GB 5009.295-2023National food safety standard - General rules for verification of chemical analysis methods (English PDF)

Foreword

This document was issued on 6 September 2023 by the State Administration for Market Regulation; Standardization Administration of the PRC and takes effect on 6 September 2024.

It is a GB standard without the /T suffix: compliance is mandatory in China.

1 Scope

GB 5009.295-2023 governs the evidence a chemical analysis method has to produce before it is written into the national food safety standard series and relied on in testing. The vocabulary is fixed first, and it is the part laboratories argue about: method validation, qualitative and quantitative methods, the concentration paid attention to, specificity, detection limit, limit of quantification, measuring range, accuracy, repeatability, reproducibility, intermediate precision, stability, robustness, linear range and detection probability. General requirements follow — which performance parameters a given method must be examined against, how representative foods are selected for the trial, what the validation samples have to satisfy, and how the resulting data are processed. The validation clause then takes each parameter in turn, from specificity through to robustness, and adds the separate case of proving one method equivalent to another. Annex A classifies food samples for validation purposes; Annex B covers the robustness test. Without a common procedure a method looks sound in the laboratory that developed it and falls apart on a different matrix or in a different pair of hands, and a disputed result turns into an argument about the method rather than about the food. For method developers, food inspection laboratories, accreditation assessors and the technical committees that adopt methods into the series.

This Standard specifies the general requirements for the validation of national food

safety standard -- chemical analysis methods.

This Standard is applicable to the validation process during the development and

revision of national food safety standard -- chemical analysis methods.

2 Terms and definitions

2.1 Method validation

Test activities that provide objective and valid evidence to confirm that the performance

parameters of an analytical method meet the intended use of the method.

2.2 Qualitative method

An analytical method that identifies the presence or absence of a target analyte in a

sample based on its specific properties.

2.3 Quantitative method

An analytical method that determines the concentration or amount of target analytes

under conditions that meet specified precision and accuracy.

2.4 Concentration paid attention to

The concentration that is decisive for determining whether the target analyte in the

sample meets the requirements of regulations or standards.

2.5 Specificity

The ability of an analytical method to distinguish between target analytes and

interfering substances.

2.14 Intermediate precision

In the same laboratory, over a long period of time (pay attention to small changes in test

conditions, such as temperature, humidity, light, etc.), the precision of independent

measurement results obtained by using the same method to test the same sample.

2.15 Linear range

Under a certain level of significance, a linear calculation model is used to define the

relationship between instrument response and concentration (or amount), which

conforms to the interval between the lowest concentration (or amount) and the highest

concentration (or amount) of the calculation model.

2.16 Detection probability

The probability that an analytical method will detect the target analyte at a specified

concentration and confidence level.

3 General requirements for validation of chemical analysis

methods

3.1 Selection of performance parameters

3.1.1 For qualitative methods, intra-laboratory and inter-laboratory validation

parameters shall include method specificity and detection limit.

3.1.2 For quantitative methods, laboratory validation parameters shall include method

specificity, detection limit, quantitation limit, measurement range, accuracy, and

repeatability. Interlaboratory validation parameters shall include the detection limit,

quantification limit, determination range, correctness and reproducibility of the method.

3.1.3 When the analysis method is sensitive to small changes in test conditions such as

temperature, humidity, and lighting, the robustness of the method shall be examined,

and the intermediate precision index shall be calculated.

3.1.4 When there are two or more methods with the same applicable scope for the same

target analyte, the equivalence of the different methods needs to be validated.

3.1.5 When literature data on the stability of a reference material are lacking, its stability

shall be validated.

3.1.6 Selection of intra-laboratory and inter-laboratory validation parameters is shown

use of the method.

b) For a sample matrix, at least 3 concentration levels are required, including the

lowest concentration, the middle concentration (generally the concentration of

concern) and the highest concentration of the method's determination range. Their

uniformity and stability are tested.

3.3.4 When selecting a standard addition sample for validation, the validation sample

can be prepared as follows.

a) The sample matrix shall be selected to be representative and meet the intended

use of the method.

b) For a sample matrix, at least 3 concentration addition levels are required,

including the lowest concentration, the middle concentration (generally the

concentration of concern) and the highest concentration of the method

determination range.

3.4 Data processing requirements

All laboratory validation data need to be subjected to outlier testing and statistical

analysis. The analysis method can use statistical methods such as Grubbs test. After

eliminating outliers through technical analysis, the number of laboratories and the

amount of validation data shall meet statistical requirements.

4 Validation of performance parameters of chemical analysis

methods

4.1 Validation of method performance parameters

4.1.1 Specificity

4.1.1.1 Validation requirements

The ability of the analytical method to differentiate between the target analyte and other

substances shall be validated. There shall be no interference factors (signals,

interference peaks, etc.) that lead to errors in the identification of the target analyte and

affect the accuracy of quantitation.

4.1.1.2 Validation method

Analyze representative samples that match the scope of the method. Check whether

there are any interferences (signals, peaks, etc.) that lead to incorrect identification of

the target analytes and/or affect accurate quantitation. A certain concentration of

substances that may interfere with the identification and/or quantification of the target

analyte is added to a representative sample and then analyzed. Check whether these

interfering substances can be effectively identified.

4.1.2 Detection limit

4.1.2.1 Validation requirements

The detection probability of the target analyte at the detection limit level shall be no

less than 95% (at the 95% confidence level). The detection limit needs to be

independently validated for each representative sample matrix.

4.1.2.2 Validation method

4.1.2.2.1 Estimation method

Methods for estimating detection limits include.

a) Signal-to-noise ratio method (applicable to spectroscopy, chromatography, and

mass spectrometry). Add target analytes to a blank sample matrix. Use the added

concentration when the signal-to-noise ratio is 3 as the estimated detection limit.

b) Blank standard deviation method (method suitable for non-zero response values

of blank samples). The detection limit is determined by analyzing a large number

of sample blanks. The number of independent tests shall be no less than 10 (n

>=10). Calculate the standard deviation of the sample blank response. The

concentration corresponding to the average of the sample blank response values

plus 3 times the standard deviation shall be the estimated detection limit.

c) Calibration equation estimation method (applicable to instrument method). If

sample data at or near the detection limit cannot be obtained, use 3S/b (S is the

standard deviation of the calibration curve, b is the slope of the calibration curve)

as the estimated detection limit.

d) Gradual dilution method (applicable to all methods). The target analyte is serially

diluted at certain concentration intervals. Test according to the method to be

validated. The concentration at which the target analyte can be detected shall be

the estimated detection limit.

4.1.2.2.2 Determination method

Select at least 20 parallel samples of blank sample matrix. Target analytes at estimated

detection limit concentrations are added separately. When the detection probability of

the target analyte is not less than 95%, it is determined as the detection limit. When

blank samples are unavailable, isotope-labeled compounds or structural analogs can be

used for validation.

4.1.3 Limit of quantitation

4.1.3.1 Validation requirements

The quantitation limit shall be validated using certified reference materials/standard

samples, quality control samples or standard spiked samples of the same concentration.

Its accuracy and precision shall meet method and regulatory requirements. The limit of

quantification plus 3 times the standard deviation of the sample at the concentration of

interest shall be less than the concentration of interest. For target analytes with limited

values, the limit of quantitation shall be 0.5 times the limit value or less.

4.1.3.2 Validation method

4.1.3.2.1 Estimation method

Methods for estimating quantitation limits include.

a) Add target analytes to blank sample matrix. Use the added concentration when the

signal-to-noise ratio is 10 as the estimated quantitation limit.

b) Three times the detection limit is used as the estimated quantification limit.

c) The concentration corresponding to the average of the sample blank response

values plus 10 times the standard deviation is used as the estimated quantification

limit.

4.1.3.2.2 Determination method

For each representative sample matrix, independent testing is performed using certified

reference materials/standard samples, quality control samples or standard spiked

samples at the quantitation limit concentration level. Test at least 6 replicate samples.

Perform statistical analysis of data. If the accuracy and precision of the measurement

results at this concentration meet the requirements of Table 2 to Table 4, this

concentration can be used as the quantitative limit of the method.

4.1.4 Measuring range

The measurement range of the method shall cover the concentration of interest of the

method. Its validation requirements are as follows.

a) The correctness and precision of at least the lowest and highest concentrations of

the method's measurement range need to be validated. For target analytes with

limited limit values, the accuracy and precision of the limit value concentration

shall also be validated.

b) If the measurement range of the method is linear and the standard curve method

Depending on the properties of the standard solution or sample solution, choose

different storage conditions and time intervals. Use the same analysis method and

analysis conditions to measure the target analyte in the standard solution or sample

solution.

4.1.7.3 Evaluation of stability validation results

4.1.7.3.1 For analytical target substances with only two time point detection results, the

stability can be evaluated using formula (1).

Where,

rhoi - the measured concentration of the target analyte at time point i;

rho0 - the initial measured concentration of target analyte;

UE - the expanded uncertainty of the instrument method (excluding sample pretreatment

steps) at rho0 concentration, UE<=15%.

4.1.7.3.2 For analytical target substances with multiple time point test results, the

stability can be evaluated using the t test or F test.

4.1.8 Robustness

Robustness testing is a testing procedure that examines the impact of environmental or

other conditional variables on an analytical method. A robust analysis method shall not

be subject to significant changes caused by small changes in environmental factors, that

is, it shall be insensitive to changes in environmental factors such as temperature,

humidity, and air pressure, as well as to other factors such as the amount of reagent

added and reaction time. The robustness test of the method is shown in Annex B.

4.2 Validation of method equivalence

4.2.1 Validation requirements

In the analytical method standards, there shall be no significant difference in the test

results of different analytical methods with the same measurement range.

4.2.2 Validation method

4.2.2.1 Select at least 3 representative samples. For each representative sample, at least

6 standard substances/standard samples or standard spiked samples with concentrations

......