Saliva based Covid test: results in 45 minutes | Hacker News Reader

Saliva based Covid test: results in 45 minutes

colorado.edu

218 pointsgoldenshale78 comments

Technically, you are really just waiting for the thermal inertia of the testing device. The reaction kinetics are really quite fast.

During the last SARS outbreak, this sort of thing was proposed: https://pubs.rsc.org/en/Content/ArticleLanding/2011/AN/c1an1...

... to do 3 minute tests. I assume grant funding dried up, since who could guess that after 2 outbreaks that we'd have a 3rd?

the method described in the article uses a constant temperature of reaction.

edited to add: https://en.wikipedia.org/wiki/Reverse_Transcription_Loop-med...

> I assume grant funding dried up, since who could guess that after 2 outbreaks that we'd have a 3rd?

Yes, that's what I've read. Any US research scientists still working on SARS-1 after 2003 were told "you're on your own."

Kudos to Canada for being as diligent as Taiwan in attacking COVID-19 - they remember it wiping out Toronto hospital staff in 2003. Check out the Youtube CBC channel for their impressive nation-wide mobilization.

Makes you wonder wtf Chinese virology labs were doing for 20 years with coronavirus if they weren't working on a cure, doesn't it? (There's 40 labs total across China.)

> Makes you wonder wtf Chinese virology labs were doing for 20 years with coronavirus if they weren't working on a cure, doesn't it? (There's 40 labs total across China.)

Like most things in a command economy, they worked on whatever was a political priority.

I'd hazard to guess it wasn't a priority in the intervening years for the same reason there was little free market incentive to address it: yes, it's bad when it happens, but it doesn't happen that often.

Ergo, they were probably just as starved of resources.

The preexisting US effort was mostly funded through DARPA for the same reasons. https://www.darpa.mil/our-research (e.g. ADEPT, P3, Prometheus, PREPARE)

Off topic, but related enough that I feel its relevant to mention:

I learned this week that many blood donation and plasma centers [1][2] are doing Coronavirus antibody tests for anyone* that donates right now.

I've heard a lot of people in my social / online circles comment how they wish they knew whether or not they had antibodies. Seems like a great way to find out.

[1] https://www.redcrossblood.org/donate-blood/dlp/covid-19-anti...

[2] https://www.lifeservebloodcenter.org/donate/recovered-covid-...

* Probably not literally anyone, but it sounds like they are giving them out pretty freely.

I recommend reading this blog post [1] about the unknowns of immunity. Antibodies are only part of the immunity picture, as this tweetstorm [2] concisely explains.

[1] https://blogs.sciencemag.org/pipeline/archives/2020/07/07/mo...

[2] https://twitter.com/EricTopol/status/1278400526716211200?s=1...

Can confirm it’s not anyone but probably anyone eligible to donate. My sister could not due to visiting Guatemala in the last year.

What's the reason here? Do they expect an infection after a year?

There are a lot of not-altogether-scientific reasons. Gay people can’t give blood. If I recall correctly British people can’t either because of Foot and Mouth disease, something that happened in the 90s.

They are based on reasons that are statistically founded in some way because not everyone's blood can be tested as extensively as anyone would like, so reducing risk factors helps. I can't speak for every country but this is at least the case for the ones I've lived in and donated in.

Also, it's no longer true in many (most?) places that "gay people" can't give blood, as the distinguishing factor nowadays is usually recent sexual activity.

Note that "recent at-risk sexual activity" specifically means (among other things) a man engaging in sex with another man in the past 12 months (at least that's the case in Australia[1]).

Practically speaking this means that the vast majority of gay men cannot donate blood -- even if the criteria don't explicitly say the words "gay men cannot donate blood". Don't get me wrong, I agree that there are purely statistical reasons why these restrictions are in place (blood banks aren't out looking for excuses to reduce their supplies) -- but it's really not accurate to say that it's no longer the case that gay men cannot donate blood.

I also hasten to note that having many sexual partners is not considered "at-risk sexual activity" (though there are some exceptions), while two monogamous men in a homosexual relationship having sex is always considered as being "at-risk".

[1]: https://www.donateblood.com.au/faq/sexual-activity

> In 2015, the US Food and Drug Administration (FDA) relaxed its guidance but still required men to abstain from sex with another man for at least a year before they could give blood, a policy that the the American Public Health Association said was “not based in science”.

https://www.theguardian.com/us-news/2020/apr/18/us-blood-don...

Note that includes gay men in monogamous relationships too. It is a ban on gay men by another name.

Probably Jacobs-Creuzfeld, not Foot and Mouth.

Can confirm: I'm not allowed to give blood in Australia because of the small risk (with a potentially very bad downside if it manifests) that I'm infected with mad cow disease, due to living up in the UK in the 90s.

They still let me sign up to the organ donor list though.

Edit: foot and mouth disease was a decade later. I don't believe it's a problem in humans?

UK: Mad Cow Disease

https://www.capeandislands.org/post/why-us-still-bans-blood-...

"I've heard a lot of people in my social / online circles comment how they wish they knew whether or not they had antibodies"

What people need to realize is that the important questions are not whether you've got antibodies but:

1 - whether you'll sick when you're exposed to the virus again

2 - how sick you'll get the next time around

No one knows the answers to these questions yet, and we also don't know how antibodies bear on the answers to these questions.

It is possible that even with antibodies you'll still get sick if you're exposed to the virus (as we know is the case for some other diseases), and even if antibodies play some role in how sick you'll get it's possible that (again, as is the case with some diseases) repeated exposure to the virus will cause even greater illness than the first time around.

People shouldn't be assuming that having antibodies means they're immune.. we just don't know enough yet to say whether that's true or not.

Even if it is true, there might be a certain level of antibodies that are required to have a certain level of immunity, and we don't know yet what level of antibodies are necessary, nor the kind of antibodies necessary,

Finally, even if antibodies do grant immunity, we don't know how long that immunity would last, nor if such immunity varies from person to person for some not yet known reason.

If people could get reinfected we would have heard of more of those cases by now. We can be pretty sure that it confers several months of immunity at minimum.

Yes, 2-6 for sure.

A few days ago spanish media reported that one of the latest outbreak in Lleida affected a residence. They tested all inmates in the residence with PCR, and found 8 positives. 5 of these 8 had COVID-19 a few months ago with the first wave (I guess march / april).

Dengue fever antibodies make the second infection lethal.

The Philippines found this the hard way when testing a vaccine for dengue.

IIRC, Dengue has 4 different strains. When you get reinfected with the same strain, then you are good. Getting one of the other strains is the problematic case.

Good thing this isn't dengue fever.

As someone who was likely exposed back in March (no tests available at the time, and antibody test now is showing negative, which I don’t believe), I will anecdotally mention a few observations:

1. While you don’t get “really” sick again, you do “feel it” when re-exposed. Specifically I noticed diarrhea and a being a bit “off” on two separate instances where I believe I was Re-exposed.

2. I am 90% certain that even with immunity you remain a carrier when Re-exposed. I don’t know how much of a carrier, but just my strong suspicion (and why I have been extra careful when I have felt “off”)

Would love more medical data to back my hypothesis/observations/thoughts, but thought I would mention it.

Mate, no offense, but this is all personal opinion and personal guesses while you’ve actually tested negative!

If I were you, have a bit less certainty in your own guesses. Next to that, every person has a different set of symptoms and experiences, your experiences aren’t all that relevant.

Completely agree, and have been extremely careful. I won’t go into the basis of my belief (which included 103.7 fever).

My main point, and why I said anything at all is this: Even if you get infected, and get through it, there is no Scientific evidence to believe you are “in the clear” now.

Hoping that is an “add” rather than a subtract from the conversation.

From all we know from other viruses you might get reinfected but with a very high probability you won’t get very sick. Also, note that immunity is not only AB induced. You have other immune components like T Cells in your body patrolling and reacting to pathogen. Otherwise everybody would have gotten really sick from SARS2 which is obviously not the case.

I don't believe the virus is known to behave like this to a very high probability. It's unknown but possible the virus behaves like dengue fever and gets worse the second time.

Well it’s a corona virus. We have four other corona viruses endemic in humans. So I think it’s ok to compare it to those others

At this point, if there was antibody-dependent enhancement, we would have seen at least one case of it clinically.

> It's unknown but possible the virus behaves like dengue fever and gets worse the second time.

That's not how dengue works. There are four variants of the dengue virus, and it gets worse the second time only if it's a different variant. Quoting Wikipedia (https://en.wikipedia.org/wiki/Dengue_virus):

"Developing a vaccine against the disease is challenging. With four different serotypes of the virus that can cause the disease, the vaccine must immunize against all four types to be effective. Vaccination against only one serotype could possibly lead to severe dengue hemorrhagic shock when infected with another serotype due to antibody-dependent enhancement. When infected with dengue virus, the immune system produces cross-reactive antibodies that provide immunity to that particular serotype. However, these antibodies are incapable of neutralizing other serotypes upon reinfection and actually increase viral replication."

Are there not three variants / strains of COVID 19 too ? .

It is still too early now to say for sure , it is probable that second time may not come or be as severe but it is not certain

There are variants of COVID-19, but there is no evidence that the variations are of any biological significance.

There is no evidence that the variations change how infectious the virus is or how severe the illness is or how often people die from it or anything else.

Certainly I am not suggesting there is evidence, I am just saying it too early to dismiss the possibility . There is no concrete evidence saying it won't happen either.

Seasonal Flu does mutate fast enough that any immunity is typically seasonal, so it unreasonable to consider the possibility of mutation

The Labcorp test in the bay is literally free with a referral or $10 to get them to refer you to themselves. It's so easy.

Covid antibodies seem to fade quickly (similar to other cold family viruses). Some people only keep them for a few months after infection.

- https://blogs.sciencemag.org/pipeline/archives/2020/07/15/ne...

- https://abcnews.go.com/Health/covid-19-antibodies-fade-month...

Also, anecdotally, a friend donated blood at the Red Cross and found he had covid antibodies. He had not been sick and has had no symptoms.

This reminds me of an excellent video on use of cheap and less sensitive tests on "This Week in Virology".

https://youtu.be/kDj4Zyq3yOA (Feel free to skip to around 6:35).

Preprint: https://www.medrxiv.org/content/10.1101/2020.06.22.20136309v...

Update: the colorado.edu article also links to this preprint when it discusses the "separate computer modelling study".

Upvoting for TWIV, which is my daily exercise podcast of choice until this thing is over (and probably afterwards too).

The visual explanation of the ideas and concepts presented there, also mentioned in the later podcast:

Medcram: “Coronavirus Pandemic Update 98: At Home COVID-19 Testing - A Possible Breakthrough”

https://www.youtube.com/watch?v=h7Sv_pS8MgQ

Also, regarding the Ct value, from:

“Interpreting the COVID-19 Test Results: a Guide for Physiatrists”

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7268832/

In PCR tests “the diagnosis is made based on the threshold cycle (Ct) value. Ct is defined as the cycle number when the sample fluorescence exceeds a chosen threshold above the calculated background fluorescence.5 In other words, the lower the Ct value of a specific gene, the more the gene exists in the sample. However, the problem with a Ct-based diagnosis is that there is no absolute or constant Ct cut-off value, and Ct cut-off values are different for each diagnostic reagent even for the same gene.”

Yep, this was an eye opening video, showing how current mindset for virus test accuracy is wrong. If early on we channeled even small portion of all the trillions used to help with covid negative effects, world would be much better off now.

The study "Saliva is more sensitive for SARS-CoV-2 detection in COVID-19 patients than nasopharyngeal swabs" [0] was published in April, so I'm glad some additional steps have been taken to test it.

[0]: https://www.medrxiv.org/content/10.1101/2020.04.16.20067835v...

This seems like a really good idea but I worry about regulatory approval. Tests like these will tend to catch any Covid-19 cases in the window where you have a real chance of infecting someone from maybe 2 days pre-symptom onset to a few days post onset. And that's really all you need from an epidemiological standpoint. But from a medical standpoint if someone comes into your ER a week post symptom onset you want to know if they've got Covid-19 whether or not they're still infected, from a medical standpoint. And in that use case this is less likely to produce the correct answer than the already existing PCR test and the time lag probably isn't that important. It would be reasonable to allow it anyways but the FDA's rules are set up for the benefit of medical uses rather than epidemiological uses and I wouldn't be optimistic about it getting allowed baring pressure from above.

Also

>One researcher spiked 30 out of 60 saliva samples with inactivated SARS-CoV-2 in the lab. Then they shuffled the samples and gave them to another scientist to test with the RT-LAMP technology.

I wonder if that required a BSL-3 facility or not. Generally you can do medical work on this virus with any sort of facility or laboratory you have on hand but you're only allowed to do science involving the virus if you have a special laboratory with negative pressure and special failsafes and so on.

Spiking inactivated virus does not require BSL-3. You only need BSL-3 for isolating and growing live virus. So the vendors that create inactivated virus are doing so in BSL-3 facilities, but once it's inactivated you can work with it at much lower stringency. Seems like BSL-2 is standard.

This test relies on RT-LAMP to generate copies of the RNA instead of PCR which the "regular" tests use. It's pretty sensitive, but not as sensitive as state of the art PCR[1]. It's also way faster. No magic, just a slightly different technique with different trade offs.

[1] https://www.frontiersin.org/articles/10.3389/fmicb.2018.0208...

They've been using a saliva test here in Australia for several weeks for children and others who struggle with the nasal swab. Accuracy is apparently only 87% though.

https://www.news.com.au/national/saliva-test-with-87-per-cen...

87% sounds pretty good to me if it encourages a lot more people to get tested. I do wonder how many are avoiding a test because the know from others' reports that it's most uncomfortable. Think about the number of people who are afraid of simple medical procedures such as injections and apply that number to those who don't wish to get "brain scraped". It's a problem.

I really do not understand what the hold up is. In California I know over ten people whose results have taken more than seven or eight days to come back. This is in the past two weeks. I am still waiting for my results from Sunday six days ago. It really is inexcusable how mismanaged this epidemic is in the United States.

Agreed. Our president and many others focused on a vaccine while ignoring the need for PPE, tests, and contact tracing. Tens of thousands of Americans have died needlessly because we have failed to control the outbreak.

perhaps you should question what they're doing with your sample

I was tested nine days ago in the bay area. Still waiting for my results.

I live in Florida and was exposed to some respiratory illness with symptoms consistent with Covid-19. As soon as I had symptoms I immediately did what I could to get tested and specifically went for the fastest PCR test I could find (I'm in between primary care providers, couldn't find a way to even try to get a rapid test in my area). My insurance provider lists the only testing location covered as an undisclosed drive through site located somewhere in a city a good hour and a half drive away from me, with the second closest location being a 2 hour drive. The county had a free testing event scheduled for 2 days later but only had 400 tests, was first come first serve, and explicitly stated that you couldn't line up before they started testing as they literally weren't opening the gate to let traffic off of the street before then. The county also listed a 10 day wait before test results would even be available which is just flat out useless. Luckily the day before the county testing event my wife managed to find free testing offered by Quest in a somewhat nearby city that advertised that results would be available in 3-5 days instead of the 10 that the county claimed. I went there, got the nasal swab done, filled out a bunch of forms with clipboards and pens that were being reused with no sanitization between people, and signed up for Quest's online accounts to get the results.

Today is literally the first day that I'm free from quarantine. Not because I got my test results back, but because the latency in testing has gotten so drawn out that even if that test eventually comes back positive, I already meet the CDC guidelines for ending quarantine. That test was completely useless for me and only serves to add some data to the statistics around the pandemic. Right now, if you want to get tested and you're not a high priority case (i.e. hospitalized), at least in my area, you can't even do that for all practical purposes. It would be arguably better if they just randomly tested only 50% and had a more reasonable backlog than testing 100% with a backlog so long that all but the high priority tests are useless by the time the results are available.

We've had literally months of time to ramp up testing nationwide. This pandemic is far from over and it's already killed 145,000 Americans. This is inexcusable incompetence by the legislative and executive branch.

I am not sure the delay is due to the backlog only .

There are only few locations that actually run the assays the logistics of getting the sample to a location is significant. Not to mention regent and other resource shortages mean some locations are off and can’t process anyone . Your sample gets moved somewhere else

Also testing is many times done like binary searches. Some 96 samples are combined and tested , if it is are negative no further tests required , if they come positive breakdown and so on . While even in the worst case you may will not need 96 separate tests, the steps are sequential, tests will take more time if more people on average become positive.

Israel is 30 seconds.

https://hamodia.com/2020/06/30/israeli-company-introduces-co...

How to learn more about this sensor?

Emirates had something about 10 minute test back in April before boarding. I never figured out how they did that.

https://www.emirates.com/media-centre/emirates-becomes-first...

OK. I need to know how.

Not sure how they can do that but my company's device can do it in 20: https://www.forbes.com/sites/johncumbers/2020/07/01/a-new-20...

They’re antibody tests, those are extremely quick.

Interesting, I'd never consent to such a test. Where are the results stored and why would I give DNA to a Dubai Health Authority?

"Because no swabs are required, and no fancy equipment is needed, the tests are less vulnerable to backlogs"

That, sadly, is nonsense unless you have thousands of technicians available to process the tests. What we have here is a classic latency vs throughput problem. 45 minutes to an answer is a lot better than 24-48 hours which is about the best achievable with the PCR tests. But even if only a third of the 45 minutes is a tech messing with "pipettes, a heating source and an enzyme mixture" and the rest is waiting for reactions at best one tech could do perhaps 30 a day. On the other hand the "fancy equipment" in PCR labs can do thousands of such tests in one go (UW Virology here in Seattle is doing over 5000/day[1]).

[1] http://depts.washington.edu/labmed/covid19/

PCR test can be done in a few hours at most. It’s mostly the whole logistical problem that takes time.

Yes, the time's not the test, it's getting the sample from where its collected to the lab (e.g. a doctors office that may only get collected from once a day), assembling a batch, running it, collating the results and sending them back out. Any batch-oriented, industrial-scale testing operation will suffer the same delays regardless of the actual process.

I competed against this team in this year’s CU New Venture Challenge. They got first, we (https://docforge.dev) got second.

They are doing some amazing work with this and SickStick.

This is still not a home test. Keep working..

This company claims to have a home testing kit:

https://nanologix.com/

https://youtu.be/HVcSchmwXgc

n=60, deactivated virus, no humans.

University of Illinois CU is running this [right now](https://illinoisnewsroom.org/university-of-illinois-creates-...). They are returning results in hours.

less space than a nomad

Yes, and we'll have fusion reactors, HIV vaccines, and a cure to cancer any year now. Science is less subjective than consumer electronics.

Antibody tests give similar results in 10min with a drop of blood.

But you have to wait longer for the antibodies to appear in the human.

True, usecase perhaps is more to confirm that what you had was covid, kind of complementary to this.

A virologist is quoted comparing the search for a corona virus vaccine to the search for an HIV vaccine? Have I misread all the science over the last six months? That seems like an absurd comparison.

Mind explaining why it's absurd?

They are very different and HIV is very nasty. As a retrovirus that targets the immune systems (T-Cells, macrophages etc) with an unusually high genetic variability it is really hard to reliably invoke the kind of immune response that normal vaccines rely on for their efficacy. Corona virus on the other hand is an RNA virus that mainly attacks the respiratory system and invokes a strong (often too strong) immune response.

They just arn't very similar beasts, if you had to draw an analogy about the difficulty of the problem it would be that we don't yet have a cure for the common cold (which is a corona virus.)

> the common cold (which is a corona virus.)

A minority of them are. Most instances of the common cold are caused by rhinoviruses.

> They just aren't very similar beasts

It took a lot of HIV vaccine research to figure that out. Before that research, nobody knew. Of course now it is knows, when the people started, it wasn't clear what the problems will be.

Likewise, a lot of research is needed this time too. And the logistic needed to have the results usable is simply huge and everything needs time even if some steps are overlapped.

On that level, of still being far from having a ready solution, one can compare, on the level of the sameness of the virus it is of course now known that the viruses behave differently.

sound promising for air travel with 59/60 accuracy and last positive sample inconclusive

Reply on news.ycombinator.com