The optimality of a given diagnostic procedure is characterised by two parameters, sensitivityand specificity. Sensitivity is the probability that the test will test positive in a person with the disease, while specificity is the reliability with which the test identifies those who do not have the abnormality/elevation of the parameter.

Accuracy and precision are also essential definitions. The former refers to the closeness of the result of a measurement to the true value of the observed phenomenon, while precision is the standard deviation of the mean, indicating repeatability and reproducibility. Precision is an important parameter for all tests, while repeatability is an important property for instruments intended for repeated use.

While sensitivity and specificity provide information about the reliability of a test, positive and negative predictive value also take into account the individual’s perspective when evaluating a test. The positive predictive value (PPV) is the probability that a positive test result gives the probability that the individual is ill (the reliability of a positive test result). The negative predictive value (NPV) is the probability that a negative result means that the individual is not ill (the reliability of a negative test result).

The negative and positive predictive value can be generalised by knowing the prevalence of a given disease/condition, so even a test with good sensitivity and specificity has a low PPI if the prevalence of the disease is low in the overall population. Using the example of an HIV test with 99.9% specificity: if the prevalence of the disease in a given population is 10%, the test has a positive predictive value of 99%, but at a prevalence of 1% the positive test result is only 91%, while at a prevalence of 0.1% the test has a positive predictive value of 50%.