Page 29 - FoodFocusThailand No.161 August 2019

SPECIAL FOCUS

Accuracy Accuracy is the closeness of test results Limit of Detection (LOD) and Limit of Quantitation (LOQ)
1obtained by test method to the true value. Certified 4LOD and LOQ are two important performance characteristics in method
Reference Material (CRM) is highest recommended to be validation, especially for analysis of trace amount of food toxicants/contaminants.
analyzed at least 7 determinations whenever possible LOD is the smallest concentration of analyte that can be detected, but not necessarily
because their highest level of traceability. In case of CRM quantified. While LOQ is the smallest quantified concentration of analyte in the
is limited, spiked sample method can be a good representation, sample that can be determined with acceptable precision and accuracy. This
at least 3 spiked concentrations (with 7 determinations for approach could only be applied to analytical procedure which exhibits baseline
each concentration) are recommended. noise by determination of Signal-to-Noise ratio (S/N). LOD and LOQ are generally
expressed as 3 and 10 times of standard deviation of response of blank sample or
Precision Precision is the determination of closeness with known low concentrations, respectively.
2of agreement replicated and independent test results,
Ruggedness Ruggedness is a parameter expressed the degree of
obtained under stipulated conditions. Repeatability is the 5reproducibility of test results obtained by the analysis of the same samples
precision obtained under “same test condition”, same
laboratory, same operator, same equipment. While under a variety of conditions, such as different elapsed testing time, different analyst,
different instrument, different pH of water, or different assay temperature.
reproducibility is another precision parameter obtained when
test results are obtained with the same method, but “under Additional Guidelines for Method Validation
changing test condition” e.g. different laboratory, different When the laboratory have used a fully validated method, but new food matrix(s)
operator, or different equipment. Precision is usually are applied, or in situation that the single laboratory has developed fully-validated
expressed in terms of Standard Deviation (SD) or Coefficient method, but no collaborative study has not yet applied; in these cases, the laboratory
of Variation (CV) of the test results. needs to extend the verification of accuracy, precision, linearity and matrix effect to
ensure the appropriate of test results. When the laboratory needs to use the test
Linearity and Range Linearity is its specification method that published in the scientific literature, however no characteristics given,
3 of test method to obtain test results which are directly in this case, laboratory needs to verify accuracy, precision, linearity as well as
proportional to the concentration/amount of analyte in the ruggedness. It is essential that appropriate validation (at least accuracy and
sample. A linear relationship shall be evaluated across the precision) is always be carried out before taking any new method into routine use.
range by visual inspection of a plot of signals as a function Verification is recommended to repeat in cases such as, when major instruments
of analyte concentration. For the establishment of linearity are replaced, new batches of major reagents are used, or even changing the
of any test, a minimum of 5 concentrations is recommended. operators.
While range of an analytical procedure is the interval between
the upper and lower of analytes concentration in the food เอกสารอ้างอิง/Reference
matrix for which it has been demonstrated that the analytical Eurachem Guide. 2014. The Fitness for Purpose of Analytical Methods, A Laboratory Guide to
procedure has a suitable level of precision, accuracy and Method Validation and Related Topic, 2nd edition.
linearity.

Method Validation Guidelines

for Chemical Analysis of Food Matrix

as Requirement of ISO/IEC17025

AUG 2019 FOOD FOCUS THAILAND 29

27-29_Special Focus_Kanitaporn.indd 29 20/7/2562 BE 14:02