print   Print full article
You have a sore throat, you might have COVID-19, so you have a test, and when the results come back you will know if you have it -right?  Well it is not quite that simple and doctors have to interpret test results in the light of information about the test and the patient. 

There is no such thing as a perfect test- one which correctly answers the question being asked 100% of the time.  Many factors come into play, not all of them are obvious.  The analytical accuracy and precision of pathology tests are typically very good.  If you do the same test multiple times on the same sample you will generally get the same answer. So why isn’t their diagnostic accuracy the same?

Let’s take the test being used to diagnose COVID-19 in Australia at the moment.  It looks for the virus RNA in a sample taken from the nose or the back of the throat.    The coronavirus is made up of a genetic material called RNA. The test has to have a certain minimum amount of RNA present to record a positive result.  You could tweak the test to find lower levels of virus RNA but in doing so you will increase the likelihood of the test giving a positive result even if there was no RNA in the sample.
 
False negatives and false positives
  • If the test gives a negative result in a person who is actually infected that is called a false negative
  • A person who does not have the infection but whose test gives a positive result is a false positive

We have seen how the number of false positives could increase by trying to push the sensitivity of the test to detect very low viral loads.  Reducing the chance of false positives by reducing the sensitivity will also (very slightly) increase the chance of a false negative.
 
Sensitivity and specificity
The balance of sensitivity (the proportion of patients who have the disease that return a positive test result) and specificity (the proportion of patients without the disease who return a negative test result) is fixed once the test method is set.  But how we interpret a negative or positive result must be made in the light of how likely the patient was to have the virus before we knew the result.

In the general population in Australia or New Zealand you are very unlikely to be infected, so a negative test result is very likely the correct result and you should be reassured you don’t have COVID-19.  If a person returns from New York by plane and two days later develops symptoms consistent with COVID-19, they are much more likely to be infected. A positive result will make it almost certain they are infected but a negative result does not rule it out, it could be a false negative, and isolation with repeat testing on another sample should be considered.
 
When a false negative can occur
So why would the test give a false negative result?  Again, there are many factors in play. 

Getting a good sample
The most important in this case is the sample swab.  Swabs that do not reach a part of the nose or throat where the virus is present will not come back positive.  Repeat testing should always be done on a fresh sample.

Timing
Timing is also important. If the sample was taken too soon after the person was exposed the viral load may not be high enough in the throat to be detected.  On the other hand, doing an RNA test when the patient has recovered will give a negative result and therefore is not useful to know if someone had the virus in the past.

Antibody (serology) tests for COVID-19 look for the antibodies produced by the body in response to the virus. Because this process takes a few days, they are very likely to give a false negative result in the first few days of infection.  The body takes time to produce the antibodies against the virus but they linger for some time – often weeks or months - and so antibody tests are useful to detect past infection and potentially immunity.  But the body also produces antibodies against other similar coronaviruses like the common cold and these may be detected by some of these tests giving false positives. 

Bottom line:  there is no such thing as a perfect diagnostic test, results must always be interpreted with caution and that is best done by a health professional.
 
 
Internal links:
How reliable is pathology testing? 
 

Last Review Date: July 17, 2020