Item — HN Reader

Is there any particular reason the US is testing so few people? They've only tested 445 people in the entire country so far[0], which is a tiny number for such a large country (contrast e.g. with the much smaller UK which has tested 7,690 people[1])? Wasn't that one of the lessons from Italy - the numbers of confirmed cases seemed to shoot up from 21 Feb mainly because that's when they started testing.

[0] https://www.cdc.gov/coronavirus/2019-ncov/cases-in-us.html

[1] https://www.gov.uk/guidance/coronavirus-covid-19-information...

8 comments

The test-kit that the CDC sent to local labs had a faulty reagent, so all test requests have to be sent back to the CDC.[0]

Local labs could develop or purchase their own test kits, but during an a public health emergency, they must get FDA authorization before launching such a test.[1]

This seems to me like a giant institutional and bureaucratic screw up.

Hopefully the FDA may be finally getting its act together and fast-tracking approval of local tests: https://twitter.com/ScottGottliebMD/status/12330216819181936...

[0] https://www.politico.com/news/2020/02/20/cdc-coronavirus-116...

[1] https://threadreaderapp.com/thread/1231944326827081729.html

In all seriousness, why should a non-invasive diagnostic test require any sort of approval (particularly during an outbreak)? This is just RT-PCR if I understand correctly - there's nothing particularly novel going on here. Just publish the primer sequences and let the biotech industry handle things.

Edit: Oh hey they did publish an RT-PCR protocol [1] plus sequences, [2] along with a disclaimer not to use them directly on human subjects. This is just silly (IMO).

> These procedures and/or reagents derived thereof are intended to be used for the purposes of respiratory virus surveillance and research. The procedures and reagents derived thereof may not be used directly in human subjects.

[1] https://www.cdc.gov/coronavirus/2019-ncov/lab/rt-pcr-detecti...

[2] https://www.cdc.gov/coronavirus/2019-ncov/lab/rt-pcr-panel-p...

My understanding is that it’s based on a law from several decades ago, when it was difficult to get your hands on a sequencer, that was meant to help empower the CDC once some kind of pandemic was declared in order to get tests out faster. Now that sequencers are commonplace, the same law instead of speeding up the deployment of these tests now slows them down.

This is something the media should be discussing more, to motivate legislators to fix ASAP.

>FDA regulates test kits but generally lab developed tests, which are designed and used in single lab, can be offered without FDA review. When HHS declares Public Health Emergency and issues declaration to support EUAs, labs must seek FDA authorization before launching new test.

Normally they can be done without approval, but there is a special regulation on the books that activates when the department of health and human services declares a public health emergency.

The natural question is then: why haven't the people in charge issued some sort of exemption or blanket authorization in this case?

Well, the result of the diagnostic test could cause harm (panic, unnecessary treatment, people not quarantined when they should be) if it was not accurate enough, but they should probably fast track something given the circumstances.

You know what causes greater panic? An outbreak that only to the attention of the media when corpses start piling up at the hospital, all because an undetected cluster wasn't identified when it was starting.

> In all seriousness, why should a non-invasive diagnostic test require any sort of approval (particularly during an outbreak)?

To make sure it's a reliable test with known error bars to avoid false positives and negatives. If your data is garbage who cares how quickly you get it?

^^ This is exactly the answer! The positive predictive value of a test changes based on the prevalence of the disease in the population. With rare conditions, you can end up with way more false positives than actual positives, which can cause actual harm to people.

"Outbreak" and "Contagion" both depict mass outbreaks of deadly viruses, and the social impacts of implementing last-ditch control efforts. There's a certain point where panic about an epidemic is more damaging than the epidemic itself, and I worry we're approaching that point with coronavirus. When everybody is freaking out is exactly when we need to enforce good public health practice, not throw it by the wayside and go full wild west.

when you conduct the assay you are handling a known pathogen

doing that without proper training equipment and expertise should be discouraged, even if it isnt the original intent of the restriction it does seem to reduce a possible avenue of amplification of the problem

The thing is, you're already dealing with a patient that might be infected and medical testing labs already process potentially infectious samples on a daily basis.

The CDC has actually issued official guidance regarding Coronavirus biosafety. [1] Other than discouraging unnecessary culturing of the virus, BSL-2 (a fairly common setup) is the main recommendation.

[1] https://www.cdc.gov/coronavirus/2019-ncov/lab/biosafety-faqs...

Maybe false negatives or even false positives could be a liability issue?

Type-1 and Type-2 errors... :-)

https://en.wikipedia.org/wiki/Type_I_and_type_II_errors

Medical testing is very, very murky. There was a time a routine medical test for me came back, indicating I possibly had Lupus. It turns out; this test has a false positive rate of about 5%.

I am a white male, and the rate for lupus for my group is astronomically low. The rate for the worst group (black females) is like 500 in 100,000. There's some evidence the rate around white males might be six times lower.

Had I known any of this at the time, or had my doctor explained it, I wouldn't have spent weeks worrying.

The fact is, medical tests just update the probability you're sick or well. And this is why they have to be well understood.

From reports, the Chinese tests have fairly high false negatives, but they’re still deploying them en masse because catching as much as is possible better than not catching anything at all. I’ve seen nothing on false positives.

If you're using an officially published primer sequence? Possible (due to supplier or user error), but _highly_ unlikely. We're pretty good at molecular biology at this point.

>a disclaimer not to use them directly on human subjects. This is just silly (IMO).

this is because if you truly understand the materials present in the assay kit, you can, with a well stocked genetics laboratory, begin recombinant proceedures

That... doesn't make any sense. What does the ability to run recombinant procedures (however ill advised doing so might be) have to do with running an assay on a human sample for diagnostic (as opposed to research) purposes?

To be perfectly clear:

* The CDC has already made an RT-PCR protocol plus associated primer and probe sequences publicly available.

* BSL-2 labs that do cell culture and employ viral vectors for transfection [1] are quite common in academia.

[1] https://www.thermofisher.com/us/en/home/references/gibco-cel...

Is it not enough to have the complete genome? https://www.ncbi.nlm.nih.gov/nuccore/NC_045512

Are you saying that if i had the test for corona, and then a bunch of knowhow and some wizbang machines I could start making the virus? I was always terrible at biology.

  why should a non-invasive diagnostic test require any sort of approval?

Because false negatives are worse than no test at all / unknown.

100 false positives for every actual positive is also pretty bad.

because of the original screw up went vaccines were first invented during he 1950s, people died..

Reagent

There's no reason to test if there aren't any confirmed cases. You can't confirm a case unless you test...
And thus, how we find it now is bursting clusters with illnesses that can't be hidden by "flu." By then it's too late.
Seems like the only way to contain it is do blanket nationwide testing every day and quickly contain anyone who tests positive, for like 6 months.
That's why it will keep spreading... Dammed is you do, Dammed if you don't..
This isn’t true. Containing the spread through mass testing would allow us to more appropriately measure the response and direct resources accordingly. Every epidemiologist I’ve read says the lack of testing is a very concerning in terms of our ability to address and respond to the pandemic.
>There's no reason to test if there aren't any confirmed cases.
There's no reason to test when there isn't a high confidence that the person has the virus. Especially when the test costs something north of $3,000 and even with good insurance only a fraction of it is being paid for based on numerous Reddit threads and a news article [0] puts a specific patient's bill at:
>$3,270 two weeks after his test. He will be responsible for $1,400 of that bill.
A Forbes article [1] two weeks ago even states:
>The cost of testing include the costs of test kits, labor, and sample processing. In some cases, storage and delivery costs would need to be added. It’s unclear what the costs of the test kits are, or ancillary items, as CDC and others haven’t revealed figures. For the sake of argument, let’s say that on the low end of the scale each test, including ancillary items such as labor and sample processing, costs $250, and at the high end it’s $1,500.
[0] - https://www.independent.co.uk/news/world/americas/coronaviru...
[1] - https://www.forbes.com/sites/joshuacohen/2020/02/12/testing-...
You have to balance that cost against the cost of not containing this outbreak, which is likely to be far, far more than that, not just in money but in lives.
It's also critically important to act quickly, as it's at the early stages when this outbreak has a chance to be stopped. If we keep waiting until it's obvious that there's a large outbreak in the US it may be way too late.
You don't have to test everyone in the US, or even anyone who has a cough, but everyone with severe symptoms of the disease should have been tested, regardless of whether they've recently been to China or had contacts with people who had.
Never a more frightening instance of "penny wise, pound foolish".
"Oh heavens no, we can't afford to preventatively test for a virulent new lethal pathogen that has already brought one of the world's largest economies to a total standstill - it costs literally dozens of dollars!"
The patient who pays for the test gains nothing from the results.
The country gains from the results, but isn't offering to pay.
If I had been to China and got ill when returning to the US, I would not be getting tested - the test wouldn't help my chance of survival, and would just cost me money and time.

The latest word from the CDC is that they’ve run 3,625 tests and that the testing capacity should double soon.
Yeah, I think the actual answer is that the number of tests on that page is not even remotely up to date, and the CDC has more important issues on its plate than updating it. As I understand it the UK testing criteria are very similar to the CDC's and also require travel to an affected country or contact with a confirmed case: https://www.gov.uk/government/publications/wuhan-novel-coron... The government here is just starting to roll out more general testing of the population to selected cities, but so is the CDC over in the US, and in both cases it's intended more to detect widespread spreading than to catch every individual case.
Why would it not be in CDC's interest to update the number of tests?

A counterpoint - testing like this takes resources, and creates a backlog for other more pressing tests for patients. I mentioned this in another comment, but if I have a patient with pneumonia, I'm more than likely going to treat it the same way regardless of what the test shows me. Getting samples from everyone takes resources that lots of hospitals don't really have.
Maybe you could get a backlog going that can test for COVID-19, when the lab's RT-PCR machines are underutilized, even after the patient leaves, so that we can get as much data as we can.
But wouldn't you need to isolate pneumonia patient if they confirmed to have COVID-19? Also, curious, would you be more inclined to use some of the medicine that other countries claimed to be effective against COVID-19? Like Remdesivir and chloroquine?
In addition, at least until the outbreak becomes an uncontrolled epidemic in the US, you'd want to trace the contacts of any person confirmed to be infected with COVID-19 in order to try to try to stop the spread of the disease.
How do we know that testing for Coronavirus creates a backlog? I don't have enough information to say how the system works, but I imagine a nurse takes an additional blood test, and then someone sends that test to the CDC. The CDC tests the blood, and sends a response to the hospital.
I am curious to know more about the actual process.
I believe a test takes 6 hours, there is already a several day backlog, and the samples are sent via FedEx.

There's a bureaucratic 'Catch-22' in each regulatory agency's rules for allowing emergency testing, & apparently no one willing to exercise the authority to break the logjam. See for example the unwound details in this thread (which is reporting roughly the same things as Dr. Gottlieb, just without the diplomatic mollycoddling of his former agency):
https://twitter.com/random_eddie/status/1232075474966368256

Yes. The tests are sufficiently unreliable that if there is no specific reason to believe that a particular person has been exposed, the results are just noise.
See https://towardsdatascience.com/statistics-and-unreliable-tes... for an analysis using Bayesian statistics that demonstrates why you would pre-filter who you give the test to, and why you shouldn't necessarily trust the test results after.
Please remember that public health organizations have over 150 years of experience with how to make the most effective use of existing resources to limit health challenges. This is a complicated optimization problem and the people doing it aren't stupid. It is better to ask why they do what they do than to assume that what they are doing makes no sense.
The problem in such cases is the prior. If a false positive is 10%, 1%, 0.1% or 0.01% makes a huge difference. With 0.01% and two different tests, you will get a false positive in every 100'000'000 tests. Testing twice solves many of the issues. It also gives information about the sensitivity and specificity of a test.
> With 0.01% and two different tests, you will get a false positive in every 100'000'000 tests.
That doesn't sound right to me (i.e. only the case if false positive distributions are independent and uncorrelated, which seems unlikely), but I don't know enough about statistics and epidemiology to dispute it.

We messed up a bunch of the first ones and we don’t really have centralized capacity to manufacture the tests at scale because of our privatized, outsourced healthcare system, we’re being prevented from using overseas ones (the senator from Hawaii is publicly begging Japan for tests), the CDC and other responsible agencies are in disarray from GOP gutting and malfeasance.
Take your pick.

If tested I have to assume the patient will be charged for it, which depending on their insurance could easily be a cost of thousands of dollars if they are in a hospital or ER. We are going to experience a situation where sick people don't get tested and continue to infect others because health care is too expensive. The cost of health care is very much going to exacerbate this situation.
Administration officials just refused to commit to making a tax payer funded vaccine affordable.
This is what end game capitalism in health care looks like.