Quality specifications based on the uncertainty of measurement.

Abstract

The result of a measurement needs to be accompanied by information on the error or uncertainty of the measurement. Either of two very closely related concepts is thus required to indicate the usefulness and relevance of a result for its purpose: decision making. This presentation will outline the difference between the concept of error and that of uncertainty and deal with procedures to estimate the latter. The use of a spreadsheet program will illustrate how the combined uncertainty is estimated in a relatively complex situation with a bearing on laboratory medicine.