Hacker News Reader

Comment by wollstonecraft | Hacker News Reader

"Test kits" are just for sample collection. You still need throughput on a qPCR machine to get results.

Shouldn't there be hundreds of labs with these machines though?

The United States government spends far more on healthcare than any other nation on Earth. That's not including private health costs.

South Korea has the capacity to test 20,000 per day and has 1/6th the population.

Shouldn't there be hundreds of labs with these machines though?

Yes. Roche just got FDA approval on a SARS-COV-2 test that will take four (4) hours that works with two of their (whatever you call the machine) models. Per Roche there are around 110 of those machines in the United States. Bloomberg had estimated throughput numbers and also indicated that Roche has tests that work with some of their other, slower equipment (that I assume is in the US as well).

Mayo Clinic recently developed a 2 hour test.

The limits on testing are, from my POV, entirely political at this point.

https://www.npr.org/2020/03/13/815522836/u-s-coronavirus-tes...

https://www.bloomberg.com/news/articles/2020-03-13/roche-get...

How does Korea have 15 minute tests (from when a sample is collected) and we're excited about a 2 hour or 4 hour one? I'm not being flippant, I imagine that whatever korea is doing they've been open about it and we know how to do manufacture everything and perform those 15 minute tests, but I imagine there must be some reason that we aren't doing them.

How does Korea have 15 minute tests

They don't. The drive through stations send samples off to a lab that uses a 5–6 hour process. If memory serves, a Hong Kong company devised a 45 minute test otherwise in most places you're looking at hours not minutes.

https://www.bbc.com/news/world-asia-51836898

You're right. I swear I saw some article that said that South Korean drive through would text results as fast as 15 minutes, but looking at a couple more reports just now all said 24 hours.

Also there are blood tests that are dirt cheap and get you the results in 15 mins, but they are only 80% accurate

Every PI at a major research university who does something related to genetics should have one. There are probably some logistics challenges in getting them all into facilities with the right biosafety level and competent technicians to run them.

The most important factor isn't how quickly one can process the test (so long as it doesn't take an excessive amount of time) or how many, but rather how many false positives you get.

The CDC, in a press conference, explained that tests in places like South Korea are producing 3% to 4% false positives. Our tests are in the 1% to 2% range.

This difference in false positives can have massive consequences as infection spreads. Doubling your false positives could easily overwhelm the medical system. Which in turn, could increase fatality.

And then there's the secondary effects, which Italy is sadly experiencing, as the medical system overloads doctors are having to choose who they simply let die. There are reports of stroke patients going without much attention in Italy because of precisely this issue.

If hospitals can't take care of people because they are overloaded due to administering large numbers of tests with double the false positive outcomes, more people die.

When things truly matter there is no substitute for quality and performance. Quantity and speed are not always the most important factors.

does money the US spends on healthcare do things other than line pockets?

This isn't correct. You also need the right primers/probes for the specific virus and controls to ensure accuracy of the test.

Reply on news.ycombinator.com