1.1.2. Test sensitivity and specificity

Many case definitions are based on laboratory or clinical tests, but tests in themselves are prone to errors either by misidentifying disease positive cases, i.e. individuals that have the disease, incorrectly as negative cases, or disease-free, or disease negative cases as positive cases. The accuracy of a test is primarily given as sensitivity and specificity.


Sensitivity is the probability that a human or animal will have a positive test result if indeed the human or animal does have a disease. This is expressed as: P(T+|D+),  where P is the probability, T+ is a  positive test result and D+ is a disease being present.  In applied epidemiology, sensitivity is often expressed as a proportion, and thus expressed as equation 1.1.2.a.

 Equation 1.1.2.a


Similarly, specificity is the probability that a human/animal will have a negative test result if indeed it is disease free. This is expressed as: P (T-|D-), where P is the probability, T- is a negative test result and D- is the disease not being present. In applied epidemiology, specificity is often expressed as the proportion of non-diseased (healthy) animals that test negative, expressed as equation 1.1.2.b.

Equation 1.1.2.b Calculating confidence intervals for a proportion