From the beginning of this pandemic, scientists around the world have been working around the clock in pursuit of answers that can effectively combat the SARS-CoV-2 virus. One of hardest things for people to grapple with, is all the unknowns: When will this end? When can I safely visit my friends and family again? What if I have it or had it and I don’t even know it?  

The increased availability of serological testing has helped ease people’s minds about their personal COVID-19 status. From a distance, serological testing may seem like a huge milestone in the marathon that is this pandemic. Unfortunately, many of these tests provide murky and inconsistent results.

Molecular vs. Serological Testing: What’s the difference?

When it comes to infectious diseases, molecular and serological are the two primary types of testing. Molecular tests facilitate detection of a pathogen while it is still circulating in the body, even picking up on fragments of the viral material waiting to be cleared from the body that may no longer be pathogenic. Up until more recently, most of the testing being conducted worldwide for SARS-CoV-2 was focused on molecular testing, employing reverse transcriptase-polymerase chain reaction (RT-PCR) to diagnose and confirm cases for both surveillance and clinical treatment purposes (1).

These molecular diagnostic tests lack a key ingredient in the recipe for successful long-term outbreak response: they are unable to detect those who were previously infected and cleared the infection. So while it is obviously important to be able to diagnose those who are currently infected, to proceed safely in the direction of societal “normalcy” until we have a vaccine, we need to know who has already been infected and recovered. This is where serological testing comes in.     

Serological tests (also known as antibody tests) pick up where molecular tests leave off by measuring the body’s immune response to pathogens, providing insight into both current and prior infections. The tests use blood samples to detect the presence, or absence, of antibodies produced in response to the virus (2,3). In the case of SARS-CoV-2, these serological tests are especially important in helping to better estimate prevalence, because not everyone who has had COVID-19 had the option or opportunity to be tested before the virus left their body, and it’s estimated that as many as 25% of infections are asymptomatic (4,5).

Antibodies Don’t Always Equal Immunity

Antibodies, or immunoglobulins (Ig), are produced by B cells and play a crucial role in a highly specific defense against new antigens, especially those from bacteria, fungi and viruses (6). Two classes of antibodies in particular, immunoglobulin M (IgM) and immunoglobulin G (IgG), are the most common targets of serological assays, due to the roles they play in both the targeting and destruction of new infections. Around 7 days post-disease onset, about 50% of patients begin developing antibodies that specifically react to SARS-CoV-2 proteins (3,7).

IgM antibodies are initially produced as the first line of defense against the virus. Current studies demonstrate that they appear to peak around 12 days following infection, maintain an ample supply for up to 35 days but then rapidly decline (7,8). IgG antibodies, which are generated later and serve more as a long-term immune response, have been observed to peak around 11-17 days after SARS-CoV-2 infection (7) and linger for up to 49 days (which is when the cited study concluded; 8). Depending on the disease it’s protecting you from, IgG can typically linger in the blood and provide immunity anywhere from months to a lifetime, but it’s hard to say for sure how long IgG antibodies last for SARS-CoV-2.

At this point, our best comparative models for estimating SARS-CoV-2 immunity are SARS-CoV-1 (the virus that causes SARS and whose genome is ~76% similar to SARS-CoV-2) and some other coronaviruses that can cause the common cold. In the former, IgG antibodies have been found in recovered SARS patients for up to 3 years (9); however, the latter seems to possess a considerably shorter immunity though the data are limited (10).

There are still too many unknown variables in this immunity equation. We don’t have the full picture of what human immunity to SARS-CoV-2 looks like or how long it lasts; immune responses between patients vary far and wide, and it’s not clear why. It’s additionally unclear if a person might still be a source of contagion to others even after symptoms disappear and IgG antibodies are developed, or if the human immune response to SARS-CoV-2 prevents you from being re-infected later on (11). There’s also uncertainty surrounding the cross-reactivity potential of the COVID-19 serology tests, indicating the possibility that antibodies from other coronaviruses that the patient has previously been exposed to might be flagged in addition to those from SARS-CoV-2 infections.

Sensitivity and Specificity: A Testing Tug of War

Under normal circumstances, the U.S. Food and Drug Administration (FDA) performs an internal quality check before antibody tests are allowed on the market. However, given the escalating urgency of the situation, the FDA initially waived the internal review of the COVID-19 antibody tests, which has resulted in over 120 labs and manufacturers flooding the market with tests that lack FDA review and validation, and wildly vary in accuracy (12,13). And even for those tests that are accurate, interpreting the results is difficult because of the unknowns listed above. 

In response, the FDA has stepped up their oversight of the assays being brought to the market and have now provided a summary of the expected performance of the tests it has now authorized (14).

In situations like this, it’s especially helpful to understand how test accuracy is determined, and what the percentage of accuracy actually means. Serological tests are evaluated based on two key accuracy metrics: sensitivity and specificity.

Sensitivity is the test’s ability to give a positive result when it’s supposed to be positive, or the likelihood that your test will detect the desired target. A highly-sensitive test should detect almost everybody who has the disease and have a low rate of false-negative results (e.g., the test indicates the absence of antibodies when they exist; 15).

Specificity is the test’s ability to indicate a negative result when it is supposed to be negative, or the likelihood that the test won’t be tricked by something other than the desired target. A highly-specific test should correctly rule out almost everybody who does not have the disease and have a low rate of false-positive results (e.g., the test indicates the presence of antibodies when they don’t exist; 15).

Sensitivity and Specificity Decoded  

How sensitive is the test?
Translation: How many actually COVID-positive people does it correctly identify as COVID-positive?

                How specific is the test?
                Translation: How many COVID-negative people does it correctly identify as COVID-negative?

                What is the false-negative rate?
                Translation: How many people who were actually COVID-positive were told that they were not?

                What is the false-positive rate?
                Translation: How many people who were not actually COVID-positive were told that they were?

The rates of sensitivity and specificity have a generally inverse relationship: increasing sensitivity can lower the rate of specificity while a more specific test will often not be as sensitive. Casting a wider net can snag more targets including more than just the one you want, while a smaller net might snag the only target type you’re after, but not catch them all.

This relationship between sensitivity and specificity contributes to the difficulty of creating a serological test that is both highly-sensitive and highly-specific which, in an ideal situation, we would want all of our antibody tests to be. So what percentages of these rates should we be on the lookout for? According to Robert Garry, a Tulane University virologist, “Sensitivity and specificity rates of 95% or higher are considered a high benchmark, but those numbers are difficult to hit; 90% is considered clinically useful, and 80 to 85% is epidemiologically useful” (10).

However, the unfortunate truth is that even those remarkably high, positive-sounding numbers are not as helpful or positive when applied in real-time testing of a large population. Let’s break it down:

In this scenario, let’s pretend 5% of the population has been infected with COVID-19. In theory, if you serologically test one million people randomly, testing should yield 50,000 positive results and 950,000 negative results. But let’s say that the test you’re administering to those one million people is only 95% sensitive and 95% specific. That test would correctly yield 47,500 positive results and 902,500 negative results, but also saddle 50,000 people (unknowingly) with a false result. Of those 50,000, 2,500 people would receive a false-negative result (when they are actually positive) and 47,500 people would receive false-positive results (when they are actually negative) (10).  

Unintended Consequences

As the above example illustrates, it’s extremely important for everyone to understand the limitations and implications of these seemingly high-accuracy tests, especially people who are using the results of commercially-available antibody tests to guide their own health behaviors and decisions.  

It’s also important to understand that the consequences of testing errors are not equivalent, in terms of false positives versus false negatives. While a false negative could prevent a person from returning to work (which is objectively bad), a false positive could lead to a full-fledged epidemic chain reaction, ignited by a single person who’s test results led them to believe that their (false) immunity status entitles them to forego preventative health measures (which is objectively much, much worse).

Serology tests are potential sources of incredibly valuable information for public health authorities that can help them better estimate the incidence of infection as well as continue to guide public health control measures, like social distancing (16). In order for serologic tests to live up to this potential, it’s vital that three critical areas of uncertainty be cleared up in the weeks to come.

The most pressing is the need for serology test validation, which is already underway; it’s crucial to ensure that the available tests are comparable, accurate and overseen by the FDA. Secondly, though most experts assume COVID-19-recovered patients will possess some degree of immunity, additional research into whether a correlation exists between a specific antibody level and immunity should be pursued. And finally, it will be beneficial to determine if protective immunity is induced by infection, and if so, how long it can last (17).

Visit our website for resources to support:
SARS-CoV-2 Viral Research
SARS-CoV-2 Serology Testing

References

1. Antigen and Molecular Tests for COVID-19. Johns Hopkins University Center for Health Security.
2. Loeffelholz, M.J. and Tang, Y.W. (2020) Laboratory diagnosis of emerging human coronavirus infections – the state of the art. Emerg. Microbes Infect. 9(1):747–756.
3. Cevik, M., Bamford, C. and Ho, A. (2020) COVID-19 pandemic – A focused review for clinicians. Clin. Microbiol. Infect. S1198-743X(20)30231-7. [Published online ahead of print]
4. Mizumoto, K. et al. (2020) Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020. Euro. Surveill. 25(10): 2000180.
5. Lai, C.C. et al. (2020) Asymptomatic carrier state, acute respiratory disease, and pneumonia due to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2): Facts and myths. J. Microbiol. Immunol. Infect. S1684-1182(20)30040-2. [Published online ahead of print]
6. Justiz Vaillant, A.A. and Ramphul, K. (2020) Immunoglobulin. StatPearls.
7. Long, Q-X. et al. (2020) Antibody responses to SARS-CoV-2 in patients with COVID-19. Nat. Med. [Published online ahead of print]
8. Tan, W. et al. (2020) Viral kinetics and antibody responses in patients with COVID-19. Preprint at medRxiv.  
9. Wu, L-P. et al. (2007) Duration of antibody responses after severe acute respiratory syndrome. Emerg. Infect. Dis. 13(10):1562–1564.
10. Patel, N. (April 9, 2020) Why it’s too early to start giving out immunity passports. MIT Technology Review.
11. Su, A. (March 13, 2020) They survived the coronavirus. Then they tested positive again. Why? Los Angeles Times.
12. Brennan, Z. and Lim, D. (April 27, 2020) FDA pushed through scores of inaccurate antibody tests without agency review. Politico.
13. McCoy, K. and Heath, D. (May 8, 2020) Coronavirus antibody tests are available around the country. Here’s why they may provide a false sense of security. USA Today.
14. EUA Authorized Serology Test Performance. U.S. Food and Drug Administration. Accessed May 14, 2020.
15. Understanding medical tests: sensitivity, specificity, and positive predictive value. HealthNewsReview.org
16. COVID-19 Serology Surveillance Strategy. Centers for Disease Control and Prevention.
17. Developing a National Strategy for Serology (Antibody Testing) in the United States. Johns Hopkins University Center for Health Security.

For more information about SARS-CoV-2 PCR and Serological Testing, please visit our website.

Related Posts

Updated 2/20/2021

The following two tabs change content below.

• Bio
• Latest Posts

Natalie Larsen

Science Writer

Natalie is a Science Writer at Promega. She earned her B.S. in Microbiology from the University of Wisconsin-Madison, and her Associate's Degree in Science from Cottey College. In her spare time, she can be found playing volleyball, making music, chipping away at her never-ending stack of craft projects, and volunteering with animals.

Latest posts by Natalie Larsen (see all)

• The Purr-suit of Happiness: The Science Behind Feline Purring - February 9, 2023
• Making “Scents” of the Mysterious Science of Plant Odors - July 19, 2022
• Empowered Women Empower Women: Closing the Confidence Gap in STEM - March 31, 2022