While the pandemic is real, the fear and some of the extreme responses may have been blown out of proportion. One of the conspiracy theories that is going around is that we’re having a “casedemic,” which stems from the lack of understanding of basic molecular biology and epidemiology.
While there is some truth to the casedemic arguments, the real situation is more nuanced than conspiracy theorists believe. This article will analyze both sides of the story and provide the most updated information about COVID-19 tests. As a molecular geneticist, I will help you understand the nuances of these tests and how they pertain to the current pandemic situation.
What Is a Casedemic?
In his video “Crucial Viewing,” dated August 12, 2020, Ivon Cummins suggested that COVID-19 cases rose due to increased testing rather than the actual number of cases. He reinforced his position by posting a graph showing how the number of cases went up but the number of deaths stayed very low.
The casedemic theory states that we don’t have a pandemic, but rather a fear-mongering scheme created by manufactured positive test results. People who believe in this theory also believe that the scheme is created to allow the governments to become more draconian and to vaccinate everyone.
Daily cases: UK
Daily deaths: UK
Daily cases: France
Daily deaths: Frances
If you crop this graph with a cutoff in August, as did Ivor Cummins in his Youtube video, you will see that the cases rose, but the death rates did not. Source: coronavirus.jhu.edu.
Interestingly, in December, as we write this article, the hospitalization and death rates due to COVID-19 are sharply increasing in the northern hemisphere. However, the percentage of deaths isn’t as high as early in the pandemic. Therefore, we’re having a true increase in the number of cases, and thus very sick people, after all.
Is a COVID-19 Casedemic Real?
For the sake of the argument, let’s assume there might be some truth to the following facts:
- The COVID-19 tests aren’t very accurate; more tests lead to more false positives.
- Any COVID-19–positive deaths from all causes would be considered COVID-19 deaths.
Now, let’s try to understand the facts behind each of these claims.
How Do COVID-19 Tests Work?
Currently, there are three main ways to test for COVID-19: You can detect the viral genetic material, its protein, or your immune response to the virus.
1) Detection of SARS-CoV2 Genetic Material (RNA)
The technology relies on pairs of oligos that recognize the unique genetic sequence in the viral genome. The RNA would only be copied into DNA and amplified if the starting sample contains those recognized sequences.
Subsequently, the DNA is amplified so that it can be visualized using DNA-specific fluorescent dyes. Each round of amplification doubles the amount of double-stranded DNA in the test tube.
In theory, if the sample contains the viral RNA, there will be a detectable amount of double-stranded DNA in the test tube. In practice, this test is so sensitive that it can pick up artifacts or contaminants with sufficiently similar genetic sequences, especially with higher rounds of amplification. However, the labs have ways to account for this issue (more on this later).
Real-Time Reverse Transcription Polymerase Chain Reaction (rRT-PCR)
RT-LAMP is an upgrade from RT-PCR. It uses an enzyme that doesn’t require changes in temperature (isothermal), eliminating the need for the thermal cycler machine. The amplification step takes only around 30 minutes and generates DNAs that form hairpin loops. Then, a specific dye that binds to this hairpin loop structure makes it possible to detect the presence of the hairpin DNAs in positive samples.
A new protocol for RT-LAMP bypasses the requirements for RNA isolation, making RT-LAMP-based tests more portable. This test has excellent specificity (99.5%) but slightly lower sensitivity (86%) when the RNA isolation is bypassed.1 This test is about as sensitive as the RT-PCR test at 30-cycle thresholds.
Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)–Based Methods
CRISPR-based detection methods rely on the same loop amplification at a single temperature like RT-LAMP. However, instead of using the dyes, a CRISPR enzyme (Cas12 and Cas13) selectively cuts the SARS-CoV2 DNA sequence that it recognizes. After the CRISPR process, the sample can then be placed on a lateral flow strip. Uncut DNA (negative results) will get caught at the control line. However, CRISPR-cut DNA will travel further up the strip and result in another line. The final result of this method looks very similar to a home pregnancy test.2,3
Image caption: SHERLOCK RT-LAMP + CRISPR-based detection method developed by the Zhang group at Broad Institute, MIT.3
2) Detection of a SARS-CoV2 Protein (Antigen)
Tests that detect SARS-CoV2 proteins, called antigen tests, detect one of the viral proteins with antibodies. The antibodies could be bound to a nitrocellulose membrane or in a solution. These tests are faster and less expensive. However, they’re less sensitive, with on average around 56.2% sensitivity.4
Image caption: How rapid antigen tests work
These tests only measure your antibody (a type of immunologic protein that can bind to the virus) response to SARs-CoV2. The problem is that antibodies may only arise about a week into the infection, and some people produce no antibodies.5 IgM indicates an active infection, whereas IgG indicates a past infection. Therefore, these antibody tests are not diagnostic, but they can tell if you’ve had the infection in the past. They may also be a useful epidemiologic tool to estimate the number of people who have already had the disease. The antibodies are associated with protection from infection and may last up to 6 months.6
How Accurate are These COVID-19 Tests?
As is the case with other medical diagnoses and tests, the diagnosis process should combine other information with the test results. These include symptoms, presentation, history, and ruling out other similar diseases. Very rarely is a medical test 100% accurate, so the doctor is responsible for weighing all pieces of information to diagnose the disease. Your doctor may diagnose you with probable or suspected COVID-19 if you have shortness of breath and loss of smell even though the COVID-19 test came back negative, for example. Sensibly, in the U.S., all COVID-19 tests require prescriptions from healthcare professionals.
In short, COVID-19 tests are not 100% accurate, especially when conditions are rarely perfect and the test cannot be repeated. COVID-19 patients start shedding virus particles a few days before symptoms onset, and the most virus about 7–10 days after symptoms onset. Missing this window can result in false negatives. Ideally, a patient who tests negative should be retested, but this doesn’t happen often. As a result, these tests can miss up to 40% of true cases.7 Therefore, inconclusive results or a history of exposure alone typically warrants a self-isolation period.
Image caption: Timeline for viral load, symptoms, and positivity on rRT-PCR and serological (antibody tests).8
However, the medical staff who order and perform these tests are accounting for these as they would routinely do.
Here are some common terms used to assess the accuracy of tests:
- Sensitivity refers to how often the test turns positive results if they have the disease. It may also be displayed as “Limit of Detection” or the lowest concentration of the analyte (RNA in this case) that the test can detect.
- Specificity is how often the test produce negative results for if they don’t have the disease, or if they have a different viral infection
- Positive predictive value (PPV) is the probability that someone with a positive test will truly have the disease.
- Negative predictive value (NPV) is the probability that someone with a negative test will truly not have the disease
PPV and NPV are different in symptomatic and asymptomatic patients and also depend on the population prevalence. The PPV decreases in asymptomatic people, or when there is no outbreak or low prevalence in the population. Conversely, NPV increases in these situations. Therefore, if you see PPV dropping in some populations as the outbreaks resolve, it’s because that’s how PPV works not because the labs are incompetent.
Here is what we currently know about the PPV, NPV, sensitivity, and specificity of each type of test.
Test type | Developed by | Sample type | PPV (95% confidence interval) | NPV | Sensitivity | Specificity | Reference |
---|---|---|---|---|---|---|---|
rRT-PCR |
CDC |
NP, nose, or mouth swap |
100% (77.2–100) |
100% (96.4–100) |
|
|
9 |
SHERLOCK (CRISPR) |
Feng Zhang, Broad institute |
|
98.4 |
93.4 |
|
|
3 |
RT-LAMP |
Meta/systematic reviews |
Purified clinical samples |
|
|
94% (90–96) |
99% |
10 |
RT-LAMP |
Meta/systematic reviews |
Unpurified samples |
|
|
78% (65–87) |
99% |
10 |
Antigen tests |
Meta/systematic reviews |
|
|
|
56.2 %(29.5–79.8) |
99.5% (98.1–99.9) |
4 |
In order for tests to receive emergency use authorization, they have to display very high specificity. To do so, the test developers need to rigorously validate the tests with blank, human, and other very similar viruses. Here’s an example of such validation results with the CDC rRT-PCR tests.
Source: https://www.fda.gov/media/134922/download
The organizations that develop these tests rigorously validate them. There are many ways to minimize false positives so that the tests are highly unlikely to give false results. These include the following ways:
- Designing the primers that target the amplification region only specific to the viral genome
- Computationally checking the primer sequences against all similar viral and human genomes using a widely available software called BLAST (Basic Local Alignment Search Tool)
- Testing different annealing temperatures (the temperature that the oligos bind to the viral genetic material) to maximize sensitivity, while minimizing artifacts.
- Testing multiple negative controls or samples that are not supposed to turn up positive results, such as blank samples, RNA from other viruses, and swaps from healthy humans
- Testing multiple positive controls or samples that are supposed to turn up positive results, including the virus itself and COVID-19 patient samples
- Testing multiple viral genes using multiple pairs of oligos in a single sample
- Including a pair of oligos that target a human RNA, such as the RNase P gene. This is to confirm that the RNA in the samples are not so degraded that they cannot be tested, and also that the enzymes and other components in the test are working correctly.
These practices are documented with the results published in the CDC rRT-PCR panel catalog.9 As a molecular biologist who is familiar with the technology, I am satisfied with the extent that they validated the test.
Also, each lab has to validate their conditions, machines, and reagents to determine their own specificity, sensitivity, PPV, and NPV. During this validation process, they identify the Ct value that identifies the lowest amount of viral RNA the test can identify. Based on the CDC’s own validation, rRT-PCR can reliably pick up about 3.16 viral RNA copies per microlitre at around 32.5–35.8 cycles.10 Any less viral RNA and it starts to come back negative at Ct of 40 cycles. Therefore, the determination of Ct for each test is not arbitrary.
#4–#7 are always tested in parallel when a medical lab test suspected COVID-19 samples. If the positive controls (#5 and #7) works while the negative controls (#6) didn’t produce positive results, then the positive results in the same batch are credible. However, if the positive controls didn’t come up positive, the negative results will not be believable and the lab will need to repeat the test.
What is a COVID-19 Case?
There are three different categories of COVID-19 cases — probable, suspected, and confirmed.11 A small percentage of confirmed cases via nucleic acid testing may be asymptomatic. The reported case counts include only those confirmed cases. However, we know that there are many other people with the virus who were never tested.
You can “have” COVID-19 and belong to any of these categories, depending on whether you have had a nucleic acid test (RT-PCR) or RT-LAMP.
Is Doing rRT-PCR Over 40 Cycles Turning Up False Positives?
Forty cycles is a lot of cycles to run a PCR. However, the positive results are usually called at lower cycle numbers than 40. According to the report by Dr. Marie Landry MD, Director of Clinical Virology Laboratory of Yale-New Haven Hospital, running the tests to 40 cycles is necessary to turn up positive for some hospitalized patients. Those with pneumonia may have the little virus in their upper airway, so sputum or bronchoalveolar washing is more likely to test positive. Whereas, recovered patients can shed non-infectious viral RNAs and test positive months after recovery, such that isolating based on these test results is no longer recommended.12 Again, this emphasizes the need to combine test results with symptoms and presentation.13 Obviously, the high-Ct positive test in a hospitalized patient will be interpreted much differently than an asymptomatic one.
For a test to be called positive, the positive and negative controls still have to show up correctly, even at high Ct. If the negative control becomes positive, the test results will be considered inconclusive.
At Yale-New Haven Hospital, only about 14% of tests were positive at Ct at or over 30. Also, MIT’s COVIDPass program ran over 156,000 tests, and under 90 of them were positive.14 As of Aug 15, Public Health Ontario tested 850,000 samples with ~17,500 positives. Among these, 30 of them were false positives.
The NY Times article that raised this issue didn’t disclose the data or source of data where it claimed the majority of cases in Massachusetts, New York, and Nevada were based on Ct over 37 for us to verify.
An analysis by Cohen and Kessler showed that the rRT-PCR false-positive rates of other viruses are between 0.6–16.7%. Based on their combined analysis of studies up to April 2020, the average false positive rate is around 3% for SARS-CoV2 RT-PCR. Therefore, it’s inconceivable that up to 90% of COVID-19 positive cases are false positives.15
Why Isn’t the Death Rate as High Now as it Was Early in the Pandemic?
Early in the pandemic, we had major shortages of tests because we were short on chemicals and laboratory supplies. Also, the CDC had to discard the first tests they developed due to excessive false-positive rates from cross-contamination. Therefore, only a fraction of true cases were tested. By September 30, one modeling study estimated that the true number of cases was between 6–24 times the number of reported cases.16 Another study estimated this range to be 3–20 times.17
On top of that, there are reporting delays and errors at all levels. In the U.S., there is no central medical system, which makes it challenging to systematically collect data that will be useful for research. Therefore, the reported number of cases and death rates are inaccurate. Rather, policy decisions need to take into account many factors including the number of cases, percent positive tests, deaths, hospitalization rates, and individual patient factors that determine risks such as age and health status.
By the summer, we were testing milder cases and testing them earlier as testing resources caught up. Also, the younger and healthier populations were now the ones getting infected, rather than the older and more vulnerable populations. Most importantly, we’d learned a lot more about better treatments for hospitalized cases. All of these factors significantly cut down the death rates. The summer and increased blood vitamin D levels may have also reduced the mortality and severity of the disease.18
Based on this information, the casedemic arguments are not valid because:
- COVID-19 tests have been rigorously validated, and the false-positive rates are at most 16.7%
- We are experiencing a spike in deaths, although the death rates have gone substantially down due to improved treatments.
- In many jurisdictions, policy decisions are based on the right information, including hospitalization rates.
References
1. Dao Thi, Viet Loan, Konrad Herbst, Kathleen Boerner, Matthias Meurer, Lukas Pm Kremer, Daniel Kirrmaier, Andrew Freistaedter, et al. “A Colorimetric RT-LAMP Assay and LAMP-Sequencing for Detecting SARS-CoV-2 RNA in Clinical Samples.” Science Translational Medicine 12, no. 556 (2020). https://doi.org/10.1126/scitranslmed.abc7075.
2. Broughton, James P., Xianding Deng, Guixia Yu, Clare L. Fasching, Venice Servellita, Jasmeet Singh, Xin Miao, et al. “CRISPR-Cas12-Based Detection of SARS-CoV-2.” Nature Biotechnology 38, no. 7 (2020): 870–74. https://doi.org/10.1038/s41587-020-0513-4.
3. Joung, Julia, Alim Ladha, Makoto Saito, Nam-Gyun Kim, Ann E. Woolley, Michael Segel, Robert P. J. Barretto, et al. “Detection of SARS-CoV-2 with SHERLOCK One-Pot Testing.” The New England Journal of Medicine 383, no. 15 (2020): 1492–94. https://doi.org/10.1056/NEJMc2026172.
4. Dinnes, Jacqueline, Jonathan J. Deeks, Ada Adriano, Sarah Berhane, Clare Davenport, Sabine Dittrich, Devy Emperador, et al. “Rapid, Point‐of‐Care Antigen and Molecular‐Based Tests for Diagnosis of SARS‐CoV‐2 Infection.” Cochrane Database of Systematic Reviews, no. 8 (2020). https://doi.org/10.1002/14651858.CD013705.
5. Li, Kening, Bin Huang, Min Wu, Aifang Zhong, Lu Li, Yun Cai, Zhihua Wang, et al. “Dynamic Changes in Anti-SARS-CoV-2 Antibodies during SARS-CoV-2 Infection and Recovery from COVID-19.” Nature Communications 11, no. 1 (2020): 6044. https://doi.org/10.1038/s41467-020-19943-y.
6. Lumley, Sheila F., Denise O’Donnell, Nicole E. Stoesser, Philippa C. Matthews, Alison Howarth, Stephanie B. Hatch, Brian D. Marsden, et al. “Antibodies to SARS-CoV-2 Are Associated with Protection against Reinfection.” medRxiv, (2020), 2020.11.18.20234369. https://doi.org/10.1101/2020.11.18.20234369.
7. Woloshin, Steven, Neeraj Patel, and Aaron S. Kesselheim. “False Negative Tests for SARS-CoV-2 Infection — Challenges and Implications.” The New England Journal of Medicine 383, no. 6 (2020): e38. https://doi.org/10.1056/NEJMp2015897.
8. La Marca, Antonio, Martina Capuzzo, Tiziana Paglia, Laura Roli, Tommaso Trenti, and Scott M. Nelson. “Testing for SARS-CoV-2 (COVID-19): A Systematic Review and Clinical Guide to Molecular and Serological In-Vitro Diagnostic Assays.” Reproductive Biomedicine Online 41, no. 3 (2020): 483–99. https://doi.org/10.1016/j.rbmo.2020.06.001.
9. “CDC 2019-nCoV Real-Time RT-PCR Diagnostic Panel — Instructions for Use.” Centers for Disease Control and Prevention, December 1, 2020. https://www.fda.gov/media/134922/download.
10. Subsoontorn, Pakpoom, Manupat Lohitnavy, and Chuenjid Kongkaew. “The Diagnostic Accuracy of Isothermal Nucleic Acid Point-of-Care Tests for Human Coronaviruses: A Systematic Review and Meta-Analysis.” Scientific Reports 10, no. 1 (2020): 22349. https://doi.org/10.1038/s41598-020-79237-7.
11. “WHO COVID-19 Case Definition.” Accessed January 15, 2021. https://www.who.int/publications/i/item/WHO-2019-nCoV-Surveillance_Case_Definition-2020.2.
12. KDCA. “Findings from Investigation and Analysis of Re-Positive Cases.” May 19, 2020. https://www.cdc.go.kr/board/board.es?mid=a30402000000&bid=0030&act=view&list_no=367267&nPage=24.
13. Landry, Marie L. “Your Coronavirus Test Is Positive. Maybe It Shouldn’t Be.” New York Times, August 29, 2020. https://medicine.yale.edu/labmed/sections/virology/COVID-19%20Ct%20values_YNHH%20Aug.%202020%20_395430_36854_v1.pdf.
14. “Was My PCR Test Result a False Positive?” Accessed December 30, 2020. https://medical.mit.edu/covid-19-updates/2020/11/pcr-test-result.
15. Cohen, Andrew N., and Bruce Kessel. “False Positives in Reverse Transcription PCR Testing for SARS-CoV-2.” bioRxiv. medRxiv, May 1, 2020. https://doi.org/10.1101/2020.04.26.20080911
16. Reese, Heather, A. Danielle Iuliano, Neha N. Patel, Shikha Garg, Lindsay Kim, Benjamin J. Silk, Aron J. Hall, Alicia Fry, and Carrie Reed. “Estimated Incidence of COVID-19 Illness and Hospitalization — United States, February–September, 2020.” Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of America, November 25, 2020. https://doi.org/10.1093/cid/ciaa1780.
17. Wu, Sean L., Andrew N. Mertens, Yoshika S. Crider, Anna Nguyen, Nolan N. Pokpongkiat, Stephanie Djajadi, Anmol Seth, et al. “Substantial Underestimation of SARS-CoV-2 Infection in the United States.” Nature Communications 11, no. 1 (2020): 4507. https://doi.org/10.1038/s41467-020-18272-4.
18. Thompson, Derek. “COVID-19 Cases Are Rising, So Why Are Deaths Flatlining?” The Atlantic, July 9, 2020. https://www.theatlantic.com/ideas/archive/2020/07/why-covid-death-rate-down/613945/.
4 Comments
Hi Nattha, very informative and well written.
Good Article. There r so many articles by renowned professionals. Unfortunately any or most articles that challenge the status quo are censored. You must know that. Many medical people are speaking put and are loosing their jobs for talking. That is a red flag, big red flag. I would ask why we are all supposed to take a vaccine especially all these fast tracked vaccines? Why r we starting to force people to comply. If you don’t follow the new rules you will be restricted in your movement. Debate needs to happen in the open not censored. If you want to be part of the experiment thats great. People need choices with no repercussions to their lives. I get tested 3 times a week. There is no way that these tests are taken correctly from the WHO recommendations. If they r show me, id like to see. When you shut down the world and force bankruptcies, unnecessary suffering, for a virus that we aren’t even sure if it was from the wild or man made doesn’t kill most healthy people?? But all the people at the top are becoming wealthier and wealthier. The technocrats. What is really happening? I would like a little more clarification before i line up.
“Hi Tim, Thank you for your comment. The WHO’s notice is a reminder of what we previously knew the entire time. I included the information in the article, i.e.:
The more Ct –> the less viral load.
The positive predictive value goes down with known prevalence in the population. “
“Hi Dr. Wannissorn, thank you for the very informative article. Since there’s obviously an understandable delay between the time you wrote this and time of publishing, would you kindly comment on how this new information the WHO released just a few days ago relates to your findings? thanks! https://www.who.int/news/item/20-01-2021-who-information-notice-for-ivd-users-2020-05
“