Are throat swabs effective in diagnosing COVID-19?

Are throat swabs effective in diagnosing COVID-19?

Are throat swabs effective in diagnosing COVID-19?

In this article, we will analyze whether throat swabs are actually effective and reliable in diagnosing COVID-19.

As we have already discussed in-depth in other articles that it is necessary to verify whether a patient is infected with the coronavirus, a specific diagnostic test is performed on a biological sample. Usually, the sample consists of a small amount of mucus taken from the nose (nasal swab) or from the throat (throat swab). Whenever possible, sputum or bronchial wash samples can be taken from the deep airways, but they are much less common, more laborious to obtain, and must be performed with specific circumstances.

From the beginning of the epidemic to today, nasal and throat swabs have certainly been the most collected samples, for both screening and also diagnosing the presence of the virus; thanks to the accessibility of the upper airways and the ease of performing the sampling.

Do all mucus samples from infected patients come out positive?

In a situation where these diagnostic tests are so widely used to identify infected people within the population, it is important to consider their limitations. What certainty do we have in finding the virus in the nasal or throat swabbed mucus samples of infected patients? If the tests performed on these samples gave a high number of false negatives, as in negative results despite the fact that people are actually infected, this means that a large part of the infected population could have escaped the radars and silently contributed to the spread of the infection.

Clinical impressions

When we started on the first swabs, we anxiously waited for the results to understand if the patients we had contact with were positive or not.

We had great confidence that the swab could give us comprehensive information.

As time passed, however, more and more patients who had a negative swab were coming into the hospital with fever and/or breathing difficulties. After several days and a worsening of their clinical conditions, these patients then had positive swabs! Even after one, two, three or four negative swabs!

We therefore began to ask ourselves: what if swabs have a low probability of identifying the positive subjects? We have already underlined in our article on swab tests that if the result is positive, it is not mistaken (1). However, little is yet known about the actual ability of the test to detect those who have COVID-19.

 

How do we know if throat swabs are reliable to diagnose COVID-19?

Because of such results, we began to study the phenomenon. To understand if the test applied to the swabs is able to identify something, we must first establish a method to identify that “something” that must be identified.

In other words, when one wonders if a method is effective, it must be compared with another method. We incurred a problem though: is there another method to compare the swab test with? Unfortunately the answer is NO. Swab testing is currently the only useful method in identifying patients with the coronavirus.

From this fact, we can understand that the matter is rather intricate because there is no other method, or “gold standard” as we say in the medical field, against which we can compare the swab test to.

The only method available to confirm the effectiveness of the swab test is to verify its result on ALMOST CERTAINLY POSITIVE patients, because they are unequivocally ill with COVID-19. In fact, only patients with a COVID-19 CLINICAL DIAGNOSIS are considered a good comparison group to evaluate the swab test.

 

The study

A first partial answer comes from a pilot study conducted in China on 205 patients diagnosed with COVID-19. In the study, the COVID-19 diagnosis was performed through the presence of:

  • typical disease symptoms (fever, dry cough, and fatigue);
  • radiological tests (diagnosed pneumonia on chest x-ray or CT scan);
  • detection of the Coronavirus CoV2 in a biological sample.

Various biological samples were examined in these patients. Research was done on the coronavirus samples found in different bodily fluids or secretions to find out where it is most frequently found.

During the study, the doctors took six different types of biological samples from the patients in addition to those normally taken during a nasopharyngeal swab. This was taken at various times during the course of the disease.

In total, the 1070 samples included specimens of blood, feces, urine and also deep respiratory tract material (called “sputum”). For the more serious patients who underwent mechanical ventilation because they were no longer able to breathe on their own, samples of bronchial washes and biopsies (small samples) of respiratory tissues were also available.

The researchers performed the diagnostic test to detect the presence of SARS-CoV-2 in all the samples, trying to detect the genetic material of the virus. Ideally, most of the samples should have been positive; at least those taken from the respiratory tract where the virus normally lives.

 

THE RESULTS: The nose and throat samples are not as reliable as we thought

The results of the study (shown in the table) were quite unexpected.

tabella-efficacia-tamponi

 

The samples analyzed were specimens from:

  • Bronchial wash: 15 samples, 14 positive, Reliability (2): 93%
  • Bronchoscopy biopsy: 13 samples, 6 positive, Reliability (2): 46%
  • Sputum: 104 samples, 14 positive, Reliability (2): 72%
  • Nasal swab: 8 samples, 5 positive, Reliability (2): 63%
  • Throat swab: 398 samples, 126 positive, Reliability (2): 32%
  • Stool: 153 samples, 44 positive, Reliability (2): 29%
  • Blood: 307 samples, 3 positive, Reliability (2): 1%
  • Urine: 72 samples, 0 positive, Reliability (2): 0%

Data analysis

Only 32% of THROAT SWABS of those infected were positive, that leaves 2/3 of the cases incorrectly diagnosed. The confirmation of these data could have important consequences on the counting of real cases in Italy and in the rest of the world, as well as on the strategies of contagion containment.

The virus is present in a nonnegligible way in the individuals’ feces. This is proof for cases of oral-fecal transmission. We may also relate this to the presence of diarrhea as a warning symptom of COVID-19.

Things got even more complicated when students found out that not all biological samples from the same patient are positive. For example, there was a case where a sputum sample came out as positive though the faeces specimen and throat swab, from the same patient, were negative. On the other hand, the opposite also took place: positive results from the throat swab, but faeces and sputum samples tested negative.

 

Conclusion: Are throat swabs effective for diagnosing COVID-19?

Overall, what does this information tell us? It is shown that the diagnosis from nasal or throat samples alone may not be reliable because it is incorrect in 1/3 and 2/3 of cases respectively. To minimize the results of false positives as much as possible, i.e. results that indicate no infection in patients while actually infected, it would be useful to perform the test on samples of the lower respiratory tract, which seem to be more reliable. However, the method is certainly not applicable on a large scale. In addition, a good strategy could include testing multiple samples from the same patient to decrease the likelihood of error and therefore ensure the reliability of a negative result.

 

Med4Care Marco De Nardin

Dr. Marco De Nardin, M.D.

Notes

  • (1) the specificity is very high, at least according to the results we could find on the internet: https://www.med4.care/covid-19-diagnostic-testing/
  • (2) Reliability refers to Sensitivity. This term has been used to allow everyone to understand the “common” meaning of the term.

References:

  • Wang W, Xu Y, Gao R, et al (2020) Detection of SARS-CoV-2 in Different Types of Clinical Specimens. JAMA

https://doi.org/10.1001/jama.2020.3786

https://jamanetwork.com/journals/jama/fullarticle/2762997?resultClick=1

 

Leave a Reply