Test, Redux
Six plus months into this pandemic, my spouse has overheard many calls. So when recently he listened to me answer the same question — if different ways — 10 times, he asked “How can you stand answering over and over again?”
I hadn’t even noticed.
I once had a professor who, after I’d asked to explain something for the umpteenth time, then apologized for not getting it sooner, say that my not getting it meant he didn’t understand it well enough to explain it clearly.
And that’s how I view those questions — ask me 10 times, I’ll try 10 different ways to explain it. I’m not going to grow frustrated (unless you’re my 10 yr old asking why he needs to clean his room). I want you to understand it.
Any scientist who says “It’s too complicated, it’s too technical” doesn’t understand it well enough to explain, or doesn’t care.*
So with that intro, I’m going to explain tests again.
Antigen Tests
SARS-CoV-2’s Spike and Nucleocapsid proteins are targets for antigen testing. These proteins are on the surface of the virus and show up when the virus is released from the cell. If you’ve got viral antigens being expressed, you’ve got active (probably infectious) virus. Sometimes that replicating virus makes you feel sick. Sometimes it doesn’t.
Antigen tests look for the expression of these proteins. The proteins are very preserved, which means that the tests are very specific for the SARS-CoV-2 antigens. They’re not going to pick up another virus.
The test strip has a specific antibody on it that will only recognize the SARS-CoV-2 antigen. When the sample is placed on the strip, if the antigen in your nasal swab is there, it will attach to the antibody and show up with a line on the strip. If there is no antigen, nothing shows up.
But because it depends on active virus present, and viral load may be low, sometimes a false negative occurs. That means there IS virus, it’s just not there in high enough numbers for the swab to catch it. This is risky from a disease control perspective, because the person thinks they’re in the clear (See White House testing).
The upside to this test is that if it IS positive, it’s a pretty clear indication you have a current, active infection. While PCR is the “gold standard,” it is so sensitive that it can pick up trace amounts of viral RNA long after a person is no longer infectious.
FDA Approval
But wait. Many antigen tests are only FDA “approved” for symptomatic individuals. This is because they can only guarantee the results with pretty good certainty IF the person has symptoms. If they don’t have symptoms and they’re infected, viral load is probably lower, and they don’t show up positive, even though they’re infected. And since a false negative is more likely with asymptomatic people, the FDA doesn’t recommend relying on antigen tests for someone suspected of having COVID due to contact if they don’t also have symptoms.
But sometimes asymptomatic people have high viral loads, especially kids.
So can these tests be used for people without symptoms? The FDA says yes. It’s up to their health care provider to decide.
What does “Off label” use mean?
Here’s an example. The HPV vaccine was first approved for females aged 9–26. Because most women are infected by the age of 26 and because the vaccine was targeting the main cause of cervical cancer (which only affects women), it was studied in this age group and licensed for this population.
But that doesn’t mean a 27 yr old woman can’t get it. If her health care provider deems her to be at risk, or that she may not yet be exposed, the HCP can give her the vaccine. Insurance may not pay, but such “off label” use was allowed.
Similarly, a 22 yr old man could get the vaccine. In fact, once they realized that 1. girls got infected from boys and 2. it likely prevented male cancers, more gave it, even though the FDA hadn’t authorized it for use in men at the time.
COVID Antigen tests work the same way. A clinician can decide if the individual warrants an antigen test. If it’s positive, and you’re in an area of high transmission, assume it’s positive. You can confirm it with a PCR within 48 hours if in doubt.
The FDA is wary of negative results when someone is asymptomatic, so recommends not relying on the antigen test if you have reason to suspect they’re infected. Instead, the FDA (and CDC, and DSHS) recommend confirming negative results with a PCR.
CDC, FDA, DSHS, APHL as well as many scientists around the country, agree that in areas of high transmission, a positive is a presumed positive is a positive and should be treated accordingly.
APHL
The Association of Public Health Laboratories makes recommendations to the Council of State and Territorial Epidemiologists. On September 24, the APHL** issued a statement on antigen tests:
- Rapid Antigen tests have high specificity and lower sensitivity. (low false positive and higher false negative)
- In populations with a high positivity rate, positive Ag test results indicate that individuals are infected and presumed to be infectious.
- A false positive is more likely to occur when the overall positivity rate in a population is low. If either a false positive or false negative is a concern, these should be verified by PCR test.
The APHL explains that Rapid Antigen tests are best used in screening situations when a quick result is necessary for infection control measures. However, in communities with high transmission, the likelihood there’s a false positive is small.
In sum, if a community has active transmission and there is reason to believe an individual is infected (symptoms or contact or living in a high transmission area), assume a positive result to be accurate and proceed accordingly. If in doubt, follow up with a PCR test.
The old adage, “if it walks like a duck and quacks like a duck, it’s probably a duck” applies. In Bexar County today, and in most cities in the US, there is sufficiently high SARS-CoV-2 transmission to err on the side of caution: accept a positive antigen test as positive, regardless of symptoms.
*There are plenty of topics for which I can explain the basics, but lack sufficient knowledge to explain more thoroughly. So I’ll send you to my biochemist colleagues (my micro students can attest to that).
**According to APHL, the tests reviewed, including the one used most commonly in San Antonio, have >97% true positive rates.
Update:
Nevada announced on Oct 2 they would be halting its use of Antigen tests in nursing homes following an analysis that found a 60% false positive rate using two tests (Quidel and BD). This is the first report to indicate anything more than a 3% false positivity rate and is worth further investigation. Notes to keep in mind:
* Sample size = 39 — very small. Most studies have sample sizes > 100 (still on the small side but better)
* Positivity rate = 0.4% — 1.6% (depending upon which results you believe) — very low. Recommendations are for use in populations with high transmission. We know we’re likely to end up with more false positives when infection is unlikely. This is why we don’t typically run Rapid Flu tests in July — a positive in July isn’t likely accurate. But a Rapid Flu test in January is. You don’t look for zebras in Texas, or bears in Sub-saharan Africa.
*Timing — An antigen test is only going to be positive for a few days while the person is most infectious (see graph from the Mina Lab). CDC and FDA recommend confirmatory PCR within 24–48 hours. The Nevada Division of Public Health state it’s possible some of those 39 tests were not repeated within the 48 hour window.
*User error — this is always a potential problem. “Possible reasons for conflicting test results include lack of compliance with the manufacturer’s protocols; inadequate training on the testing procedure…” The tests are simple, but do require training and caution in doing them correctly. With such a small sample size, minor user errors can lead to significant data discrepancies.