Technical guidelines for the production and acceptance of a European Reference Material

General principles

  • Every European Reference Material project must be executed in accordance with internationally accepted guidelines such as ISO-Guides 30-35 (and more particularly ISO Guide 34, specifying the general requirements for the competence of reference material producers), WHO Guidelines (for clinical reference materials), the ISO-Guide to the Expression of Uncertainty in Measurement (GUM), etc.

  • Details of a finalised European Reference Material project are summarised in a submission document, which specifies the selection of base material, technical aspects of the production process of the candidate material and its subsequent storage, homogeneity control, short- and long-term stability testing, characterisation of the batch, long-term storage conditions, and planned stability monitoring. This document is submitted to the ERM technical committee for review and technical acceptance. Finally, the submission is forwarded to the ERM panel for approval, and the material can be released for sales.

European Reference Materials samples

  • European Reference Materials sample containers (bottles, vials, ampoules, etc.) have to be tight and must contain not less than the stated amount of material.

  • Samples must have appropriate homogeneity and stability for their intended use.

  • Whenever relevant (e.g., in case of clinical ERM-CRMs), the commutability of the European Reference Material has to be assured.

  • Packaging must be appropriate for the purpose.

  • Labelling must be clear and of good quality.

Homogeneity testing

  • Between-bottle homogeneity must be examined. Previous experience might be available demonstrating that a specific CRM can be prepared with negligible heterogeneity (for instance some solutions). However, even in such cases homogeneity testing is required to confirm the appropriate homogeneity of the material.

  • Any possible between-bottle heterogeneity must be accounted for in the overall estimated CRM uncertainty. This requirement applies even when no statistically significant between-bottle variation is present. In this case the method variation or the actual calculated between-bottle variation (whichever is larger) should be included in the overall uncertainty.

Stability testing

  • Long-term stability must be positively demonstrated by appropriate statistical extrapolation for the certified European Reference Material's shelf-life to be within the certified uncertainty; the related uncertainty contribution shall be part of the estimated CRM uncertainty.

  • Exceptions can be made, when instability of the material can be considered negligible (element composition of steels or alloys). In such cases the uncertainty contribution of long-term stability is set zero. In all other cases this parameter is determined from regression analysis of the results of the long-term stability study.

  • Whenever possible, stability studies should be carried out using isochronous measurements.

Batch characterisation

  • When a quantity that is not method dependant is certified, batch characterisation can be carried out

  1. by a primary method of measurement, or

  2. by laboratory inter-comparison, using, whenever possible, two or more independent analytical methods.

  • When a method-dependant quantity is certified by a laboratory inter-comparison, all laboratories must strictly comply with the imposed standard operating procedure.

  • The methods used must:

  1. be applied under metrologically valid conditions,

  2. have been properly validated before use (see ISO/IEC 17025: 5.4.5)

  3. have precision and accuracy compatible with the target uncertainty.

  • Each set of measurements must:

  1. be traceable to the stated references, and

  2. be accompanied by a GUM-compatible uncertainty statement whenever possible.

  • Participant laboratories must:

  1. have the required competence for the execution of the task in the European Reference Material project,

  2. be able to achieve traceability to the required stated references,

  3. be able to estimate its measurement uncertainty in accordance with GUM,

  4. have in place a sufficient and appropriate quality assurance system.

  • Estimation of uncertainty of batch-characterisation:

  1. Whenever possible, standard uncertainties of batch-characterisation (uchar ) by both primary methods of measurement or laboratory inter-comparisons have to be based on the combination of detailed uncertainty budgets of the contributing values.

  2. If insufficient information is available for values based on laboratory inter-comparisons, uchar can, in the absence of uncertainties common to all participants (e.g., common standard), also be approximated by the standard deviation of the mean of laboratory means.

  3. be able to estimate its measurement uncertainty in accordance with GUM,

  4. If such common uncertainties are present, they have to be added to the uncertainty of the mean of laboratory means.

Establishment of a certificate

  • The project responsible of a European Reference Material project establishes a draft certificate (possibly complemented by a certification report) in accordance with ISO Guide 30, containing all information relevant to and needed by the user.

  • Certified values must be traceable to stated references and be accompanied by a GUM-compatible expanded uncertainty statement valid for the entire shelf-life of the European Reference Material.

  • After approval by the European Reference Materials panel, the certificate and report are made available, both on the European Reference Materials-web site and in printed form, and the European Reference Material is put on sale.

Transfer of samples to final storage place

  • To avoid degradation after certification, all samples are best transported to the location which offers the conditions designated for the final storage of the European Reference Material before measurements are started.

  • Otherwise, they must be transported from door to door keeping them at all times under such storage conditions for which it has been demonstrated that there is no influence on the certified values.

Latest update 1 March, 2007