ANNEX IISupplementary provisions
ANNEX II ‘ANNEX II GENERAL PROVISIONS ON METHODS OF ANALYSIS FOR FEED A. PREPARATION OF SAMPLES FOR ANALYSIS 1. Purpose The procedures described in this Annex concern the preparation for analysis of samples, sent to the control laboratories after sampling in accordance with the provisions laid down in Annex I. The laboratory samples must be prepared in such a way that the amounts weighed out, as provided for in the methods of analysis, are homogeneous and representative of the final samples. In addition to the procedures described in this Annex, the guidelines for sample preparation as provided for by EN ISO 6498 shall be followed. 2. Precautions to be taken The sample preparation procedure to be followed is dependent on the methods of analysis to be used and the constituents or substances to be controlled. It is therefore of major importance that that the followed sample preparation procedure be appropriate for the used method of analysis and for constituents or substances to be controlled. All the necessary operations must be performed in such a way as to avoid as far as possible contamination of the sample and changes of its composition. Grinding, mixing and sieving shall be carried out without delay with minimal exposure of the sample to the air and light. Mills and grinders likely to appreciably heat the sample shall not be used. Manual grinding is recommended for feed which are particularly sensitive to heat. Care shall also be taken to ensure that the apparatus itself is not a source of contamination. Homogenisation of the sample by preparing a slurry by high shear mixing with water has proven to provide in certain cases more homogeneous sub-samples than dry homogenisation/grinding, in particular in case of heterogeneously distributed chemical substances. However also homogenisation by sufficient dry grinding might provide homogeneous subsamples. In certain cases, such as for the determination of rye ergot, harmful botanical impurities, etc., the homogenisation of the sample cannot be done by grinding but by sufficiently mixing the sample. If the preparation cannot be carried out without significant changes in the moisture content of the sample, determine the moisture content before and after preparation according to the method laid down in Part A of Annex III. 3. Procedure 3.1. General procedure The test aliquot is taken from the final homogenised sample. Coning and quartering is not recommended because this might provide test aliquots with high splitting error. 3.1.1. Feed which can be ground as such — Mix the final sample and collect it in a suitable clean, dry container fitted with an air-tight stopper. Mix again in order to ensure full homogenisation, immediately before weighing out the amount for analysis (test aliquot). 3.1.2. Feed which can be ground after drying — Unless otherwise specified in the methods of analysis, dry the final sample to bring its moisture content down to a level of 8 to 12 %, according to the preliminary drying procedure described under point 4.3 of the method of determination of moisture mentioned in Part A of Annex III). Then proceed as indicated in point 3.1.1. 3.1.3. Liquid or semi-liquid feed — Collect the final sample in a suitable clean, dry container, fitted with an air-tight stopper. Mix thoroughly in order to ensure full homogenisation immediately before weighing out the amount for analysis (test aliquot). 3.1.4. Other feed — Final samples which cannot be prepared according to one of the above procedures shall be treated by any other procedure which ensures that the amounts weighed out for the analysis (test aliquot) are homogeneous and representative of the final samples. 3.2. Specific procedure in case of examination by visual inspection or by microscopy or in cases where the whole aggregate sample is homogenised — In case of an examination by visual inspection (without making use of microscope), the whole aggregate or final sample is used for examination. — In case of a microscopic examination, the laboratory may reduce the aggregate sample, or further reduce the reduced sample. The final samples for defence and possibly reference purposes are taken following a procedure equivalent to the procedure followed for the final sample for enforcement. — In case the whole aggregate sample is homogenised, the final samples are taken from the homogenised aggregate sample. — For the determination of rye ergot and harmful botanical impurities, the final sample has to be divided into 2 subsamples of equal weight of approximately 500 grams. One subsample is examined. In case the result of the subsamples is equal or below 50 % (analytical threshold) of the maximum level, the sample is compliant with the maximum level. If the result is above 50 % of the maximum level, another subsample needs to be examined and the average of the result of the 2 subsamples is used for checking compliance with the maximum level. 4. Storage of samples Samples must be stored at a temperature that will not alter their composition. Samples intended for the analysis of vitamins or substances which are particularly sensitive to light shall be stored in such conditions that the sample is not adversely affected by light. B. PROVISIONS RELATING TO REAGENTS AND APPARATUS USED IN METHODS OF ANALYSIS 1. Unless otherwise specified in the methods of analysis, all analytical reagents must be analytically pure (a.p.). When trace analysis is carried out, the purity of the reagents must be checked by a blank test. Depending upon the results obtained, further purification of the reagents may be required. 2. Any operation involving preparation of solutions, dilution, rinsing or washing, mentioned in the methods of analysis without indication as to the nature of the solvent or diluent employed, implies that water must be used. As a general rule, water shall be demineralised or distilled. In particular cases, which are indicated in the methods of analysis, it must be submitted to special procedures of purification. 3. In view of the equipment normally found in control laboratories, only those instruments and apparatus which are special or require specific usage are referred to in the methods of analysis. They must be clean, especially when very small amounts of substances have to be determined. C. APPLICATION OF METHODS OF ANALYSIS AND EXPRESSION OF THE RESULTS 1. Extraction procedure Several methods determine a specific extraction procedure. As a general rule, other extraction procedures than the procedure referred to in the method can be applied on the condition that the used extraction procedure has been proven to have the equivalent extraction efficiency for the matrix analysed as the procedure mentioned in the method. 2. Clean-up procedure Several methods determine a specific clean-up procedure. As a general rule, other clean-up procedures than the procedure referred to in the method can be applied on the condition that the used clean-up procedure has been proven to result in equivalent analytical results for the matrix analysed as the procedure mentioned in the method. 3. Number of determinations In case of the analysis of undesirable substances, if the result of the first determination is significantly (> 50 %) lower than the specification to be controlled, no additional determinations are necessary, on the condition that the appropriate quality procedures are applied. In other cases a duplicate analysis (second determination) is necessary to exclude the possibility of internal cross-contamination or an accidental mix-up of samples. The mean of the two determinations, is used for further assessment. In case of the control of minimum or maximum levels of feed additives, if the results of the first determination is above the minimum level or below the maximum level no additional determinations are necessary, on the condition that the appropriate quality procedures are applied. In other cases, a duplicate analysis (second determination) is necessary to exclude the possibility of internal cross-contamination or an accidental mix-up of samples. The mean of the two determinations is used for further assessment. In case of the control of the declared content of a substance or ingredient, if the result of the first determination confirms the declared content, i.e. the analytical result falls within the acceptable range of variation of the declared content, no additional determinations are necessary, on the condition that the appropriate quality procedures are applied. In other cases a duplicate analysis (second determination) is necessary to exclude the possibility of internal cross-contamination or an accidental mix-up of samples. The mean of the two determinations, is used for further assessment (the average analytical result falls or not within the acceptable range of variation of the declared content). In some cases this acceptable range of variation is defined in legislation such as in Regulation (EC) No 767/2009 and Regulation (EU) 2019/4 of the European Parliament and of the Council ( 1 ) . 4. Reporting of the method of analysis used The analysis report shall mention the method of analysis used. 5. Reporting of the analytical result The analytical result shall be expressed in the manner laid down in the method of analysis to an appropriate number of significant figures and shall be corrected, if necessary, to the moisture content of the final sample prior to preparation. Most regulatory levels (e.g. maximum level, minimum level) in EU animal feed legislation are established relative to a feed with a moisture content of 12 %. Therefore, in these cases, in order to assess the analytical result measured on the sample against the regulatory level, the analytical result first needs to be divided by the dry matter content of the sample (in %) multiplied by 88, as indicated in the following formula: where: Mc : moisture content of the sample (in %). 100 – Mc therefore represents the dry matter content of the sample (in %). R ana : analytical result as measured on the sample. R 12 % : result for a feed with a moisture content of 12 %; to be assessed against the regulatory level. In addition, if the following conditions are met: — the result of the analysis is significantly (> 50 %) lower or higher than the labelling information/specification to be controlled (depending on whether the labelling information/specification is a maximum or a minimum level), — the moisture content of the sampled feed is known and it can be determined that correction to the moisture content will not change the assessment, then, on the condition that the appropriate quality procedures are applied and the analysis serves only the purpose of checking compliance with legal provisions, the correction to the moisture content might be omitted (e.g. in cases there is no specification or regulatory level), unless it is required for interpretation. If the analytical result is corrected to the moisture content, the corresponding measurement uncertainty must also be corrected in the same procedure. In case of the determination of rye ergot or harmful botanical impurities by visual/microscopic examination correction to the moisture content is not necessary. 6. Analytical measurement uncertainty and recovery rate in case of analysis of undesirable substances As regards undesirable substances within the meaning of Directive 2002/32/EC, a product intended for animal feed shall be considered as non-compliant with the established maximum content, if the analytical result as a mean of two independent determinations, relative to a feed with a moisture content of 12 %, is deemed to exceed the maximum content taking into account expanded analytical measurement uncertainty using a coverage factor of 2 which gives a level of confidence of approximately 95 % and correction for recovery. This means, in order to assess compliance, the analysed concentration is used after being corrected for recovery and after deduction of the expanded analytical measurement uncertainty. This procedure is only applicable in cases where the method of analysis enables the estimation of the expanded analytical measurement uncertainty and correction for recovery (e.g. not required in case of visual/microscopic examination). If the analytical result of the sample taken for defence exceeds the maximum content (without taking into account the expanded analytical measurement uncertainty), this confirms the non-compliance established with the control sample, in the absence of specific national rules on this. The analytical result shall be reported as follows (in so far the method of analysis used enables to estimate the expanded analytical measurement uncertainty): (a) corrected for recovery, where appropriate and relevant, and when corrected it has to be so stated. The recovery rate is to be quoted unless intrinsic correction for bias is part of the procedure, whereby bias is the difference between the measured value and the reference concentration. The correction for recovery is not necessary in case the recovery rate is between 90-110 %; (b) as “x +/- U”, whereby x is the analytical result and U is the expanded analytical measurement uncertainty, using a coverage factor of 2 ( 2 ) which gives a level of confidence of approximately 95 %. However, if the result of the analysis is significantly (> 50 %) lower than the specification to be controlled, and on the condition that the appropriate quality procedures are applied and the analysis serves only the purpose of checking compliance with legal provisions, the reporting of the recovery rate and expanded analytical measurement uncertainty might be omitted (e.g. in cases there is no specification or regulatory level), unless the measurement uncertainty is required for interpretation. 7. Analytical measurement uncertainty and recovery rate in case of analysis of content of feed additives In order to check compliance with authorised minimum and maximum content of feed additives, the presence of a feed additive shall be considered as non-compliant with the established minimum and maximum content, if the analytical result as mean of two independent determinations, relative to a feed with a moisture content of 12 %, is deemed to: — exceed the maximum content taking into account expanded analytical measurement uncertainty and correction for recovery. This means, in order to assess compliance, the analysed concentration (i.e. mean of two determinations) is used after being corrected for recovery and after deduction of the expanded analytical measurement uncertainty, — be lower than the minimum content taking into account the expanded analytical measurement uncertainty and correction for recovery. This means, in order to assess compliance, the analysed concentration (i.e. mean of two determinations) is used after being corrected for recovery and after the addition of the expanded analytical measurement uncertainty. If the analytical result of the sample taken for defence exceeds the maximum content (without taking into account the expanded analytical measurement uncertainty), this confirms the non-compliance established with the control sample, in the absence of specific national rules on this. The analytical result shall be reported as follows (in so far the method of analysis used enables to estimate the expanded analytical measurement uncertainty): (a) corrected for recovery, where appropriate and relevant, and when corrected it has to be so stated. The recovery rate is to be quoted unless intrinsic correction for bias is part of the procedure, whereby bias is the difference between the measured value and the reference concentration. The correction for recovery is not necessary in case the recovery rate is between 90-110 %; (b) as “x +/- U”, whereby x is the analytical result (mean of two determinations) and U is the expanded analytical measurement uncertainty, using a coverage factor of 2 ( 3 ) which gives a level of confidence of approximately 95 %. ’. ( 1 ) Regulation (EU) 2019/4 of the European Parliament and of the Council of 11 December 2018 on the manufacture, placing on the market and use of medicated feed, amending Regulation (EC) No 183/2005 of the European Parliament and of the Council and repealing Council Directive 90/167/EEC ( OJ L 4, 7.1.2019, p. 1 ). ( 2 ) The confidence interval of 95 % can be achieved by making use of another factor such as the t-factor. ( 3 ) The confidence interval of 95 % can be achieved by making use of another factor such as the t-factor.