Specificity

Specificity (Sp) tells us the likelihood of a false positive test. A high specificity means that a positive test can rule in the disease. It does this by telling us the percentage of disease free patients who are correctly diagnosed as being disease free. A test that always yields a negative result when used in healthy patients is 100% specific - there are no false positives. When a person tests negative and does not have the disease, the test result is known as True Negative (TN). Using a test with poor specificity will result in a patient appearing to have a disease, when he is, in fact, healthy. This is known as False Positive (FP). A test with high specificity will have a low false positive rate. Highly specific laboratory tests are used to confirm the existence of a disease. Specificity equals the true negative rate and is independent of the disease rate in a particular community. We have a test we want to evaluate so we select 200 people, 100 have the disease and 100 do not. So, let’s see the test results of those who have the disease.

Another way of remembering this concept is to remember the acronym “SpPIn” which means, a test with high Specificity and a Positive result rules IN the diagnosis (Sackett et. al). Thus, a positive test result (P) in a specific test rules in the diagnosis. Specificity is calculated by taking the true negative rate and dividing it by the sum of false positive and true negative rate [TN/(FP+TN)].

In both diagnostic and screening situations, a test should be accurate enough that it does not produce large numbers of false-positive (poor specificity) or false negative (poor sensitivity) results. The concepts of sensitivity and specificity help to describe the validity and accuracy of a test. The higher the sensitivity and specificity is, the more useful the test. Using a test that is 100% sensitive, the test result will never be negative if a patient really has the condition (there are no false negatives). Using a test that is 100% specific, a positive test is always accurate (no false positives).

Quiz