COVID TESTING AND THE UNDERSTANDING OF IMMUNITY AND THE RESULTS

Let us start with the basics.

Sensitivity
This is the ability of the test to detect the immune protection.  If 100 patients have immunity and the test is positive in 90 of them, then it has a 90% sensitivity.  A professor at Imperial College has confirmed that the best results he has seen so far have about 90% sensitivity.

Specificity
This is the ability of the test to rule out immune protection.  If 100 patients do not have immunity, then the test is negative in 90.  It therefore has a 90% specificity.

Both sensitivity and specificity are important for a test.  For example, the presence of a heartbeat has 100% sensitivity for COVID protection, but it does not have any value as it is not specific.

However, we do not really care so much about these.  We want to know what to do if we get a positive or negative test result for our patients.

Positive Predicted Value – PPV
This is the probability that if a patient’s test says they have COVID immunity, that they really do. 

Negative Predicted Value – NPV
Similarly, this is the probability that if a patient’s test says they do not have COVID immunity then they really do not. 

So what does this mean?
The problem when determining if a test is beneficial is that the PPV and the NPV depend on the underlying prevalence within the population, as well as the sensitivity and specificity.

covid-19-testing

For example, if 5% of a population is truly immune and we use a test with 90% sensitivity, its positive predicted value is 32% and this will mean that 68% of the population will have an erroneous result.   That’s because in 10% of cases of the 95% of the population who do not have immunity the test suggests they have – which is a large number compared to the 90% of cases of the 5% of the population who actually do have immunity.

However, if 20% of the population is truly immune then that same assay positive predicted value will be 69%, meaning that 31% will have a false positive. 

What this shows is that the usefulness of a test depends in part on how many people in a population group do actually have what the test is looking for.  It also shows the importance of carefully selecting who has a test to make sure it is best implemented.

This one of the reasons why The London General Practice only provides testing through a Doctor led consultation. To find out more visit www.thelondongeneralpractice.com/covid-19-testing