“Contagiousness testing” – the TLDR version. If your goal is to find the people who are infectious before they infect other people you need testing that is rapid and frequent. Which means they need to be cheap & easy-to-use anywhere (think paper strips you spit on). So what’s the hold up? Basically, high standards…

One of the biggest hold ups with deploying rapid coronavirus tests is that the FDA isn’t authorizing them because they aren’t as sensitive as the conventional RT-PCR tests (a type of test that looks for viral RNA). The FDA is basically holding all the tests to the same really high standard. But this standard is so high – the tests are so sensitive – that tests must detect teeny tiny amounts of virus (or virus parts) in order to get approved.

I say virus parts because a lot of times what the RT-PCR tests are detecting is just viral RNA pieces, not the whole virus. Viral RNA by itself isn’t dangerous – it can’t get into our cells without having it’s coat with those spike proteins, etc. And a lot of the viral RNA being detected by RT-PCR tests isn’t even the intact genome, just pieces of it. These pieces can linger on surfaces for a long time, which has led to a lot of media coverage about how virus has been found on surfaces after days, etc. But what those scary stories are not telling you (and sometimes which the researchers haven’t even investigated) is that when they try to culture the virus (“grow it” and get it to infect other cells in a dish) they can’t. What’s on those surfaces is usually NOT infectious. 

Similarly, after the first few days, the viral RNA a person sheds is mostly “harmless” RNA. The probability of shedding “live” (culturable) virus falls sharply starting around 3 days after symptom onset and scientists have been unable to culture any virus after 9 days or so, even though patients may continue to test positive w/RT-PCR. (This is why the CDC changed its guidance on how long people need to self-isolate & no longer recommends people need 2 neg tests to reenter society) 

So, clearly, for screening to find who is contagious, the RT-PCR is “too sensitive” so holding “contagiousness tests” to these standards is misguided. And it’s counterproductive because it’s holding back the implementation of not-so-sensitive tests which are sensitive enough for what you want with a “contagiousness test.” 

In order to achieve sensitivity required by FDA, test makers use technically-intensive protocols (e.g. extracting RNA & using RT-PCR to make lots of copies of it) & fluorescence, which lets off stronger signal than colored readouts, but needs special machines & is more $$. 

Semi-“Rapid tests” like the Abbott ID use a simpler copying method than RT-PCR that’s faster, but this test is one at a time so if 1 sample takes 15 min it takes a long time to run a lot of samples, as opposed to RT-PCR, where hundreds of samples are run at once but you have to collect and process them all first. 

The really rapid tests are likely to be “antigen tests” which look for viral proteins instead of the viral RNA. There are a couple already (like Quidel’s) but they use fluorescence and need special machines. 

Unlike RT-PCR or other conventional “molecular tests,” antigen tests don’t use any amplification step, so they aren’t making any copies of the viral part before they detect it. Therefore you need higher levels of the virus for the test to find it. But, if all you want to know is whether the person is shedding enough virus to infect someone else, that’s all you need. 

The real key’s testing frequency. As long as you test really frequently (e.g. 3X a week) you’ll catch people when contagious with minimal time that they’re exposed to others. And if they’re masking, social distancing, etc. they’re unlikely to have infected anyone during that time.

You might still be thinking, what about those couple of days they were out there? Maybe we could have caught them sooner with a more sensitive test? In reality, the person probably wouldn’t have been tested at all if it weren’t for the rapid tests. If they’d gotten a RT-PCR test, they’d probably still be waiting for their results for those couple of days and, if they didn’t have any symptoms they might not think they had any need to self-isolate. 

Furthermore, at that initial testing time, the person could have been below the levels detectable by even the most sensitive test. So they’d test *negative* on that RT-PCR test, but then become infectious and they’d have no clue unless they were tested again soon, which is unlikely would happen. Instead, what’s likely to happen is the person goes about their business, gets a call in a few days that their test result is negative, and thinks they’re fine.

This isn’t even getting into fact that even in a person is truly uninfected when they get a RT-PCR test, they could get exposed & infected while waiting for test results. Yet, camps, etc. are saying if you have a negative test result from within past 12 days you can come?! 

Really cheap, rapid tests could be realistic if there was more incentive to invest in them. If, for example, the government did the sort of thing they did with vaccine development – basically offer to eat the cost if things don’t work and offer to buy up a certain (really big) number of tests so that the companies don’t have to worry about not having a market and thus can invest in really scaling up their production capacity.

These tests won’t replace the “gold standard” tests, which will still be important for symptomatic patients and for use in hospitals, clinics, high-risk environments, etc. But they would allow us to test a huge population of people we currently aren’t testing – and frequently. This is crucial because many people who are infected with the virus are asymptomatic (I think they’re thinking this is usually about half of people, but I’ve heard a lot of estimates), but they can still infect others. If we could find these people when they’re contagious, we could dramatically decrease disease spread.

That’s the main point Michael Mina makes in this TWiV podcast https://www.microbe.tv/twiv/twiv-640/

And which I try to explain in my longer post http://bit.ly/reallyrapidtests

Leave a Reply

Your email address will not be published.