Publikationsdatenbank

Harmonization - Horizontal approach in collaborative studies (2004)

Art

Vortrag

Autor

Weiß, Hartmut

Kongress

Standardization in the 3rd Millenium

Parma/Italien, 21.04.2004

Quelle

Sprache

Englisch

Kontakt

Institut für Veterinär-Epidemiologie und Biometrie

Königsweg 67
14163 Berlin
+49 30 838 56034
epi@vetmed.fu-berlin.de

Abstract / Zusammenfassung

During the last ten years we can observe a successful development of harmonized protocols for collaborative studies between the responsible organisations. Some points are not harmonized as you can see for example, when comparing standards dealing with the organization and evaluation of collaborative studies (IDF Standard 135B:1991 and ISO Standard 5725:1994). There are some differences regarding the selection of laboratories, the graphical presentation of data, the criteria for outlier identification and the use of “robust” statistical procedures, if the assumption of an underlying uniformly normal distribution is violated.
All these points are important with respect to the development of horizontal approaches in collaborative studies for the validation of quantitative microbiological techniques. In microbiological work it is not sufficient to apply qualitative presence-absence-techniques. We need in many different fields validated methods comparing samples with defined limits.
Within my presentation I shall compare reported results of various collaborative studies – performed within the last years – reanalysed with different statistical strategies, including the “robust” statistical procedures drafted as IDF/ISO Standard “Microbiology of food and animal feeding stuffs – Protocol for the establishment of precision characteristics of quantitative methods by interlaboratory studies”. I shall try to draw practical conclusions with respect to the investigated matrix and the sources of measurement uncertainty.