Teszler: Let’s Test Better
Current rapid tests aren’t nearly good enough — we need a new generation of more accurate and useful ones.
Rapid antigen tests are having something of a moment. In December, demand for the tests surged, reflecting a widespread desire to test before visiting relatives over the holidays. Last month, the Biden administration premiered its website to allow every household to order four free rapid tests. On campus, we’ve presumably administered thousands of such tests over the past weeks as people have tested out of quarantine.
But for all their newfound importance, rapid tests have a key drawback: they are not that good at their job. Though their rate of false positives is relatively low, they miss many cases that more sensitive methods can pick up, especially at the start of infection. Despite the convenience they represent, the current generation of rapid tests leave much to be desired — we need a new generation of rapid, at-home tests that are both more sensitive and accurate to manage this disease as it becomes endemic.
In a recent Arizona study, Binax Now tests (a common brand) had a 64.2% sensitivity for symptomatic patients — meaning around two-thirds of those who truly had COVID-19 tested positive on the tests. For asymptomatic individuals, the sensitivity was an abysmal 35.0%. To be fair, the tests had few false positives and were more sensitive to those with the very highest viral loads, who are most transmissible. Yet the relatively low sensitivity limits their usefulness to the average consumer. If I wanted to know if the cold-like symptoms I was experiencing were COVID-19, a negative test on antigen simply is not very reassuring, as over a third of people who truly have the disease test negative.
On the other end of things, rapid tests also seem to be causing issues at the conclusion of COVID-19 cases — people continue to test positive after day five, often up to day 10, a trend confirmed by emerging studies. Although Dartmouth has not released data on how long the average quarantine is, anecdotal reports suggest a large number of students are not being released on day 5 as they continue to test positive.
Now, there are a few possible explanations. One is that the CDC has totally erred in its new 5-day isolation guidance and that the large majority of people are still quite transmissible after quarantining for five days. Yet, another possible explanation remains — that rapid tests are picking up viral remnants rather than live virus itself; lab results have suggested this could happen, although this explanation appears unlikely, since it takes a significant viral load to cause a rapid test to be positive. Or it could be some mix, with small amounts of live virus and dead virus, in which case someone is barely transmissible. In the words of Matthew Binnicker, a virologist at the Mayo Clinic, “There is no test that we have for infectiousness.” Rapid tests are not perfect measures of how infectious someone is, despite the fact they have some general correlation with viral load.
What this all means is rapid tests are rapidly becoming an annoyance, rather than the convenient solution they are supposed to be. At the start of symptoms, they fail to pick up COVID-19, leaving people in the lurch and unsure if they actually have the disease. And at the other end, people are trapped in quarantine by positive tests — some of whom are likely still infectious, as well as others who may not be infectious or who are only mildly contagious.
What we need, in the theoretical sense, are tests that are more sensitive at the beginning of the illness and more attuned to live viruses specifically at the end. This is easier said than done, but already a few solutions seem promising. While we seemingly go all-in on rapid antigen tests, we must also be sure to support the next generation of testing.
On the problem of low sensitivity, there’s already an in-between class of tests — rapid nucleic acid tests. These tests vary somewhat widely in their details, but they all work on the same general principle of amplifying the detection of coronavirus genetic material, making them far more sensitive than antigen-based tests. Some of these tests are even available for the home consumer, such as the Detect and Cue Health systems, with sensitivity rates above 95%. These tests usually consist of both a terminal, essentially a heater to allow reactions to be processed at the right temperature, and the individual test kits used for each test. They still miss a handful of cases compared to standard RT- PCR — but are much better than antigen tests. There are also rapid PCR tests that analyze for a number of types of viruses — including flu — one of which I received from Dick’s House this fall after running a fever. Why Dartmouth doesn’t offer this option to more symptomatic or recently exposed people is puzzling to me.
The big obstacle to these tests remains cost and availability. As of now, those two tests are significantly more expensive than rapid antigen testing, at several hundred dollars for the system and a few tests. The same used to be true for rapid antigen tests, which were very rare and difficult to find for months. Once these issues are solved, though, one could imagine a future where our test drop boxes could incorporate rapid nucleic acid terminals, where you deposit a sample and get results back in less than an hour. I would have greatly appreciated that option for the times I was exposed to friends with COVID-19 this term — who maybe wouldn’t even have exposed me if they had known they had COVID-19 earlier but had no choice but to wait 24-48 hours for test results.
Testing to leave isolation presents a bit more of a dilemma, and more research is likely required to find a reliable method for identifying when live virus has significantly decreased. One potential area to start is viral culture, already the standard method for testing for several viruses, such as herpes. These tests rely on the growth of live virus, and take several days and a secure lab to perform. But as Operation Warp Speed demonstrated for vaccines, sufficient money dedicated to a scientific problem can yield results. Developing viral cultures with several hour turnaround time, for instance, could help us get the most useful information as to whether people should leave isolation — determining how much live virus they have, and how much chance they stand to spread the disease.
All this might seem a bit extreme. Some may wonder that as COVID-19 becomes endemic, will we even need such a level of testing? Could we not more or less ignore the illness, treating it much like other respiratory infections, which we largely do not test for? In an ideal world, perhaps — but given the ability of COVID-19 to evolve and break through immunity, it seems likely this illness will continue to resurface and cause illness in vulnerable populations. Better testing can provide valuable information to help prevent negative outcomes. We have effective antiviral treatments, but they are limited by the ability to quickly get back an accurate positive result to confirm infection. More sensitive rapid testing could get treatment into the hands of those who need it and ensure that people minimize the amount of time staying home in quarantine.
I do not know — and indeed, many experts do not know — what the best plan for future COVID-19 management should look like. But better rapid testing would be a key tool to managing the spread of COVID-19, no matter the future of the virus.