Introduction

It is important and required by ISO/IEC 17025:1999 that analysts are aware of the uncertainty associated with each analytical result and estimates that uncertainty. The measurement uncertainty may be derived by a number of procedures. Food analysis laboratories are required, for Codex purposes, to be in control¹, use collaboratively tested methods when available, and verify their application before taking them into routine use. Such laboratories therefore have available to them a range of analytical data which can be used to estimate their measurement uncertainty.
These guidelines only apply to quantitative analysis.
Most quantitative analytical results take the form of “a ± 2u or a ± U” where “a” is the best estimate of the true value of the concentration of the mesurand (the analytical result) and “u” is the standard uncertainty and “U“ (equal to 2u) is the expanded uncertainty. The range “a ± 2u” represents a 95% level of confidence where the true value would be found. The value of “U“ or “2u” is the value which is normally used and reported by analysts and is hereafter referred to as “measurement uncertainty” and may be estimated in a number of different ways.
Terminology
The international definition for Measurement Uncertainty is:
"Parameter, associated with the result of a measurement, that characterises the dispersion of the values that could reasonably be attributed to the measurand”²
NOTES:
1. The parameter may be, for example, a standard deviation (or a given multiple of it), or the half-width of an interval having a stated level of confidence.
2. Uncertainty of measurement comprises, in general, many components. Some of these components may be evaluated from the statistical distribution of results of a series of measurements and can be characterised by experimental standard deviations. The other components, which can also be characterised by standard deviations, are evaluated from assumed probability distributions based on experience or other information.
3. It is understood that the result of a measurement is the best estimate of the value of a measurand, and that all components of uncertainty, including those arising from systematic effects, such as components associated with corrections and reference standards, contribute to the dispersion. .”

Recommendations

1. The measurement uncertainty associated with all analytical results is to be estimated.
2. The measurement uncertainty of an analytical result may be estimated by a number of procedures, notably those described by ISO (1) and EURACHEM (2). These documents recommend procedures based on a component-by-component approach, method validation data, internal quality control data and proficiency test data. The need to undertake an estimation of the measurement uncertainty using the ISO component-by-component approach is not necessary if the other forms of data are available and used to estimate the uncertainty. In many cases the overall uncertainty may be determined by an inter-laboratory (collaborative) study by a number of laboratories and a number of matrices by the IUPAC/ISO/AOAC INTERNATIONAL (3)or by the ISO 5725 Protocols (4).
3 The measurement uncertainty and its level of confidence must, on request, be made available to the user (customer) of the results.

References

1. “Guide to the Expression of Uncertainty in Measurement”, ISO, Geneva, 1993.
2. EURACHEM/CITAC Guide Quantifying Uncertainty In Analytical Measurement (Second Edition), EURACHEM Secretariat, BAM, Berlin, 2000. This is available as a free download from http://www.eu rachem.ul.pt/
3. “Protocol for the Design, Conduct and Interpretation of Method Performance Studies”, ed. W. Horwitz, Pure Appl. Chem., 1995, 67, 331-343.
4. “Precision of Test Methods”, Geneva, 1994, ISO 5725, Previous editions were issued in 1981 and 1986.

¹ As outlined in Codex GL 27-1997 “Guidelines for the Assessment of the Competence of Testing Laboratories Involved in the Import and Export of Food"

² International vocabulary of basic and general terms in metrology, ISO 1993, 2^nd Edition.