[Concept of reliable laboratory test procedures (author's transl)].

Considering the multitude of tests in clinical chemistry and hematology, their low diagnositc sensitivity and specifity, and the low incidence of the diseases in a non selected population it is not possible presently to recommend a general screening procedure based on the definition of the reference, the methodological reproducibility and the diagnostic information of a test procedure it is possible to calculate the number of requests necessary to get any desired number of pathological results. "Tell me how many pathological results you want and I'll tell you how many requests you need". In order to get an optimal diagnostic information for clinical chemistry and hematological tests, it is necessary to increase the prevalence by proper selection of the patients and then to order test procedures specific for the suspected disease.