Testing for COVID-19 can be limited in efficacy and accuracy — even lab-quality, reverse transcription polymerase chain reaction (RT-PCR) tests, considered the gold standard for detecting the virus. However, when armed with the appropriate information about the various strengths and weaknesses of tests, frontline providers are better positioned to make appropriate decisions on how to manage patients.
For example, while RT-PCR tests are more accurate than rapid antigen tests when it comes to detecting cases, the sensitivity of these tests in practice is only about 80%, according to Daniel Rhoads, MD, section head of microbiology at the Cleveland Clinic. A good bit of this drop-off in sensitivity concerns timing. “The one variable to identify in your initial clinical history is the time from symptom onset to testing,” explains Rhoads, vice chair of the microbiology committee of the College of American Pathologists. “If it has been a long time from symptom onset to testing, the sensitivity decreases. We know that because typically there is a lot of shedding in the first several days of the virus. Then, as the patient gets further away from that initial symptom onset, the amount of virus that is being shed is decreasing.”
Consequently, this variable is important for clinicians to consider in a case in which the RT-PCR result is unexpectedly negative for COVID-19, but there is strong suspicion the patient is positive based on other factors. In such a case, Rhoads says it is reasonable for clinicians to conclude the patient is far off from symptom onset — but he or she still has COVID-19. If clinicians remain unsure, imaging may help them reach a diagnosis.
Another important variable is what Rhoads terms pretest probability. This is essentially the prevalence of the virus in the community. When disease prevalence increases, the positive predictive value of a test also increases because there will be fewer false-positive results.
“The prevalence [of COVID-19] has swung drastically over short courses of time, going from 40% positive to 20% positive in testing. It is important to consider prevalence when evaluating patients,” Rhoads says.
Rhoads predicts COVID-19 will ease into a seasonal pattern similar to that of influenza. If that happens, “the prevalence in the community should heavily influence clinical suspicion: not suspecting it when it is not circulating, and being skeptical of a negative result when the prevalence in the community is high,” he says.
Although most EDs can access RT-PCR testing, is there a place for using rapid antigen tests in the ED? “If the goal is the quickest turnaround possible, then rapid antigen tests are typically faster than RT-PCR,” Rhoads says. “If the goal is the most sensitive test possible, then RT-PCR is more sensitive than antigen testing.”
For example, Rhoads notes many organizations are using the rapid antigen tests for screening purposes. When there is a high prevalence of COVID-19 in a community, EDs may want to test everyone who presents for care to prevent virus transmission. However, clinicians must interpret the results of a rapid antigen test in line with guidance from the CDC. “If you are using the antigen test for screening, and a result is negative, the probability is very high that the patient is not shedding [the virus],” Rhoads offers.
If the patient is symptomatic of COVID-19, but tests negative, Rhoads recommends ordering a nucleic acid amplification test, which is more sensitive and could indicate whether the antigen test result was a false-negative. “Any time an antigen test is positive, the CDC recommends accepting that this result is likely to be true,” Rhoads says.
When determining which test to use, start by identifying goals. “Think about what would be the risks of a false-positive or false-negative result,” Rhoads says. “Determine what is most important, and then make decisions based on those variables.”
W. Frank Peacock, MD, FACEP, professor, research director, and vice chair of research at Baylor College of Medicine, led research into the performance of one commonly used antigen test, BinaxNOW.1 Out of group of 735 participants who presented to the ED for care, 623 reported COVID-19 symptoms such as fever, shortness of breath, or coughing. All participants underwent rapid antigen tests and RT-PCR tests for comparison.
Ultimately, the RT-PCR tests delivered positive results in 173 cases, while the rapid antigen tests delivered positive results in 141 cases. Peacock and colleagues found that among the patients who reported experiencing symptoms for more than two weeks, the positive rate was about half that of those who presented for care and testing earlier in the course of their illness. Further, among patients who reported experiencing symptoms for less than seven days, the sensitivity of the antigen test was 84.6% and the specificity was 98.5%.
From these findings, the researchers recommended the use of antigen tests for patients who have had symptoms for up to two weeks. However, data collection for this study occurred between July and November 2020, before virus variants were widespread in the United States. “Whether [BinaxNOW] will perform equally with different variants is unknown,” the authors wrote.
Peacock says the sensitivity rate of the antigen test means one out of six cases may go undiagnosed, but performance improves with repeated testing. “If you do the test two days in a row, the probability of missing the diagnosis comes out to be one in 13,” Peacock reports. He adds that if clinicians were to perform the test a third time, the probability of missing a case “is so low there is no point in continuing to do it. With a single test in somebody who is not tremendously symptomatic, if the result is positive you are done. If it is negative, you have to think that maybe you should do a test tomorrow also.”
Considering that testing over multiple days is not practical in the ED, clinicians with access to relatively rapid RT-PCR testing probably will opt for the more sensitive test. However, the antigen test offers a good option to urgent care centers or EDs that lack access to a sophisticated lab.
“Whenever you do a point-of-care test, you make a sacrifice on sensitivity. That applies to a lot of point-of-care tests,” Peacock says. “What you get out of it is speed and ease.”
As with the RT-PCR test, an antigen test works best early in the illness course, when the viral load is high. “The rapid antigen tests aren’t able to detect really low levels of antigens, so you want to [perform the test] during the sweet spot. If a patient comes in and tells you they have been sick for two days, that is the sweet spot.”
There have been some conflicting observations about the ability of asymptomatic people who are COVID-19-positive to spread the virus to others. “What we are pretty sure about is that if someone has some level of immunity and they are infected with [the virus], generally their symptoms are less severe, the duration [of the illness] is shorter, and the amount of virus shedding is either less or [the viral shedding is of] shorter duration,” Rhoads explains. “If someone is asymptomatic, it is probably because there is some level of immunity. The duration of their viral shedding is probably going to be shorter, which should theoretically decrease the opportunity to spread the infection. But that doesn’t mean that they can’t [spread the virus] or that they have not or will not. It just decreases their risk.”
Peacock says as he and colleagues were collecting data on antigen tests, they also were trying to evaluate asymptomatic people, which was difficult. “We found that the rate of positivity among asymptomatic patients was 1%, so that means you would have to test a boatload of people to find one positive result,” he says. “With omicron, there are a lot more people [because] it is a much more contagious [variant]. The rate may change, but asymptomatic people usually don’t have any disease. It is a very difficult population to screen for because there are so many negatives that are true.”
It is fortunate that most approved COVID-19 tests have continued to work despite the spread of multiple variants. “There are hundreds of tests that received the EUA [emergency use authorization] from the FDA, and a couple of them have been found to not be able to test for omicron because of the mutations present,” Rhoads says. “But most tests, from my perspective, are performing similarly to how they were performing a year or two ago.”2
How do the COVID-19 tests stack up against the tests clinicians use regularly for other common respiratory diseases? Rhoads says this is difficult to assess because testing generally is used as a reference standard for respiratory syncytial virus, influenza, and similar maladies. “We don’t really know how many cases are missed because everyone relies on the test as a diagnostic piece,” he says. “They typically take the lab result as the truth.”
REFERENCES
- Peacock WF, Soto-Ruiz KM, House SL, et al. Utility of COVID-19 antigen testing in the emergency department. J Am Coll Emerg Physicians Open 2022;3:e12605.
- FDA. SARS-CoV-2 viral mutations: Impact on COVID-19 tests. Dec. 28, 2021.
Considering that testing over multiple days is not practical in the ED, clinicians with access to relatively rapid RT-PCR testing probably will opt for the more sensitive test. However, the antigen test offers a good option to urgent care centers or EDs that lack access to a sophisticated lab.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content