Welcome to the NMS chemical and biological metrology website. Please
log in or
register to view restricted content.
Principle 5
Analytical measurements made in one location should be consistent with those elsewhere.
Why Is It Important?
- Disagreements between companies or other organisations over analytical measurements wastes time and costs money.
If a supplier and purchaser reach different conclusions about a product or a service in which they are trading, both will have to bear costs in resolving their differences. The laboratories used may lose business or face legal costs if the dispute escalates. Regulations cannot be fairly enforced if the relevant analytical data show inconsistencies. - Undetected long term inconsistency may be misinterpreted as real and significant effects.
Results from process QC analyses which drift over time may be misinterpreted as process variations, and trigger unnecessary investigative work, possibly even halting production processes. Local contamination might go undetected without a firm external reference comparison. Long term studies with environmental or health implications need long term consistency to make sure that no false trends are deduced from the analytical results.
What To Do
- Check or calibrate the entire method using relevant, stable, well established reference materials.
A well characterised material, preferably a certified reference material (CRM), should be put through the whole analytical procedure in the same way as a test sample. This will provide a good single point calibration of the method if the reference material closely resembles customer samples in terms of analyte/matrix combination and concentration.
The use of stable reference materials as quality control samples for the analysis in association with control charts will show up short term and long term drift. - Ensure all instruments used are calibrated with traceable reference standards.
Traceable reference standards are working standards which have been calibrated against national or international standard materials or methods. Traceability is needed to allow valid comparisons to be made with the results from other laboratories using similarly calibrated standards. - Understand the factors which contribute to uncertainty in the method and estimate the overall uncertainty in your results.
Laboratories which underestimate the uncertainties in their results risk more disputes through over-interpretation (see Figure 1).
Figure 1: Consistency and Uncertainty
Laboratories A and B, with under estimated uncertainties (based, for example, only on run to run precision), clearly do not agree on their findings for the material in question; the results are not consistent. Action would be taken to find out why.
Laboratories C and D produced the same results as A and B respectively, but with realistically assessed uncertainties, including allowance for systematic effects and long term reproducibility. They correctly interpret the results as showing no significant difference; their results are consistent with one another This does not imply that larger uncertainties are better; clearly they are not. But correct evaluation of all the contributions to uncertainty is essential to establishing consistency of results.
Laboratories which overestimate uncertainty will tend to do more work than necessary to ensure adequate accuracy.
Understanding the method and the sources and magnitude of component uncertainties is important if the uncertainty in the final results needs to be reduced to meet requirements. - Try to identify laboratories carrying out similar work and investigate the possibilities of collaboration.
Where no other means of forming an external assessment exists, you should use contacts such as professional bodies to see if informal intercomparison exercises can be set up. Sections which deal with analytical science exist in professional and industrial bodies and can be important sources of advice and a route to ad-hoc intercomparisons.
Additional Information
Certified reference materials (CRMs) and reference standards are highly specialised materials. Unfortunately, there will be many measurements for which appropriate reference materials do not exist. In this case, it is important to apply alternative approaches in order to underpin consistency and comparability. One possibility is to develop your own reference material. This should be characterised by the application of more than one well-established method. The wider the variety of methods which are used the more reliable is the material likely to be as a reference point. Collaboration and sharing of materials and results between expert laboratories is also highly recommended as a procedure for helping to ensure comparability and consistency.
Example
Improved competitiveness through traceable standards
Aircraft engines are sold into a world-wide market and therefore measurements of their emissions to atmosphere need to be accurate and internationally accepted. Through a series of international comparisons, some measurements made by UK organisations were found to be in error. These errors were attributed to the fact that the monitoring instruments were calibrated with inaccurate, commercially produced, gas mixtures which lacked traceability to recognised national or international standards. Subsequently nationally traceable gas standards were employed which resulted in internationally comparable measurements and significant economic benefits to the UK aircraft industry.
Supporting resources
Last modified on
28 April 2009.