Analytical procedures validation

For physical and chemical measurements it is essential that measurements are referenced to standards accepted by all the laboratories undertaking a particular type of measurement. In particular, the use of chemical analytical procedures validated through their application to certified reference materials (CRM) (26) is highly recommended. It is to be noted that a number of CRMs prepared with Antarctic matrices are already available, i.e., marine sediment and krill, or in preparation (26-29). These and other CRMs should be also used routinely by the participating laboratories to assure a periodical assessment of accuracy and repeatability of measurements. These laboratories should also undertake regular intercalibration studies. It is possible that in the first application of these exercises systematic errors will be found, but better results are expected in subsequent rounds together with a general improvement in the performance of laboratories and data quality. 

The logic and consequences of such an approach are discussed, with background definitions and terminology as a starting point. Close attention is paid to the concept of measurement uncertainty as providing a single-number index of accuracy inherent in the procedure. The appropriateness of the uncertainty-based approach to analytical measurement is stressed in view of specific inaccuracy sources such as sampling and matrix effects. And methods for their evaluation are outlined. The question of a clear criterion for analytical procedure validation is also addressed from the standpoint of the quality requirement which measurement results need to meet as an end-product. 

---