Guide for implementation of the Standard EN ISO 19036:2019
A guide for implementation of the Standard EN ISO 19036: 2019 Microbiology of the food chain — Estimation of measurement uncertainty for quantitative determinations, has been prepared by the Working Group, reviewed by the respective networks of NRLs, and approved by the Standing Committee on Plants, Animals, Food and Feed.
According to the Guide on the implementation of the Standard EN ISO 19036:2019, matrix uncertainty corresponds to the uncertainty associated with taking the test portion from the laboratory sample; it is therefore different from the uncertainty linked to sampling (taking a sample from a batch of production). It is linked to the degree of the spatial heterogeneity of the microorganisms within the matrix. It is thus expected to be large for solid or multi-component matrices, and low for homogeneous matrices. It is assumed in the standard as being independent of the analytical method used (including target bacteria) and of the laboratory estimating it.
The standard allows using values of matrix uncertainty already obtained for laboratory samples expected to have a similar matrix uncertainty (matrix homogeneity). We thus suggest using values of matrix uncertainty, per matrix type, obtained in a collaborative study organised by the three EURLs together with the associated NRLs networks.
The matrix uncertainty values obtained in this collaborative study can be used by any laboratory for the calculation of combined uncertainty when analysing the same matrices or matrices expected to have a similar matrix uncertainty.
How to provide additional data:
We aim at continuing this study and enlarging the list of matrices and corresponding values of matrix uncertainty. To submit new data, please contact any of the three EURLs by email (Campylobacter, CPS and Lm) to receive the experimental design and the file for reporting data. Prerequisites to include new data are that:
- experimental design followed the setup according to EN ISO 19036 (clause 6.3),
- method as described in the report has been followed,
- valid estimates of matrix uncertainty have been retrieved,
- values correspond to matrices not already included in the list.