RAT, but for everything
July 29, 2022 5:12 AM Subscribe
How realistic is my dream of a test, similar to a COVID Rapid Antigen Test, to tell you if you're sick or not?
I enjoy the relief that a negative RAT gives, but then I wonder - I'm still potentially sick with something else, I wonder if I am contagious?
Feels a bit science fiction-y but I'd love a breathalyser style device or yes nasal swab test that will tell me if I'm sick and/or contagious (or is it just allergies?) Is this something that's remotely feasible or is this real science fiction territory?
I enjoy the relief that a negative RAT gives, but then I wonder - I'm still potentially sick with something else, I wonder if I am contagious?
Feels a bit science fiction-y but I'd love a breathalyser style device or yes nasal swab test that will tell me if I'm sick and/or contagious (or is it just allergies?) Is this something that's remotely feasible or is this real science fiction territory?
What exactly do you mean by "sick" here?
We can make a lateral flow / RAT test for almost any specific protein, including almost any specific virus (or family of viruses), but if you have an infection it could be one of a million different viruses, including many that are not well studied. So choosing which ones to test for would be a big problem for this plan.
Also, everyone is always carrying bacteria and viruses of many kinds without being "sick". You could have traces of anything without necessarily being infectious or unwell enough to need to change your behaviour. So deciding what threshold to put on this test and what to do with the results needs thinking through further for this plan.
posted by richb at 6:23 AM on July 29, 2022 [9 favorites]
We can make a lateral flow / RAT test for almost any specific protein, including almost any specific virus (or family of viruses), but if you have an infection it could be one of a million different viruses, including many that are not well studied. So choosing which ones to test for would be a big problem for this plan.
Also, everyone is always carrying bacteria and viruses of many kinds without being "sick". You could have traces of anything without necessarily being infectious or unwell enough to need to change your behaviour. So deciding what threshold to put on this test and what to do with the results needs thinking through further for this plan.
posted by richb at 6:23 AM on July 29, 2022 [9 favorites]
It’d be pretty straightforward to do this as a PCR test.
I went for a PCR test earlier this week and along with COVID I was automatically tested for Influenza A/B and RSV. It's not everything that one could be infected with (by a long shot) but I suspect it is a big chunk of the things going round and likely to cause the most problems.
posted by Cheese Monster at 6:25 AM on July 29, 2022 [1 favorite]
I went for a PCR test earlier this week and along with COVID I was automatically tested for Influenza A/B and RSV. It's not everything that one could be infected with (by a long shot) but I suspect it is a big chunk of the things going round and likely to cause the most problems.
posted by Cheese Monster at 6:25 AM on July 29, 2022 [1 favorite]
Best answer: They definitely do make multiplex lateral flow tests for other purposes, so I think you could make one that tests for COVID, RSV, flu and, like, a couple of common adenoviruses. But the rapid antigen tests test for specific antigens, so you have to actively decide to test for things - you can't just say "all respiratory infections" because they have different markers AND the amount that would be worth reporting on is going to differ between pathogens (a dangerous amount of coronavirus might be different from a dangerous amount of RSV).
Still, let's imagine you had a multi-antigen test that had been validated to work for, say, 100 different pathogens. At this point, false positives become a real problem. This happens a lot with non-invasive pregnancy screens - all tests have a some level of false positives, and the more things you test for, the more false positives you're going to get. If the disease is rare in the population, a positive result is often much more likely to be false than true. And if there are 100 things you're screening for in the test, the chances you're going to get a false positive is 100x what it would be if you were only testing for one thing.
posted by mskyle at 6:28 AM on July 29, 2022 [6 favorites]
Still, let's imagine you had a multi-antigen test that had been validated to work for, say, 100 different pathogens. At this point, false positives become a real problem. This happens a lot with non-invasive pregnancy screens - all tests have a some level of false positives, and the more things you test for, the more false positives you're going to get. If the disease is rare in the population, a positive result is often much more likely to be false than true. And if there are 100 things you're screening for in the test, the chances you're going to get a false positive is 100x what it would be if you were only testing for one thing.
posted by mskyle at 6:28 AM on July 29, 2022 [6 favorites]
Best answer: When I took a PCR confirmation test for COVID the lab used a BioMerieux BioFire FilmArray, which can perform multiple PCR tests in parallel. I was given the Respiratory panel which looks for 22 different markers at the same time including most influenzas and COVID-19. They make assay panels for other groups such as 43 types of bloodborne infections and some genetic markers as well.
TLDR the technology for what you want exists and could be expanded to include most definitions of "sick". However, the US FDA severely frowns upon people performing their own diagnostic tests at home for a lot of things unless it's backed up with medical care.
Like, you can go on eBay and buy a BioFire analyzer, but you'd need a source for the reagent cartridges and typically the manufacturer wants to see your lab licensing before they will ship. That's not to say you can't find a company with less scruples and obtain one eventually.
posted by JoeZydeco at 7:34 AM on July 29, 2022 [2 favorites]
TLDR the technology for what you want exists and could be expanded to include most definitions of "sick". However, the US FDA severely frowns upon people performing their own diagnostic tests at home for a lot of things unless it's backed up with medical care.
Like, you can go on eBay and buy a BioFire analyzer, but you'd need a source for the reagent cartridges and typically the manufacturer wants to see your lab licensing before they will ship. That's not to say you can't find a company with less scruples and obtain one eventually.
posted by JoeZydeco at 7:34 AM on July 29, 2022 [2 favorites]
Best answer: Pixelcorp does a wider-spanning test modeled upon their COVID home test -- not "everything" but several things!
posted by churl at 8:48 AM on July 29, 2022 [1 favorite]
posted by churl at 8:48 AM on July 29, 2022 [1 favorite]
I vote science fiction.
Because, isn't this essentially what Theranos claimed they were going to do*? It didn't go well.
* if we're being picky, Theranos said they aimed to use blood tests rather than nasal swabs or whatever - but i can't help thinking that the whole detect-anything-instantly! idea may hit similar issues no matter the specific mechanism of the test
posted by rd45 at 9:18 AM on July 29, 2022
Because, isn't this essentially what Theranos claimed they were going to do*? It didn't go well.
* if we're being picky, Theranos said they aimed to use blood tests rather than nasal swabs or whatever - but i can't help thinking that the whole detect-anything-instantly! idea may hit similar issues no matter the specific mechanism of the test
posted by rd45 at 9:18 AM on July 29, 2022
There was a research project in the early 1990s (I think) which put chemical tests on the ends of hypodermic needles. You'd stick one into your skin and it would test for a chemical. The idea wasn't to find a specific disease but to perform a standard blood test on the spot, with no blood being removed.
As a side note, this is important not because it's quick and convenient, although that makes a huge difference, but because it's not unheard of for doctors to order enough blood tests on a newborn to kill it. One of the first things impressed on neonatal nurses is that you keep very careful track of blood tests on babies and refuse to do them past a certain point.
The outfit doing this was trying to make a chip which did every common test at once. I never heard anything more about them, which is unfortunate because even doing a single test instantly and accurately (and I think it's common knowledge that medical testing is not that accurate) would be a very good thing.
The test you're looking for would presumably look for specific antigens, which would mean a separate test for each virus or bacterium. It's easy enough to do this for one thing, but I can't see any way to make a test that picks up not just any known infection but also unknown ones.
One could go with a general approach - look for the signs which identify the body being stressed or defending itself. Given the state of computers and programming it would be simple to make a system which could do a battery of tests and also offer a diagnosis, and systems like this were postulated in the 1960s and in some use by the early 1980s.
Such a system would be of some use if it could tell if you were infected with something, but immensely more valuable if it could diagnose the majority of conditions and not just infectious agents.
There are a lot of things which can go wrong with the human body, but the number isn't infinite, and it's not as large as one might think. Even such a system couldn't administer all the tests at once it's well within the reach of modern technology.
A friend had a mechanical system which could do some of this, and it dated to about 1920.
I was once given an analogous thing, made for a different purpose, and it sold for about $12. I'm not sure why nobody has done this yet.
posted by AugustusCrunch at 12:51 PM on July 29, 2022
As a side note, this is important not because it's quick and convenient, although that makes a huge difference, but because it's not unheard of for doctors to order enough blood tests on a newborn to kill it. One of the first things impressed on neonatal nurses is that you keep very careful track of blood tests on babies and refuse to do them past a certain point.
The outfit doing this was trying to make a chip which did every common test at once. I never heard anything more about them, which is unfortunate because even doing a single test instantly and accurately (and I think it's common knowledge that medical testing is not that accurate) would be a very good thing.
The test you're looking for would presumably look for specific antigens, which would mean a separate test for each virus or bacterium. It's easy enough to do this for one thing, but I can't see any way to make a test that picks up not just any known infection but also unknown ones.
One could go with a general approach - look for the signs which identify the body being stressed or defending itself. Given the state of computers and programming it would be simple to make a system which could do a battery of tests and also offer a diagnosis, and systems like this were postulated in the 1960s and in some use by the early 1980s.
Such a system would be of some use if it could tell if you were infected with something, but immensely more valuable if it could diagnose the majority of conditions and not just infectious agents.
There are a lot of things which can go wrong with the human body, but the number isn't infinite, and it's not as large as one might think. Even such a system couldn't administer all the tests at once it's well within the reach of modern technology.
A friend had a mechanical system which could do some of this, and it dated to about 1920.
I was once given an analogous thing, made for a different purpose, and it sold for about $12. I'm not sure why nobody has done this yet.
posted by AugustusCrunch at 12:51 PM on July 29, 2022
You might be interested in reading about testing for C-reactive protein, which is a marker of inflammation. Problem is, inflammation comes from many sources, so it really needs to be triangulated with other information.
posted by deludingmyself at 1:49 PM on July 29, 2022
posted by deludingmyself at 1:49 PM on July 29, 2022
Best answer: I hate to break it to you, but a RAT test does not tell you if you're sick or not.
It tells you if you're contagious or not.
For me, this makes the tests actually more interesting. They answer the question of "If I go out tonight, will I be putting people's lives at risk?" This is the primary piece of information needed to help stop the spread of a disease. That makes them more useful to me than a PCR test!
But I'm kind of with everyone else: asking if you are "sick" or not is a value judgement more than an objective measure of biological factors. It's like asking if the weather is "good" or "bad": weather's a roiling chaotic system of heat and pressure and humidity, and one person's life-ending drought is another person's "fantastic holiday weather". Having loads of healthy, young-looking, fast-reproducing cells that don't ever die is called cancer, and we kind of don't like that very much.
posted by rum-soaked space hobo at 4:48 PM on July 30, 2022 [1 favorite]
It tells you if you're contagious or not.
For me, this makes the tests actually more interesting. They answer the question of "If I go out tonight, will I be putting people's lives at risk?" This is the primary piece of information needed to help stop the spread of a disease. That makes them more useful to me than a PCR test!
But I'm kind of with everyone else: asking if you are "sick" or not is a value judgement more than an objective measure of biological factors. It's like asking if the weather is "good" or "bad": weather's a roiling chaotic system of heat and pressure and humidity, and one person's life-ending drought is another person's "fantastic holiday weather". Having loads of healthy, young-looking, fast-reproducing cells that don't ever die is called cancer, and we kind of don't like that very much.
posted by rum-soaked space hobo at 4:48 PM on July 30, 2022 [1 favorite]
« Older They went off to the east, at least I think that's... | Online publishing platform with marginalia support... Newer »
This thread is closed to new comments.
posted by u2604ab at 6:00 AM on July 29, 2022