Everybody is changing its basic behavior (now: often wash hands, do not shake hands or kiss, avoid crowds, think twice before visiting elderly... then: avoid sandboxes and playgrounds for children, avoid rain, avoid certain foods like mushrooms or salads...).

In a few words, it is a deadly shadow spreading on the world... it shakes the world and it will pass.

a lot of this data are tools are utterly pointless unless you have also transparency over how many tests are being done each day.

garbage-in/garbage-out

Not only how many, but also how they are sampled. A random sample across the entire population will, likely, discover fewer cases than a sample of hospital personnel and that, fewer than a sample of patients with pneumonia.

A US specific question - Does anyone know of a dashboard where you can see the trend of the Covid-19 positive cases by state? I would find it helpful to see if my state is trending up/flat/down. For example, in MA the cases had been doubling by day, except Yesterday they only jumped from 91 to 95. It would be good to see the trend for MA as well as for other states.

Unfortunately the availability of tests is so limited we really have no idea of how many cases there are in any given state at the moment.

That said, the CDC publishes this data on a state basis daily here: https://www.cdc.gov/coronavirus/2019-ncov/cases-in-us.html

It could probably be pretty readily inferred from "pneumonia and flu cases in which medical aid has been sought" data by extracting the normal seasonal trend from the current data.

In the UK, hospitals/local health centres track this data (and deaths are tracked down to quite a fine resolution, differentiating types of flu, for example) and presumably they're putting it on their networked computers so in theory the data is there in real time.

The gov.uk site "Number of coronavirus (COVID-19) cases and risk in the UK" has now been updated -- to remove any localisation of data.

I suppose they think "if people can't easily access the data they can't know how bad things are", which when you need to convince people of how bad things are in order to control an epidemic seems like a "curious" move.

Someone is going to say "but you can drill down on this other site and get figures for England" ... to which I say, yes, but gov.uk is for the UK as a whole and is the authoritative website. I wish they'd just admit no-one is publishing infection spread data for Wales.

Is this political? Well (non-Tory) Scotland seem to be able to keep their data up front and accessible.

Different numbers are being reported from this site [0]

Which would be a more definitive source?

[0] https://www.worldometers.info/coronavirus/#countries

I find that BNO [0] is the fastest in reporting new cases (both in aggregate and in detail) and gives sources so you can see what region of the country is affected.

[0] https://bnonews.com/index.php/2020/02/the-latest-coronavirus....

Their map is crap though. They indicate the southern US as nearly free of cases. I had initially been viewing it alone, thinking things seemed remarkably fine. Data seems accurate, but don't trust their map.

They used to be. In the last few days, they started falling behind (and as others said, the map is very out of date).

However at least for Germany the embedded map is very outdated, possibly hasn't been updated for a week or more?

More than a week. Confirmed cases are 30 in Florida, presently. But strangely, the map historically indicated much more, then suddenly removed many cases.

One reason more to never look at secondary sources. Stick to the primary ones.

Nowadays every "data scientist" can pull a data set from the internet and toy around with it. Build a map and website and put it online. Without review of neither the results nor the source material. Share it on Twitter, reddit, Facebook and the numbers are out there. Unverified and impossible to take back.

I really wished people would stop doing that. Play around for yourself if you want, but keep it to yourself.

Why does speed matter in that regard? Honestly asking. Because I prefer confirmed and reviewed numbers, quality, over speed here. Daily is more than enough for me. I just wished the WHO situation reports had more depth to the data. But I am not necessarily the target audience of those.

Speed matters to a degree. A lot of resources around, including the ones linked in the Wolfram documents, are outdated by number of days or do not disclose when was the series for a region updated last. For instance, as I am writing this on March 12, Wolfram says last update is March 11. It lists Sweden with 355 cases. That number is actually from March 10 and the actual value should be 500 for March 11 and 565 as of now. It also does not give you sources for changes so you don't get an idea of where these cases are. Both BNO and worldometers give these sources. To me personally it is important to know what the growth is in my neighbourhood rather than in the whole country.

For the same reason, I am not looking at the map so I do not care that much that it is outdated. It conveys zero information about the geographic distribution of new cases/deaths within most countries. I am yet to see a resource for the whole world that is as detailed as what China/SK/SG have.

And before I am told that I should just look at local news. I am an expat with family and friends scattered across Europe. I want to know what's the risk for them in one place not 5.

Wolframvirus

would be interested in your thoughts on how covid will change the election process in the US. you spent some time in the trenches, so I really, really would love to see a write-up (speculative or not idc :))

The most likely outcome in my view is that a number of elderly legislators will die before the election. Even last winter, before the virus hit, the actuarial chances of everyone in the Senate living until November were down at around 14%.

Much of Congress and the men running for President are in a very vulnerable age group, and the nature of politics makes it hard to self-isolate.

https://docs.google.com/spreadsheets/d/1gWVuTPWLXZH3nB5oVMAm...

appreciated, thanks for the link. covid19 will be a real wildcard adding plenty of volatility then.

https://www.youtube.com/watch?v=c6eSPWaUsgY

TIL the onion had a youtube channel. this is gold thank you!

Reply on news.ycombinator.com