If you read the BBC article, you'll see that they didn't just rely on local sources, but interviewed experts from around the world.

Yeah, I read that. Some of the worst statistical proofs I've seen have been written by economists who know something about statistics but don't understand epidemiology, and are advocates trying to win an argument. This guy, a biostatistician, can explain the problems with Wyner better than I can:

https://liorpachter.wordpress....
Bits of DNA
A note on âoeHow the Gaza Ministry of Health Fakes Casualty Numbersâ
March 8, 2024
by Lior Pachter

In a Tablet Magazine article titled âoeHow the Gaza Ministry of Health Fakes Casualty Numbersâ posted on March 6, 2024, Professor of Statistics and Data Science Abraham Wyner from the Wharton School at the University of Pennsylvania argues that statistical analysis of the casualty numbers reported by the Gaza Ministry of Health is âoehighly suggestive that a process unconnected or loosely connected to reality was used to report the numbersâ....

The BBC did a good job of fact-checking and talking to experts about the Hamas and Israel death statistics. The people who study and collect wartime casualty statistics thought that their numbers were pretty accurate, although probably understated.

BBC, Checking Israel's claim to have killed 10,000 Hamas fighters, 29 February 2024
https://www.bbc.com/news/world...
Israel claimed to have killed 10,000 Hamas fighters, but has no way to distinguish between Hamas fighters and ordinary civilians. Also, according to Haaretz and +972, whose reporters are in regular contact with IDF soldiers and commanders, the IDF counts all male deaths as Hamas fighters. On the ground, the IDF has free-fire zones, or "Death Zones," in which they assume that everyone in the zone is a terrorist, and can kill them -- men, women, and children. That includes 6-year-olds like Hind Rajab.

BBC, How the dead are counted in Gaza.
https://www.bbc.com/news/world...

Gandalf's horse was Shadowfax, not Equifax.

Until Apple or Samsung put their efforts into AR, I don't see anyone else being able to go to market with a VR solution. Even Google tried and failed with the Google Glass a decade ago. While I love the idea, I don't think the market is there for this yet (and might not be for a long time outside of large cities).

I'm pretty sure HIPAA laws (in the US, anyway) would prevent such activity, for the same reason(s) that California deemed it acceptable to retract their requirement that HIV-positive people disclose their situation with potential partners.

Germany seems to be dong pretty well.

https://91-divoc.com/pages/cov...
COVID-19 Cases by Country

Active cases, and new cases per day, have been declining since 6 April.

Here's an authoritative report on Germany.

https://www.microbe.tv/twiv/tw...
TWiV 601: Das coronavirus with Christian Drosten
April 14, 2020
Coronavirus expert Christian Drosten joins Vincent to provide a view from Germany on COVID-19 and SARS-CoV-2.

Here's a video from a Youtube channel called Real Engineering, produced by Brian McManus, a biomedical and aerospace engineer in Galway, Ireland, who worked for Medtronic. Good explanation of ventilator design.

https://www.youtube.com/watch?...
A Guide To Designing Low-Cost Ventilators for COVID-19
Real Engineering
Apr 4, 2020

He's critiquing the cheap ventilators that people are proposing, including this MIT design. https://scitechdaily.com/mit-p...

(This is my understanding of what he's saying. People with more expertise and time than I have are welcome to add corrections.)

The problems with the MIT design are that (1) it doesn't have a cycle that responds to the patient's attempts to breathe (2) it doesn't have positive end expiratory pressure (PEEP). This is what those advanced computerized controls are for.

(1) The patient breathes in a normal cycle. If the ventilator ignores that cycle, the patient with try to fight it, and it will be very uncomfortable. To overcome this, they have to sedate the patient, which has anesthesia risks of its own, and after a week of sedation, it can be difficult to get the patient's breathing cycle back again. Some people never come off the ventilator. To prevent this, commercial ventilators can sense when the patient is trying to breathe, and synchronize their cycle to assist the patient's own breathing cycle, rather than fight against it.

(2) The windpipe divides repeatedly like a tree, until it ends in alveoli, the little sacs that transfer oxygen to the blood. In covid-19 pneumonia, the virus infects these sacs, they become inflamed, fill with fluid, and can't function. If, on the exhale cycle, the pressure on the alveoli goes too low, the healthy alveloi will collapse. To prevent this, the ventilator must maintain enough pressure (PEEP) at the end of the cycle to prevent the alveoli from collapsing.

The MIT article (an MIT press release) says, "it is crucial for such a system to not damage the bag." Well, it's even more crucial that the system not damage the alveoli.

Once a patient goes on artificial ventilation, their prospects aren't too good. (The grim reality about covid-19 is that once a patient develops pneumonia, their chances of survival even with a ventilator are around 30-50% https://www.npr.org/sections/h...)

That SciTech Daily article cites a paper, Design and Prototyping of a Low-cost Portable Mechanical Ventilator https://www.researchgate.net/p... from 2010. Ten years is a long time in respiratory therapy. They've done a lot of studies to improve survival since then; for example, they found that patients in a prone position had higher survival than patients in a supine position (70% rather than 50% in one paper, as I recall).

The MIT paper says in their abstract, "Future iterations of the device will include a controllable inspiration to expiration time ratio, a pressure relief valve, PEEP capabilities and an LCD screen." So they don't have PEEP.

McManus said at the end that there was an Indian design that avoided these problems, and was run by an android phone. Because of the shortage of ventilators, the Indian government doesn't allow them to be exported, but the Indian designer wants to license them out.

From an MD perspective, here's a video with a good explanation of pneumonia, particularly covid-19, and how it's treated. Nice explanations of x-rays.

https://www.youtube.com/watch?...
3 COVID-19 Cases As Described By Doctors In China
Feb 23, 2020
Chubbyemu

Particularly good collection of articles about artificial ventilation in the references:

Ventilation with Lower Tidal Volumes as Compared with Traditional Tidal Volumes for Acute Lung Injury and the Acute Respiratory Distress Syndrome. N Engl J Med 2000; 342:1301-1308

Neuromuscular Blockers in Early Acute Respiratory Distress Syndrome. N Engl J Med 2010; 363:1107-1116

Prone Positioning in Severe Acute Respiratory Distress Syndrome. N Engl J Med 2013; 368:2159-2168

First Case of 2019 Novel Coronavirus in the United States. DOI: 10.1056/NEJMoa2001191

https://www.msn.com/en-us/news...

It's a pretty short story, not one of those leaders the WSJ was famous for. I don't understand how this could require 5 reporters.

It doesn't answer the question that most of the people who follow this (like Terry Gross) are asking:

Who exactly made the decision not to use the WHO/German test after they had trouble with their own test?

Was it Azar?

Were these decisions made by the civil service staff, or by the Trump appointees?

There are different and distinct uses for tablets and laptops. Until we get full power-user options for tablets such as shell access and the ability to install arbitrary files, its hard to do all workflows through a tablet. Even being able to ssh into a remote service and do tests, its still not completely viable for many of my workflows. The one advantage the Microsoft Surface has is that its a full Win10 operating system on standard processors, giving me a lot of flexibility without having to root the device.

I'd love to see VSCode or similar programs take this on as a challenge and provide the ability for me to work with my git repos, and provide some remote shells to test my code in. With Azure its all within their ecosystem if they are willing to spend the time / energy.

Beautiful video of the Roche cobas 6800/8800 Systems.

https://diagnostics.roche.com/...

DNA testing kits have 2 parts. (1) First you get a nose or throat sample with a Q-tip and put it in a tube. (2) Then you extract the DNA and see whether it contains the sequence you're looking for.

The cobas 6800/8800 Systems is the second part.

The press release is here.

https://www.roche.com/media/re...

They've really automated that stuff since I did it in a biology lab.

They should rate these machines by the number of biology graduate students they can replace, like horsepower. "This machine can do the work of 1,000 graduate students."

You know that Roche is a Swiss company, right?