In this series of COVID FAQs, the SciMoms answer frequently asked questions about the coronavirus or COVID-19 pandemic. In this post, we focus on diagnostic testing. What is a COVID-19 test? How accurate are COVID-19 diagnostic tests? Why is testing important?
What is a COVID-19 test?
A COVID-19 test is when a biological sample, such as saliva, a nasal swab, or a blood sample, is taken from an individual and that sample is tested for signs of COVID-19. There are three important roles that testing provides: diagnostic, screening, and surveillance.
Diagnostic testing helps to identify current infections in people with symptoms or people with a recent, known exposure to SARS-CoV2. Diagnostic testing is used to test symptomatic people, people who have been identified through contact tracing, and people who report that they have been exposed to a known or suspected case of COVID-19.
Screening testing is used to identify infected but asymptomatic people who have no known or suspected exposure to SARS-CoV2. The purpose of this testing is to identify people within a group who may be contagious to prevent further transmission of the virus. Screening usually involves testing a group of people who congregate, regardless of symptoms, in settings such as long-term care facilities, workplaces, or schools.
Surveillance testing refers to testing for the purposes of ongoing, systematic collection of data for planning and implementation of public health policies. The purpose of surveillance testing is to monitor the level of infection in a community or population, and to determine the incidence and prevalence of disease. This type of testing is usually done on de-identified samples (results that are not linked to people) and does not involve reporting a diagnosis to individuals.
What is contact tracing?
Contact tracing is one of the oldest public health tools to track infections (the canonical example is the 1854 cholera outbreak in London). It has been used successfully to contain other infectious diseases. The system works by retracing a patient’s steps to determine where they’ve been and who they’ve been in contact with.
As soon as someone receives a positive test or is presumed to be positive, that person is immediately isolated and contact tracers work to identify everyone that person came into contact with in recent days. Individuals that may have been exposed are then contacted so that they can quarantine to contain the spread of COVID-19. This allows tracers to determine the network of disease transmission.
There are certain characteristics of COVID-19 that make contact tracing more difficult for this disease than for others. These are similar to the very characteristics that made SARS-CoV2 able to become a pandemic (see our Intro to COVID-19 FAQ for details). These were explained well on the Dear Pandemic Facebook page (run by The Nerdy Girls, a group of epidemiology professors):
- Because people are contagious even BEFORE developing symptoms, tracing (identifying and notifying contacts) needs to happen quickly. But tracing takes time.
- There is also currently no treatment to offer to individuals who test positive, so there is less personal incentive to cooperate. Instead, contacts are asked to quarantine, not for their own benefit, but to reduce exposure to others.
- The period of quarantine, 14 days from the time of exposure, is a long one. This poses challenges in terms of time off from work, especially for those who cannot work from home.
- For contact tracing to succeed, there have to be enough tests to keep up with the pace of infection (and ideally, a faster turnaround for results). Few places in the US have met this goal.
In light of these challenges, cooperating with contact tracers in a timely manner is critical. If you are contacted by a contact tracer, it is very important to cooperate, comply with quarantine, and get tested if they request it.
Why is testing important?
Since the start of the pandemic, health agencies around the world have been pushing for coordinated programs of testing and contact tracing to combat the SARS-Cov2 pandemic.
Without adequate testing, we are not able to test exposed individuals in a timely manner before they spread the disease. And we are blind to the very information we need to develop appropriate public health measures to slow the spread of SARS-CoV2.
The personal actions we can each take to protect ourselves and our communities — the 3Ws (wear a mask, wash your hands, watch your distance), quarantine after suspected exposure, and avoid gathering indoors or in large groups — can only go so far. A strong national response must include coordinated programs of testing and contract tracing. These programs are arguably the most important tools we have to combat this pandemic. Without testing and tracing, we are left with only blunt public health measures, like lockdowns, to slow the spread, which is why the COVID response in the US has been reduced to a binary choice between reopening or stopping the pandemic.
Testing can help reduce the need for lockdowns
Lockdowns are a blunt tool for slowing the spread of COVID-19. Ultimately, lockdowns are not sustainable but, with testing and tracing, officials can implement targeted lockdowns while still minimizing risk. Implementation of coordinated testing and contact tracing, in combination with widespread adoption of the 3Ws, would allow us to control the spread without the need for prolonged lockdowns. These measures are critical to slowing the spread and controlling the pandemic in the absence of a vaccine or effective treatments. (Visit our FAQs on COVID-19 spread and masks for more info).
Take the following analogy: a disease outbreak is like a leak somewhere in a building. A lockdown is akin to shutting down water to the entire building in order to stop the flow of water. Shutting off water buys you time so that you can prevent damage to the property, reduce harm to the residents of the building, and identify the location of the leak. But with the proper tools (electronic sensors, ultrasonic detectors, etc), you can pinpoint the location of the leak without having to shut off water to the entire building. You might even be able to prevent leaks before they happen. In this analogy, COVID-19 tests are one of those tools.
Testing and tracing is easier and more effective when cases are low, which we can see when we look at the success of countries that implemented testing and tracing early on. With these programs in place, they have been able to both get and keep spread under control. Testing and tracing become harder when there are many cases and community spread is high. In this scenario, even more testing is required. Lockdowns might become necessary when testing and tracing are not keeping pace with the spread.
What types of tests are available?
Roughly speaking, there are diagnostic tests and antibody (or serological) tests. Diagnostic tests are used to determine if a person is currently infected with SARS-CoV2. Antibody (or serological) tests are used to determine if a person was previously infected with SARS-CoV2, by testing for antibodies to SARS-CoV2 in the bloodstream. Here, we will focus on diagnostic testing.
Diagnostic tests look for material from the SARS-CoV2 virus in a sample on the day of testing. These tests can be done from samples collected from nasopharyngeal swabs, nasal swabs, or saliva. Diagnostic tests include: molecular tests (typically PCR-based tests) that measure whether there is RNA from the virus SARS-CoV2 in the sample and antigen tests that measure whether there are viral proteins from the virus SARS-CoV2 in the sample. (See our FAQ on coronavirus basics for more info on coronavirus proteins and structure.)
How do diagnostic tests work?
Molecular or PCR tests for SARS-CoV2
The novel coronavirus (SARS-CoV2) is an RNA virus (see the our FAQ on coronavirus basics). Most currently available diagnostic tests for SARS-CoV2 use PCR to test for the presence of the virus’ RNA.
Once a sample (nasopharyngeal, nasal, or saliva) is collected, scientists extract RNA from the collected sample. Since the amount of virus that is obtained from the patient is usually very small, most currently available molecular tests rely on the polymerase chain reaction (PCR). PCR makes many copies of specific segments of the virus’ genetic material. The resulting amount of material at the end of the reaction is used to determine how much material was present in the original sample.
PCR works on DNA and SARS-CoV2 is an RNA virus. So scientists must use a process known as reverse transcription-polymerase chain reaction or RT-PCR. This makes DNA from the RNA before copying.
PCR-based diagnostic tests for SARS-CoV2 test for multiple specific pieces of viral RNA. Testing for more than one segment of the RNA provides greater accuracy.
qPCR, which stands for quantitative PCR, allows scientists to measure how much DNA is present in real-time by measuring light emitted by fluorescent dyes that bind to the DNA. The more DNA that is present, the greater the amount of light emitted at each cycle. Most diagnostic tests for the novel coronavirus SARS-CoV2 use a combination of these methods: quantitative reverse transcription PCR.
Antigen tests for SARS-CoV2
Antigen tests can also be used for diagnosis of COVID-19. These tests are generally faster and cheaper, but less accurate, than PCR-based tests. These tests look for proteins on the surface of the virus instead of viral RNA. (See Intro to COVID-19 FAQ for viral basics). Like PCR-based tests, antigen tests can be used to determine if someone is currently infected with SARS-CoV2.
These tests use similar technology to a rapid strep test or at-home pregnancy tests. They work by using antibodies to determine if proteins in the SARS-CoV2 viral coat are present in a sample. This is different from an antibody test in that it does not test for antibodies in a sample (antibody tests will be covered in the next FAQ).
In these tests, a sample from a nasopharyngeal or nasal swab is placed on a small piece of film or membrane. This film contains antibodies against viral proteins at a specific location on that film. These antibodies are attached to molecules that produce a signal when the virus sticks to the antibody. As the sample travels through this film or membrane, if the antigen is present it will stick to those antibodies and form a visible line.
How accurate is current diagnostic COVID-19 testing?
Usually, a new diagnostic test is compared to the gold standard diagnostic test for that disease. However, when a disease is new, there is no pre-existing test to compare to the new tests. Scientists must estimate these numbers based on the technical properties of the test and real-world clinical data. Since SARS-CoV2 is a new virus and there is no previously developed gold standard test, data on the accuracy of the tests has been developing as the pandemic progresses. Without a standard, it was difficult to get clear answers on accuracy. As scientists learn and more tests are performed, they can get a better and more reliable estimate of a test’s accuracy.
There are many metrics that scientists use to evaluate how accurate a test is. These terms are being frequently used to talk about COVID-19 testing in many discussions of testing, so we’d like to go over the basics here.
For diagnostic testing, there are two possible test results: positive or negative. But sometimes, test results might be inaccurate. This means there are four possible outcomes.
- True positive: Sick people correctly identified as sick
- True negative: Healthy people correctly identified as healthy
- False positive: Healthy people incorrectly identified as sick
- False negative: Sick people incorrectly identified as healthy
Using these numbers, scientists can calculate how well a test will work as a diagnostic, screening or surveillance tool. A test with high false positive or false negative rates would not be very useful on its own for diagnosis. However, surveillance testing might not require as accurate a test if testing is frequent enough. Some of the terms you may see in reports about testing include:
- Specificity: the percent of negative samples that are correctly identified. This tells us how well a test can correctly identify people without the disease. A test with high specificity will not mistakenly identify healthy people as sick.
- Sensitivity: percent of positive samples that are correctly identified. This tells us how well a test can correctly identify people with the disease. A test with high sensitivity will not miss many people who are sick.
- Accuracy: percent of total samples that are correctly identified as positive or negative. This measure gives an overall estimate of how well a test can correctly identify people with or without the disease.
Accuracy of molecular or PCR tests for SARS-CoV2
Because PCR is based on the unique genetic sequence of SARS-CoV2, it has high specificity and sensitivity. From a technical perspective, false positives and false negatives due to the technology are rare. This is explained very well in this article in Science:
“Genetic tests use the polymerase chain reaction (PCR) to amplify tagged DNA or RNA sequences, making it easy to reliably identify just a few copies of a virus. That gives PCR tests for the SARS-CoV-2 virus about a 98% sensitivity and near perfect selectivity, meaning almost every active infection is detected and only in very rare cases does someone uninfected receive a positive test.”
However, there are two major issues that can lead to these false negatives in practice: poor sample quality and the timing of testing.
Poor sample quality: A PCR test may fail if insufficient quantity or quality of RNA is collected from a sample. Inclusion of controls and quality control steps can help to determine if a test is negative due to sample failure. This would be reported as inconclusive and not as a negative test result.
This is why nasopharyngeal swabs have been the preferred method for sample collection. The concentration of virus recovered from these samples is higher than those collected from nasal swabs or saliva, so when done properly this method ensures that sufficient quantities of virus are collected. However, it is possible to get enough RNA from nasal swabs or saliva samples and PCR tests have been developed that work from these samples. The FDA recently approved an Emergency Use Authorization for SalivaDirect, developed at Yale, and other labs (including my department at Michigan State Univeristy) around the country have been developing similar protocols.
Timing of sample collection: Because viral load changes throughout the course of disease, the timing of testing is critical. If a test occurs too soon after exposure, the test may be negative because viral load is still too low to detect, even though the person is actually infected. If a test occurs too late after symptom onset, the viral load may be too low as the virus is cleared from the body. Early data suggested false negatives might be as high as 20%, even on the most accurate day of testing. This is consistent with what we know about viral load and disease courses. This high false negative rate is why quarantine is still recommended after potential exposure or before merging quarantine bubbles.
Accuracy of antigen tests for SARS-CoV2
Antigen tests are faster, cheaper and more amenable to point-of-care use than most molecular tests, but accuracy is highly variable. Because the accuracy of these tests is so variable, results must be interpreted with caution. As with other types of antigen tests, a negative result on a rapid test is often followed up with a more accurate diagnostic test to make sure it is truly negative. Even a positive result on these tests is often considered “probable” until follow-up with PCR is done because of the false positive rate.
Recently, two new rapid antigen tests (Xpert Express and ID Now) were approved by the FDA via Emergency Use Authorization. Preliminary findings show that these two tests are actually quite accurate when compared to the PCR tests (see Table 2 in the link for all the nerdy details).
Some scientists think that antigen tests may be very useful for surveillance testing because they are cheap and fast. Frequent, repeated testing regardless could make up for the inconsistent accuracy.
What does a negative diagnostic test mean?
A negative diagnostic test means that the sample that was tested did not have detectable levels of virus. However, as discussed above, given the relatively high rate of false negatives and the timing of the disease course, it is not that simple to interpret what this means.
This graphic from the “Dear Pandemic” Facebook page shows the timeline of COVID-19 infection. The variability in the incubation period is part of what leads to many of the false negatives.
A negative test cannot definitively tell you that you will not be contagious at any point after your sample was collected. You could have been infected close to the day of testing so your viral load was too low for detection on the day of testing. Thus, a negative test should not be taken as license to relax physical distancing measures, like the 3Ws (wear a mask, wash your hands, and watch your distance).
Relaxing physical distancing as a result of a negative result can have severe consequences. For example, someone who is contagious might feel confident to continue activities in the community and could infect many others. Therefore, due to the high false negative rate, a single negative result should not be used to justify relaxing your protective measures.
Overall, this means that a positive test is useful for knowing if someone is infected with SARS-CoV2, but a negative result is not as useful for confirming that someone is not infected. A negative result alone should not be taken as license to forgo quarantine after a suspected or known exposure.
When should you get tested?
After a potential exposure or after traveling:
Because of the variability in the incubation period, it is difficult to set a specific day that is the best day to get tested. Thus, after a potential exposure to COVID-19, a 14-day quarantine is the recommended course of action. This covers the entire 2-14 day range of the incubation period for this virus. Testing immediately after exposure may not provide enough time for the virus to replicate to a level high enough to detect.
The best timing would be as late as possible in the 14-day window. However, you may have to break quarantine to get tested, and if you have properly quarantined, by the time you get your test result, your quarantine period will likely be over. It is critical to remember that a negative result at any time during the incubation period does not provide a shortcut to this quarantine period. Thus, a 14-day quarantine following exposure is the recommended course of action.
As mentioned above, one of the challenges for contact tracers in this pandemic is the lack of individual incentives for testing. If you are sick, you will call a doctor regardless of whether you had a positive test and there are few treatments available for COVID-19. If you are negative, you still need to quarantine because it could be a false negative. So the personal incentive to get testing is low because the result has little effect on recommended behavior. Despite this, for the health of your community, testing in combination with proper 14-day quarantine remains critical for surveillance and contact tracing. A positive test is helpful for contact tracers to trace the path of the virus, but a negative test does not clear the contact from the need to quarantine.
Before traveling or gathering with people outside your household:
To gather safely, without physical distancing, all individuals must have no exposure for the entire 14 day period before the date of the gathering. Similarly, prior to travel, to ensure that you are not spreading SARS-CoV2, you must have no exposure for the entire 14 day period before the date of the gathering
This means avoiding any possible exposures for a full 2 weeks prior to gathering or traveling. People could also potentially get tested as late as possible in that 14-day period to have results before the date of the gathering (i.e. if testing turnaround is 3 days, get tested on Day 11) to make sure you do not have an asymptomatic infection. All of the caveats discussed about accuracy and false negatives apply to these scenarios as well. A negative does not replace a the full 14 day quarantine. Quarantine is the only way to guarantee safety for a non-physically distant gathering or to ensure that you are truly virus-free before traveling.
If you cannot quarantine for a full 14 days before or after travel, gathering with people outside your household, or other potential exposure, it is critical to maintain good preventive practices like the 3Ws (wear a mask, wash your hands, and watch your distance).
Are rising numbers of COVID-19 due to more testing?
No. This is like saying if I am pregnant and don’t take a pregnancy test, I won’t be pregnant. The pregnancy exists whether I pee on that stick or not.
It is true that if we test more we KNOW about more cases, but those cases are out there whether we test for them or not. If we don’t test, those cases still exist but we would not know who was sick or where the disease was spreading. We can tell if case numbers are due to testing by comparing increases in testing with increases in cases and by looking at the test positivity rate.
- Scientists can predict how many more cases we should see with increased testing and then compare actual cases to these predicted cases to see if the increase in case numbers is due only to more testing. Currently, in many areas of the US, we are seeing the increase in cases is more than would be expected from an increase in testing alone.(This NY Times article shows graphs for each state).
- The “test positivity rate” is the percent of tests performed that are positive. If the number of actual cases is staying the same or decreasing, but more tests are done, the test positivity rate would fall. If the test positivity rate goes up, it means the rise in cases is outpacing increases in testing. Current recommendations are a test positivity rate of below 5% over the past 14 days before reopening should be considered. Countries that have been successful at containing outbreaks have a test positivity rate below 2%. While test positivity rates in the US have been falling, many US states still have test positivity rates above the 5% target. (You can find charts here from Johns Hopkins Coronavirus Resource Center.)
Testing allows scientists to learn more about spread, transmission, disease course, testing accuracy, and everything else about this disease. Testing also lets us know which measures to slow transmission are working. Knowing more about the disease helps us take more effective and targeted measures to contain it.