(Spanish language version: https://theconversation.com/pruebas-covid-de-pcr-o-antigenos-conoce-cuales-son-las-diferencias-)
If you want to learn more, please visit our website Accu-Tell.
At this point in the pandemic, you or someone you know has probably received at least one COVID-19 test. But do you know which kind of test you got and the strengths and weaknesses of these different tests?
I&#;m a molecular biologist, and since April I&#;ve been part of a team working on a National Institutes of Health-funded program called RADx that is helping innovators develop rapid tests to detect when a person is infected with SARS-CoV-2, the virus that causes COVID-19.
Two major types of tests are used to diagnose infection with SARS-CoV-2: molecular tests &#; better known as PCR tests &#; and antigen tests. Each detects a different part of the virus, and how it works influences the test&#;s speed and relative accuracy. So what are the differences between these types of tests?
The first step for either kind of test is to get a sample from the patient. This can be a nasal swab or a bit of saliva.
For PCR tests, the next step is amplification of genetic material so that even a small amount of coronavirus genes in the patient&#;s sample can be detected. This is done using a technique called a polymerase chain reaction. A health care worker takes the sample and treats it with an enzyme that converts RNA into double-stranded DNA. Then, the DNA is mixed with a solution containing an enzyme called a polymerase and heated, causing the DNA to separate into two single-stranded DNA pieces. The temperature is lowered, and polymerase, with the help of a small piece of guide DNA called a primer, binds to the single-stranded DNA and copies it. The primers ensure that only coronavirus DNA is amplified. You&#;ve now created two copies of coronavirus DNA from the original one piece of RNA.
Laboratory machines repeat these heating and cooling cycles 30 to 40 times, doubling the DNA until there are a billion copies of the original piece. The amplified sequence contains fluorescent dye that is read by a machine.
The amplifying property of PCR allows the test to successfully detect even the smallest amount of coronavirus genetic material in a sample. This makes it a highly sensitive and accurate test. With accuracy that approaches 100%, it is the gold standard for diagnosing SARS&#;CoV&#;2.
However, PCR tests have some weaknesses too. They require a skilled laboratory technician and special equipment to run them, and the amplification process can take an hour or more from start to finish. Usually only large, centralized testing facilities &#; like hospital labs &#; can conduct many PCR tests at a time. Between sample collection, transportation, amplification, detection and reporting, it can take from 12 hours to five days for a person to get results back. And finally, they aren&#;t cheap at $100 or more per test.
Rapid, accurate tests are essential to contain a highly contagious virus like SARS-CoV-2. PCR tests are accurate but can take a long time to produce results. Antigen tests, the other major type of coronavirus test, while much faster, are less accurate.
Antigens are substances that cause the body to produce an immune response &#; they trigger the generation of antibodies. These tests use lab-made antibodies to search for antigens from the SARS-CoV-2 virus.
To run an antigen test, you first treat a sample with a liquid containing salt and soap that breaks apart cells and other particles. Then you apply this liquid to a test strip that has antibodies specific to SARS-CoV-2 painted on it in a thin line.
Just like antibodies in your body, the ones on the test strip will bind to any antigen in the sample. If the antibodies bind to coronavirus antigens, a colored line appears on the test strip indicating the presence of SARS-CoV-2.
Antigen tests have a number of strengths. First, they are so easy to use that people with no special training can perform them and interpret the results &#; even at home. They also produce results quickly, typically in less than 15 minutes. Another benefit is that these tests can be relatively inexpensive at around $10-$15 per test.
Antigen tests do have some drawbacks. Depending on the situation, they can be less accurate than PCR tests. When a person is symptomatic or has a lot of virus in their system, antigen tests are very accurate. However, unlike molecular PCR tests, antigen tests don&#;t amplify the thing they are looking for. This means there needs to be enough viral antigen in the sample for the antibodies on the test strip to generate a signal. When a person is in the early stages of infection, not a lot of virus is in the nose and throat, from which the samples are taken. So, antigen tests can miss early cases of COVID-19. It&#;s also during this stage that a person has no symptoms, so they are more likely to be unaware they&#;re infected.
A few antigen tests are already available over the counter, and on Oct. 4, , the Food and Drug Administration granted emergency use authorization to another at-home antigen test. The U.S. government is also pushing to make these tests more available to the public.
Want more information on covid rapid blood test? Feel free to contact us.
At RADx, the project I am a part of, we are currently conducting clinical studies to get a better understanding of how antigen tests perform at various stages of infection. The more data scientists have on how accuracy changes over time, the more effectively these tests can be used.
Understanding the strengths and limitations of both PCR and antigen tests, and when to use them, can help to bring the COVID-19 pandemic under control. So the next time you get a COVID-19 test, choose the one that is right for you.
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Figure 1 illustrates the major steps performed in the data preprocessing, starting from a table containing 727,483 N-gene Ct-value records. For each person, we defined an infection event if there was either no previous infection or at least 90 days had passed from the previous infection (i.e., from the first positive PCR test in that interval)15. In the second step, only Ct records related to the patient&#;s first positive PCR test of the first infection event during the study period were included. Filtering out extreme Ct-values (rounded values above 35), the resulting data table contained 685,250 Ct-value records, each corresponding to a unique individual. Out of 676,552 persons with a confirmed infection case during the study period whose N-gene rounded Ct-value was 35 or below, 644,845 (~&#;95.3%) were tested in the five largest labs. These records also included 41,065 individuals who performed an additional AG test within 24 h after the PCR test, and before the PCR test result was obtained.
Figure 1Flowchart of individuals in the study, depicting the process of data filtering.
Full size image
Table 1 presents the distribution of different covariates in the overall population (Box B, Fig. 1) and compares it to the study subpopulation (Box D, Fig. 1). Most of the tests (more than 85%) were done during January and February which represents the peak period of the Omicron surge in Israel10. About 40% of the study subpopulation is of the age group 16&#;39 while elderly people aged 60 or more account for less than 6%. The age distribution differs from that of the overall population where individuals aged 16&#;39 comprise less than 35% and elderly individuals account for 15% of the PCR samples. Interestingly, there are more women than men in both the overall and the study population (60% vs 40%). There are some differences in the distributions of vaccination status and lab between the overall population and the study population. The rightmost column of Table 1 presents the percentages of positive AG tests in the different groups, which vary from 58 to 79%, in all groups except tests performed during March and April showing a larger probability (85&#;87.5%) of a positive AG test. There is quite a clear association between a positive AG result and time, where a larger proportion of positive results are obtained in later months. Other interesting findings are a smaller proportion of positive AG tests in children 0&#;2 and 3&#;4 compared to adults, a larger positive proportion in men compared to women, and differences among the labs.
Table 1 The number of PCR samples in the data with and without an additional AG test, and the AG detection rates within each category stratified by different variables.Full size table
Figure S1 in the supplementary material compares the Ct distribution of those who underwent both PCR and AG tests to the entire study population which underwent PCR tests, regardless of whether or not an AG test was performed within 24 h. The Ct values in the general population tend to be somewhat lower (mean 26.25 compared to 26.09), but the difference between the distributions is quite small with the mean difference far below one Ct cycle.
Figure 2 illustrates the relation between Ct-level and AG detection probability. The vertical bars represent 95% confidence intervals around the proportions in each Ct category (integer values). The dashed line represents an estimated univariate logistic model over the Ct-values. As can be seen from the graph, the empirical AG detection probability closely approximates a logistic curve. This graph asserts that a decreased Ct level (i.e., increased viral load) renders AG detection more probable in a nonlinear fashion. Note that the detection probability for Ct-values less than 23 is very high. Figure S2 repeats the analysis in Fig. 2, stratified by the different covariates, while Fig. S3 presents stratification by lab. It shows a new, more detailed look at the results provided in the rightmost column of Table 1. The differences in AG detection probabilities between months, sex, and age groups are clearly observed. One can use these plots to evaluate detection rates in worst/best case scenarios. For instance, in cases where samples were taken in January vs March (see Fig. S2 for differences in detection rates).
Figure 2The AG detection probability vs Ct-levels for gene N, measured by 5 labs between January and April . Vertical lines are proportion and 95% confidence interval calculated separately for each Ct integer value. The dashed line is a univariate logistic regression of Ct value on a positive AG test. The gray lines are the estimated multivariate logistic curves for each individual in the study population. The data that was used to create this graph is described in Box C of Fig. 1 (Box D for the gray curves).
Full size image
Table 2 reports the results of a multivariate logistic regression on a positive AG test, with Ct as a continuous exposure, adjusting for all covariates as mentioned in the Methods section. The Ct coefficient is &#; 0.27, translating to an odds ratio estimate of about 0.76. This is quite a large effect for a continuous variable ranging from 15 to 35. There is a significant effect for children aged 0&#;2 and 3&#;4, showing smaller probabilities to be detected by AG (OR&#;=&#;0.63 and 0.70, respectively). Males have a significantly larger probability of being confirmed by an AG test, with an odds ratio of 1.32. In addition, there are clear differences between the labs, possibly due to differences in assays, or different methodological standards. Interestingly, vaccination seems to be slightly positively correlated with the probability of AG detection. Finally, the probability of detection increases almost monotonically with time (i.e., calendar date), even after correcting for the Ct level. Mass AG testing in Israel started in January; at that time many teams were less experienced. Thus, accumulated experience in swab sample collection and rapid tests conduction, and possibly an improved sensitivity of AG kits, might all explain the increase in AG detection over time. The shaded area in Fig. 2 displays the variability in AG detection probabilities among individuals. It was obtained by estimating and plotting the logistic curves of all individuals in the study group, using the multivariate logistic regression results.
Table 2 Multivariate logistic regression results.Full size table
Finally, we used our model to estimate the expected number of undetected individuals in the hypothetical case that only AG tests were performed. This was done by applying the results of the multivariate model to the overall population who have full covariate data (Fig. 1, Box B). Using only AG tests would result in 70% detection rates. In other words, out of the 639,423 infections that were confirmed by PCR, only 450,438 positive tests would be obtained if the PCR test were replaced by AG tests, thus missing about 188,985 individuals in this period. A similar analysis predicts an increase from only 62 to 87% detection rate on days 0&#;13 (January) and days 98&#;111 (April) from study follow up, respectively.
If you want to learn more, please visit our website serum d dimer.