I'm comparing a bunch of SARS-CoV2 rapid antigen tests:


Columns 4 and 6 list the values for sensitivity and limit of detection (LOD). How come that a test with a several times lower limit of detection can have a worse sensitivity?

As an example, consider tests #2 and #4:

#2 - Sensitivity: 97.7% - Limit of detection: 2.0*10^2.4 = 502 TCID50/ml

#4 - Sensitivity: 91.4% - Limit of detection: 2.5*10^1.8 = 158 TCID50/ml

Shouldn't these values correlate, i.e. the more sensitive a test is the smaller the amount it is still able to detect?

  • 1
    $\begingroup$ Detection limits depend on both sensitivity and the relevant noise levels. A short list of detection limit publications is here. $\endgroup$
    – Ed V
    Jan 8, 2021 at 19:54
  • 1
    $\begingroup$ You have to ask the authors how did they determine the limit of detection. In the German Standard DIN 32645, Nachweisgrenze has a specific protocol. $\endgroup$
    – AChem
    Jan 8, 2021 at 22:17

1 Answer 1


In this situation, "sensitivity" refers to how often the test is able to correctly detect the presence of analyte. A test with high sensitivity means that there are few false negatives.

Limit of detection refers to how much of the analyte must be present before it is considered positive.

Imagine two different tests: test A uses an antibody with very high affinity for a specific spike protein sequence. Because of the high affinity, it can detect the spike protein even when there's not much around (good limit of detection). However, the antibody is less robust against other sequences, so it sometimes reports a false negative in the presence of spike protein variants (poor sensitivity).

Conversely, test B has an antibody with medium affinity for several different spike protein variants. It relies on high levels of the protein being present, but it works with many different variants and rarely gives false negatives. This test would have a high (aka poor) limit of detection and high (aka good) sensitivity.

  • $\begingroup$ The usual definition of sensitivity in chemical analysis is that it is the slope of the response function or calibration curve, as relevant. It is also as you have given it, especially for ROC methodology. And, unfortunately, it is sometimes mistakenly used as detection “power”, whatever that is supposed to mean. Of course, the standards folks also weigh in. ;-) $\endgroup$
    – Ed V
    Mar 13, 2021 at 16:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.