The positive predictive value (PPV) is a parameter for evaluating medical testing procedures. It indicates how many people who have been given a certain diagnosis by a particular test procedure are actually ill. For the calculation of the PPV, according to Bayes, the specificity of the test and the prevalence of the disease are of particular importance.
The prevalence indicates the occurrence of cases of the disease in the population at a given moment. While the specificity can be determined very precisely, the prevalence is currently unknown. It can only be estimated roughly and varies considerably from region to region. Therefore, the PPV should be interpreted with great caution at the moment.
As an example, for the test we use, which has a specificity of 99.6 percent, an assumed prevalence of 5 percent results in a PPV of 93 percent. This means that 93 percent of positive results are actually positive, while 7 percent are false positive. Assuming a prevalence of 10 percent with the same specificity, this results in a PPV of 96 percent (see chart).