you may also have seen news stories suggesting  that some COVID tests are too sensitive,  

or that cheaper, faster (and even less  sensitive?) tests might be better. But  

why would they be better? And better for what? According to public health experts at Harvard  

and Brown University, in the fall of 2020 the  US needed to do 8x more testing to successfully  

suppress COVID-19, India needed to do 4X more  testing, Brazil needed to do 66X more testing,  

Bolivia needed to do 140X more testing, and  Mexico needed to do over 400X more testing.  

The idea behind saying “COVID tests are too  expensive and too sensitive” is that part  

of what's kept countries from doing enough  testing is the kind of tests being used. 

The current gold-standard for COVID testing is a  nasal swab PCR [polymerase chain reaction] test.  

This test is highly sensitive, which means  it's able to detect the presence of the virus  

even at very low levels of viral load. That's  great for a doctor who wants to know whether  

their patient was infected with COVID-19. But if  you're screening an entire community for COVID  

outbreaks, then PCR tests have a few downsides. First, PCR tests are expensive. To avoid breaking  

the bank, many places are focusing on testing only  people who already have COVID symptoms or who were  

recently exposed to someone with COVID. Using  only expensive tests means that we don't test  

(and therefore, don't catch!) most people who  are contagious without showing symptoms – and  

we know that a significant proportion of COVID  transmission comes from these asymptomatic people. 

Second, PCR tests take time: usually a couple of  days for results to come back from the lab. Here's  

what the typical viral load in an infected person  looks like over time – they're most infectious  

around the peak and their infectiousness declines  afterwards. So if test results take a few days to  

come back, chances are you miss this peak – which  means missing the window of time during which the  

person is most likely to infect other people. Essentially, although PCR is very sensitive,  

it's also expensive and slow, so while it's great  for diagnosis, it may not be the best test for  

screening a community. And the cheaper, faster  testing alternatives are – no surprise – not as  

sensitive as PCR. BUT - and this is a subtle point  and maybe wouldn't matter if PCR were cheap and  

fast, BUT when it comes to screening a community,  PCR's high sensitivity could itself be a downside.

PCR tests can detect very low viral loads, which  is great if you happen to test somebody very  

early on in an infection, or very late and you  want to confirm whether they were infected. But  

this sensitivity also means that a person can  test positive for the infection when they're  

basically no longer infectious – and in fact, PCR  is so sensitive that it can detect viral fragments  

and give a positive result even when there are  no viable viruses left. So it's likely that by  

using PCR tests we're quarantining & contact  tracing some people when we don't need to. 

It's not that the PCR test is a bad test, in fact  it's a great test for what it's designed to do,  

which is to accurately and reliably detect even  very low viral loads. It's just that it's more  

expensive, slower, and perhaps even more sensitive  than would be ideal for screening a community. 

A less sensitive test could catch just  as many highly infectious people as PCR,  

but it would miss people who are post-infectious,  which you could argue is a good thing for  

screening purposes since those people  are no longer a significant threat. The  

downside is that you'd also miss people who are  pre-infectious, which is undoubtedly a bad thing. 

But the math works out that if you have a test  that's just a little bit less sensitive than PCR,  

but you use it a lot more frequently,  you're actually more likely to detect an  

infection compared to infrequent PCR testing.  So can you identify more infections overall. 

All of this suggests that if you're trying  to screen a community for COVID outbreaks,  

you can get more testing bang for your  buck with rapid testing than PCR testing;  

that is, you can identify more infectious  people more quickly for less money.  

As long as the rapid tests are only slightly  less sensitive than PCR tests but a lot cheaper. 

Rapid testing is not a replacement  for PCR testing in a clinical setting,  

nor is it a replacement for masks,  physical distancing, handwashing,  

and so on. And if PCR testing cost 1/10th  of what it does and took 1/10th the time,  

we probably wouldn't be having this discussion.  But if we're going to test widely and frequently  

enough to suppress COVID (and we really  should), then we can get more for our money  

with tests that are cheaper, faster, and - perhaps  counter-intuitively - a little bit less sensitive. 

Ok, so that's the argument for rapid  testing, and here are the caveats. 

No medical tests are perfect, and a negative test  result isn't a foolproof guarantee of not being  

infected. Because rapid tests are less sensitive  than PCR tests, this makes it more likely that  

infected people will slip through the cracks  of a test. So as long as COVID is a problem,  

it's important to still take precautions  like wearing a mask and social distancing  

even if you test negative. And for  people who were likely exposed to COVID,  

health authorities recommend that they quarantine  for 14 days regardless of a negative test result,  

and that their negative rapid test  results are verified by a PCR test. 

When a rapid test is just a little less  sensitive than PCR, then the downsides of  

missing some infected people can be balanced out  by the benefits of more widespread and frequent  

testing. But this only works up to a point. If  the sensitivity of the rapid test is too low,  

the cost of missing many infections will  outweigh the benefits of widespread testing. 

The graph we showed for viral load was based  on measurements of patients' viral loads,  

and while all else being equal, a higher viral  load tends to mean you're more infectious,  

there are a lot of other things that  can affect how infectious you are 

We didn't really mention what happens when you  aren't infected but the test gives a positive  

result anyway – this isn't likely on any given  test, BUT if you're testing a huge number of  

people frequently, then just by statistics alone  you can end up with a large number of false  

positive results. The rate of false positives  is comparable between PCR tests and rapid tests,  

but matters more with cheap rapid tests  since the whole idea is to do way more tests. 

Finally, vaccines. So won't all of this  be irrelevant once vaccines come out?  

Well, vaccines will certainly help, but  unfortunately, testing will probably still  

play a major role in suppressing COVID.  It's really a topic for another video.

A very big thankyou to Brilliant  for sponsoring this episode of  

MinutePhysics - with MinutePhysics, I try  to bring you deeper level explanations,  

while Brilliant engages you to solve  fascinating, challenging problems yourself,  

guiding you to a deeper level understanding  of science and mathematics. Like their course  

on knowledge and uncertainty - what's the most  efficient way to zero in on the true criminal  

from a lineup of suspects? Or, how worried  should you be if you know a medical test is 90%  

sensitive and you get a positive result? To improve your knowledge and reduce your  

uncertainty, go to brilliant.org/MinutePhysics/  and sign up for free. The first 200 people will  

get 20% off the annual Premium subscription,  with full access to all of Brilliant's courses,  

puzzles, & daily challenges. Again,  thanks to Brilliant for their support.