Diagnostic validity as a theoretical concept and as a measurable quantity.

The analytical result of a laboratory examination is a scientific fact and has no medical meaning as such. It must be interpreted to become a medical finding. To explain the very complex cognitive procedure of the interpretation a three-level model is used. In an environment of cost containment in health care systems the quality of medical laboratory findings is very important. Analytical results are monitored by quality control procedures. For measuring the performance of medical findings the concept of the 'validity' of a laboratory test is used. Validity means the 'degree of achieving the objective'. Accordingly, a valid laboratory finding is one which correctly answers the question which the physician at the sick-bed directs to the laboratory. Quantitative measures for the validity of interpretation can be developed by an analysis of the underlying classification processes. Characteristic indices describing the validity quantitatively in terms of conditional probabilities can be derived from decision tables. Examples of 'validity indices' are diagnostic (or prognostic) 'sensitivity' and 'specificity'. These indices are powerful tools for developing strategies for the clinical use of laboratory examinations in diagnosis, prognosis and therapy management. Moreover, validity indices are appropriate output quantities for the estimation of effectiveness and efficiency of a diagnostic or prognostic examination.