DON’T GET TESTY – PART 3. Schools can open safely (and avoid re-closing) if they don’t choose the wrong SARS-CoV-2 screening test

[An edited shorter version also appears on Medscape]

As schools re-open, periodic screening for SARS-CoV-2 may add a layer of protection. Conversely, choosing the wrong screening test may result in avoidable school re-closings.

In Part 1 and Part 2 of this series, I explained why once weekly SARS-CoV-2 testing by polymerase chain reaction (PCR) is a losing strategy; weekly PCR guarantees false reassurances, needless quarantines, and unnecessary shutdowns. What about an alternative approach?

There is an invaluable figure by Michael Mina and colleagues. At-a-glance, the figure tells you everything you need to know about SARS-CoV-2 screening tests. Well, almost everything. Actually, in order to better convey some critical points, it might be good to consider a modified version:

Figure 1.

Infection progression

Let’s start with the “Amount of virus.” When someone is infected with SARS-CoV-2 (day 0), an initially minuscule amount of virus rapidly starts to replicate. Viral replication proceeds faster and faster until, at some point, the virus is so abundant it can no longer be contained. The “infected” person become “infectious”—contagious. They can now share their virus with others.

The transition from not-yet-contagious to contagious might generally occur around day 3. Following day 3—usually by about day 5—Infected individuals hit their contagiousness peak. The peak represents when the immune system starts to contain the situation. So from day 5 onward, the amount of virus begins to drop.

Ultimately, after about a week of contagiousness (i.e., by about day 9), the virus is essentially vanquished. Viral particles remain but, like slain enemies on a battlefield, they pose no threat. The infected person is no longer contagious.

Nonetheless, the infected person may continue to harbor “dead” virus bodies for weeks—even months! Because the exact time to clearing the bodies is uncertain (and variable), the “Amount of virus” line in Figure 1 trails off hazily.

Regardless of when virus zero occurs, there is generally no “live,” active virus beyond about a week of contagiousness. So what does all this mean for screening tests?

Two testing strategies

With regard to screening-test detection level, there is no question PCR is “better.” It can detect virus very slightly earlier, and for much longer. However, what we really care about is not detecting any virus. What we care about is detecting contagious virus. 

For detecting contagious virus, antigen tests perform just as well. Additionally antigen tests do not create confusion by detecting virus that is no longer a threat. 

Consider two screening strategies: once weekly PCR vs. thrice weekly antigen testing.

Figure 2.

Once weekly vs. thrice weekly? That might not seem like a fair comparison. But actually, if devising a screening program based on cost, thrice weekly antigen testing (about $5 per test) is actually much MUCH cheaper than weekly PCR (around $150 per test). 

Perhaps counter-intuitively, the less expensive option—less able to detect virus—is actually the BETTER option. Here’s why:

Problems with PCR

In Figure 2, day 0 is a Sunday. Let’s say, at a weekend playdate, a student gets infected. Inside the unsuspecting student, the virus starts to replicate. But when a PCR test is done on Wednesday, while the amount of virus will be increasing it will still be below the threshold for detection.

Since PCR can often take several days for results (Figure 2 generously supposes only two days), results for Wednesday’s test come back Friday. The results are negative: true for how the student was on Wednesday; false for how the student is on Friday. The now-infectious student is thus—along with classmates, teachers, parents, and administrators—falsely reassured. The student is allowed to stay in school … and to return the following week!

The following week, the infected student is again tested on Wednesday. The results again come back on Friday. This time the result is positive: true for Wednesday; but “false” (at least in terms of contagiousness) for Friday. So after an entire week of being contagious in school, the student is only now sent home to self-isolate. Inappropriately. The classroom is quarantined. Inappropriately. People feel protected. But they have been harmed.

Nobody benefits. Everybody loses. The expensive, sensitive test fails!

In fact, because of false positives (much more on those, and on likewise harmful “indeterminates,” here), another classroom winds up shutting down: As it happens, that class’s teacher was in a situation like that shown for “Week 3” in Figure 2 (PCR truly positive for virus, but falsely positive for contagiousness). In other words, the teacher had a distant infection from which she has long recovered. She is no longer a threat. But given multiple classrooms now with “positive” tests, it appears the school has an “outbreak.” The school shuts down. Avoidably.

?The alternative

What would have happened with antigen testing? Had the play-date-infected student been screened with thrice-weekly antigen testing, that student would have been detected on Friday, almost immediately as he/she was becoming contagious. The results would have been known almost immediately (within 15 minutes). The student would have been sent home to self-isolate. Close contacts would have been sent home to quarantine. But as opposed to the alternative quarantine for “dead-virus” (pointless), this quarantine would be in the setting of “live virus.” (It might actually protect people.)

Additionally, the PCR-positive teacher would have tested negative by antigen. Appropriately. The teacher’s class would therefore not have had to quarantine. Appropriately. And there would been no perception of an “outbreak” at the school. The school would have remained open. Safely. Correctly.

In fact, unnecessary closures and quarantines might be even more avoidable with even more-regular antigen screening. For instance, if everyone were screened daily, at the start of the school, before entering classrooms, “positives” could be identified (and isolated) prior to having any chance of exposing others. Thus, any quarantines might become unnecessary. The availability of cheap, simple, rapid testing could make this possible. And for schools needing logistical help, there are end-to-end solutions to coordinate implementation.

Regardless, even without daily screenings, a strategy of antigen testing multiple times per week would allow threats to be discovered right as they are emerging. If infected individuals are somehow missed on Friday, they will surely be discovered Monday (Figure 2). Conversely, with weekly PCR, they might be missed altogether (especially true for emerging viral variants of concern, which can escape detection by PCR).

We’ve come too far. We’ve suffered so much. We can get kids learning and teachers teaching again—in-person. Safely. Affordably. Sensibly. We just need to not pick the wrong screening test!