Log transformation of proficiency testing data on the content of genetically modified organisms in food and feed samples: is it justified?

The outcome of proficiency tests (PTs) is influenced, among others, by the evaluation procedure chosen by the PT provider. In particular for PTs on GMO testing a log-data transformation is often applied to fit skewed data distributions into a normal distribution. The study presented here has challenged this commonly applied approach. The 56 data populations from proficiency testing rounds organised since 2010 by the European Union Reference Laboratory for Genetically Modified Food and Feed (EURL GMFF) were used to investigate the assumption of a normal distribution of reported results within a PT. Statistical evaluation of the data distributions, composed of 3178 reported results, revealed that 41 of the 56 datasets showed indeed a normal distribution. For 10 datasets, the deviation from normality was not statistically significant at the raw or log scale, indicating that the normality assumption cannot be rejected. The normality of the five remaining datasets was statistically significant after log-data transformation. These datasets, however, appeared to be multimodal as a result of technical/experimental issues with the applied methods. On the basis of the real datasets analysed herein, it is concluded that the log transformation of reported data in proficiency testing rounds is often not necessary and should be cautiously applied. It is further shown that the log-data transformation, when applied to PT results, favours the positive performance scoring for overestimated results and strongly penalises underestimated results. The evaluation of the participants' performance without prior transformation of their results may highlight rather than hide relevant underlying analytical problems and is recommended as an outcome of this study. Graphical abstract.

A reference method for determining the total allergenic protein content in a processed food: the case of milk in cookies as proof of concept.

The establishment of a reference method for the determination of the allergen protein content in a processed food material has been explored. An analytical approach was developed to enable the comparability of food allergen measurement results expressed in a decision-relevant manner. A proof of concept is here presented, resulting in quantity values for the common measurand, namely 'mass of total allergen protein per mass of food'. The quantities are determined with SI traceability to enable the comparability of reported results. A method for the quantification of total milk protein content in an incurred baked food at a concentration level clinically relevant is presented. The strategy on how to obtain the final analytical result is outlined. Challenges associated with this method are discussed, in particular the optimal extraction of the marker proteins, the complete digestion and release of the peptides in an equimolar fashion, the use of conversion factors to translate the amount of measured proteins into total milk protein and the estimation of the uncertainty contributions as well as of the combined uncertainty of the final result. The implementation of such a reference method for the determination of the total allergen content in a processed food is an important step, which will provide comparable measurement data of relevance to risk assessors. Graphical abstract.

Ten years of proficiency testing reveals an improvement in the analytical performance of EU National Reference Laboratories for genetically modified food and feed.

National Reference Laboratories (NRLs) in the Member States of the European Union (EU) monitor the implementation of the EU legislation on the presence of genetically modified organisms (GMOs) in food and feed. The EU Reference Laboratory for GM Food and Feed (EURL GMFF) supports the harmonisation of measurement procedures and the improvement of the analytical performance of these laboratories, among others through the organisation of a proficiency testing (PT) scheme. The PT results reported over 10 years have been analysed using common criteria applied to the reported data. The outcome revealed a gradual decrease of the relative standard deviation within the sets of the reported data with time. The extent of the deviation of the results from the assigned value also diminished between 2010 and 2019. The average deviation from the assigned value was independent of the GM content in the later PT rounds but it was affected by the complexity of the test item matrix. Performance scores were calculated for all results reported by the 62 NRLs. The number of unsatisfactory performance scores obtained decreased with time. The trends observed indicate an improvement in the analytical performance and an increased harmonisation of GMO testing within the EU enforcement laboratories.

IMEP-115: determination of methylmercury in seafood by elemental mercury analysis: collaborative study.

A collaborative study IMEP-115 was organized by the European Union Reference Laboratory for Heavy Metals in Feed and Food (EURL-HM) to validate a method for the determination of methylmercury in seafood. The method was based on a liquid-liquid extraction with an organic solvent and with an aqueous cysteine solution. The final quantitation was done with an elemental mercury analyzer. Fifteen laboratories experienced in elemental mercury analyses, from 10 European countries, took part in the exercise. Five test items were selected to cover the concentration range from 0.013 to 5.12 mg/kg. All test items were reference materials certified for the methylmercury mass fraction: DOLT-4 (dogfish liver), TORT-2 (lobster hepatopancreas), SRM 2974a (mussel), SRM 1566b (oyster), and ERM CE-464 (tuna). Participants also received a bottle of ERM CE-463 (tuna) to test their analytical method before starting the collaborative study. Method validation showed adequate accuracy and acceptable precision for all test items, thus fitting its intended analytical purpose. The repeatability RSD ranged from 3.9 to 12.3%, while the reproducibility RSD ranged from 8.4 to 24.8%.

Food allergen analysis: Considerations for establishing a reference measurement system to implement EU legislation.

Inconsistent quantification results obtained from various analytical methods for food allergen testing hamper an accurate quantitative risk assessment and its regulatory implementation. In order to overcome such problems, a concept aiming at ensuring the comparability of quantitative food allergen measurement results is presented here. It is based on an approach called reference measurement system for food allergens, which uses a commonly agreed reference, namely the 'mass fraction of total protein of the allergenic ingredient in food'. The necessary system components are outlined, consisting of a primary reference measurement method, a certified reference material and a reference laboratory. This metrology-based concept can be applied to quantify various food allergens determined with different analytical procedures. The example of 'milk in cookies' is used to demonstrate the approach.

Determination of Alternaria Toxins in Tomato, Wheat, and Sunflower Seeds by SPE and LC-MS/MS-A Method Validation Through a Collaborative Trial.

BACKGROUND: Alternaria toxins are ubiquitous contaminants in highly consumed food products. Therefore, they are candidates to be regulated by EU legislation. In this context, the availability of reliable analytical methods is a keystone both for protecting the health of citizens and smooth functioning of the European market. OBJECTIVE: This paper describes an advanced LC-MS/MS method based on isotope dilution quantification suitable for the determination of altenuene, alternariol, alternariol monomethyl ether, tenuazonic acid, and tentoxin in tomato puree, wheat, and sunflower seeds. METHODS: The method has been validated in an interlaboratory study that included the analysis of both spiked and naturally contaminated food commodities. Twenty-three participants contributed with analytical data. RESULTS: The average recoveries and relative standard deviations for repeatability and reproducibility obtained across the tested matrixes were: 97, 8.0, and 23%, for altenuene, respectively; 95, 9.2, and 17% for alternariol, respectively; 98, 6.4, and 13% for alternariol monomethyl ether, respectively; 97, 4.2, and 9.3% for tenuazonic acid, respectively; and 102, 5.6, and 15% for tentoxin, respectively. The method enabled the determination of all tested Alternaria toxins close to or below 1 µg/kg. CONCLUSION: Overall, the method showed a satisfactory trueness and precision, complying with the requirements for the monitoring of mycotoxins in food in the EU. It is currently under evaluation by the European Committee for Standardization for adoption as a standard method. HIGHLIGHTS: Isotope dilution mass spectrometry method for the determination of Alternaria toxins in food.

Determination of dissolved bromate in drinking water by ion chromatography and post column reaction: interlaboratory study.

A collaborative study, International Evaluation Measurement Programme-25a, was conducted in accordance with international protocols to determine the performance characteristics of an analytical method for the determination of dissolved bromate in drinking water. The method should fulfill the analytical requirements of Council Directive 98/83/EC (referred to in this work as the Drinking Water Directive; DWD). The new draft standard method under investigation is based on ion chromatography followed by post-column reaction and UV detection. The collaborating laboratories used the Draft International Organization for Standardization (ISO)/Draft International Standard (DIS) 11206 document. The existing standard method (ISO 15061:2001) is based on ion chromatography using suppressed conductivity detection, in which a preconcentration step may be required for the determination of bromate concentrations as low as 3 to 5 microg/L. The new method includes a dilution step that reduces the matrix effects, thus allowing the determination of bromate concentrations down to 0.5 microg/L. Furthermore, the method aims to minimize any potential interference of chlorite ions. The collaborative study investigated different types of drinking water, such as soft, hard, and mineral water. Other types of water, such as raw water (untreated), swimming pool water, a blank (named river water), and a bromate standard solution, were included as test samples. All test matrixes except the swimming pool water were spiked with high-purity potassium bromate to obtain bromate concentrations ranging from 1.67 to 10.0 microg/L. Swimming pool water was not spiked, as this water was incurred with bromate. Test samples were dispatched to 17 laboratories from nine different countries. Sixteen participants reported results. The repeatability RSD (RSD(r)) ranged from 1.2 to 4.1%, while the reproducibility RSD (RSDR) ranged from 2.3 to 5.9%. These precision characteristics compare favorably with those of ISO 15601. A thorough comparison of the performance characteristics is presented in this report. All method performance characteristics obtained in the frame of this collaborative study indicate that the draft ISO/DIS 11206 standard method meets the requirements set down by the DWD. It can, therefore, be considered to fit its intended analytical purpose.

Proficiency test on the determination of polyethylene and polybutylene terephthalate cyclic oligomers in a food simulant.

The outcome of a proficiency test (PT) organised by the European Union Reference Laboratory for Food Contact Materials (EURL-FCM) is presented. The PT was set up to assess the analytical performance of National Reference Laboratories (NRLs) and Official Control Laboratories (OCLs) in the determination of mass fractions of polyethylene terephthalate (PET) and polybutylene terephthalate (PBT) cyclic dimers and trimers in the official food simulant D1 containing ethanol and water (50:50 v/v). The EURL-FCM had developed and validated an analytical method based on HPLC-UV to monitor the homogeneity and stability of the target oligomers in the PT test items and to determine the respective assigned values, as prescribed in ISO 17043, 2010ISO 17043, 2010. The standard operating procedure of the method was provided to the participants and could be used instead of their own routine methods. Laboratory results were rated using z, z' and ζ scores in accordance with ISO 13528, 2015. The standard deviation for proficiency assessment, σpt , was set to 20 % of the respective assigned value, for all the four studied oligomers, based on the perception of experts. A total of 36 participants from 26 countries have registered to the exercise. They received two test items. Solution 1 consisted of food simulant D1 fortified with a known mass fraction of the four oligomers, while Solution 2 was obtained by a migration experiment with PET bottles and food simulant D1 and further fortification of the resulting solution with the four oligomers. The majority of the participating laboratories presented satisfactory results for the four PET and PBT oligomers. For the analysis of Solution 1, 79-88 % of the participants obtained |z (or z')-scores| below 2, while the satisfactory performances ranged from 71 to 85 % for Solution 2. This PT has been organised for the first time at EU level for the analysis of polyester oligomers and confirms that most of the NRLs are able to monitor properly these oligomers in the frame of Regulation (EU) No 10/2011.

Measurement uncertainty of test kit results--the ELISA example.

The measurement uncertainty budget for a clinical test-kit--an ELISA technique based on experimental data provided in the instruction manual--is presented. The proper characterization of the calibrators, within-plate homogeneity and the trueness of ELISA test-kits are identified as being major contributors to the combined uncertainty.

Influence of pre-heating of food contact polypropylene cups on its physical structure and on the migration of additives.

Laboratories unexpectedly carried out pre-heating of polypropylene beverage cups prior to performing a migration test in a proficiency test. Principal component analysis of the data collected showed that the preheating temperature of the cups contributed to an increased variance of the data and distinguishing pre-heating and non-pre-heating groups. This triggered to study the effect of applying such pre-heating on the physical structure of the material and on the migration of additives to food simulant D1 (ethanol 50% v/v). Several cups were pre-heated at selected temperatures and either analyzed with differential scanning calorimetry to establish the degree of crystallinity or used for the migration test. Six target additives from Regulation (EU) No 10/2011 were quantified in the food simulant using HPLC-FLD and LC-MS. Results show that pre-heating of the beverage cups led to a significant change in the degree of crystallinity, resulting in a change of analyte migration in comparison to the migration results from non-pre-heated cups.

Can official control laboratories quantify reliably fipronil in eggs? Evidence from a proficiency testing round.

A proficiency testing (PT) round dedicated to assess the competence of European Union (EU) National Reference Laboratories (NRLs), EU Official Control Laboratories (OCLs) and OCLs from associated countries for the determination of the insecticide fipronil in eggs is described. The content of the target analytes (sum of fipronil plus its metabolite fipronil sulfone, expressed as fipronil in the PT test material) was set around the regulated Maximum Residue Level (MRL) set by EU legislation. The PT was organised by the European Commission's (EC) Directorate General (DG) Joint Research Centre (JRC) - in agreement with the EC DG for Health and Food Safety (SANTE) - following a request by the Belgian Authorities. Eighty-six NRLs and OCLs from 22 EU Member States, Norway, Serbia and Albania participated. The test items used were two materials made of frozen liquid eggs, processed and characterised at the JRC facilities in Geel (Belgium). The majority of the participants (94%) were shown to have a satisfactory performance, expressed as z scores, thus demonstrating the analytical capability of most of the participating NRLs and OCLs to enforce the relevant EU Regulations ((EC) 396/2005 and 1127/2014). Furthermore, 93% of the participants who provided a compliance statement classified, correctly, one of the test items as non-compliant.

Application of the CEN/ISO Standard for Phytase Activity Measurements to a New Phytase Product: Determination of a Robust Conversion Factor by an Interlaboratory Study for Screening Feed Samples.

Background: Phytase-based preparations are important feed additives currently authorised in the European Union (EU). The European Standard (EN) and International Organization for Standardization (ISO) standard 30024 describes a harmonized method for the determination of phytase activity and is fit-for-purpose for official control of a group of phytase products. However, it is not suitable for the determination of the phytase activity of a new feed additive encoded as 4a16 in the EU Register of Feed Additives, to which a slightly different phytase activity definition has been attributed. Objective: To establish a robust conversion factor to support official control laboratories that apply the EN ISO method when monitoring feed products containing 4a16. Methods: The phytase activity of test materials was determined by the participants using the EN ISO and/or the "applicant" methods. Results: Robust relative SDs for repeatability and for reproducibility of the methods applied for the determination of the phytase activity in the materials containing the 4a16 feed additive ranged from 2.6 to 22% (EN ISO method) and from 2.4 to 39% (applicant method). Conclusions: The data obtained confirmed the performance characteristics published for other phytase-based feeds in the related standard methods. These results allowed us to estimate a factor of 2.68 to convert phytase activities measured with the EN ISO method into the enzyme activity measured with the applicant method. Highlights: The obtained conversion factor will allow EU official laboratories to screen feed samples supplemented with the 4a16 phytase by applying EN ISO Standard 30024.

A review of the work of the EU Reference Laboratory supporting the authorisation process of feed additives in the EU. [corrected].

This paper describes the operation of the European Union Reference Laboratory for Feed Additives (EURL) and its role in the authorisation procedure of feed additives in the European Union. Feed additives are authorised according to Regulation (EC) No. 1831/2003, which introduced a completely revised authorisation procedure and also established the EURL. The regulations authorising feed additives contain conditions of use such as legal limits of the feed additives, which require the availability of a suitable method of analysis for official control purposes under real world conditions. It is the task of the EURL to evaluate the suitability of analytical methods as proposed by the industry for this purpose. Moreover, the paper shows that one of the major challenges is the huge variety of the methodology applied in feed additive analysis, thus requiring expertise in quite different analytical areas. In order to cope with this challenge, the EURL is supported by a network of national reference laboratories (NRLs) and only the merged knowledge of all NRLs allows for a scientifically sound assessment of the analytical methods.

Proficiency test on the determination of pesticide residues in grapes with multi-residue methods.

This manuscript presents the results of the International Measurement Evaluation Programme 37 (IMEP-37) study, a proficiency test (PT) which was organised to assess the world-wide performance of food control laboratories on the determination of pesticide residues in grapes. This PT supports the implementation of Regulation (EC) No 396/2005 on maximum residue levels of pesticides in or on food and feed of plant and animal origin. Eighty-one participants reported results, forty from EU Member States and forty-one from outside the EU. The test item was a grape sample spiked with 20 selected pesticides. The results of the participants were rated with z- and zeta (ζ-) scores in accordance with ISO 13528 and ISO 17043. The standard deviation for the proficiency assessment, σˆ, of this PT was set at 25% for the 20 measured pesticides based on previous experience with similar measurands. The results reported to IMEP-37 showed that the participants performed satisfactorily, ranging from 81% (carbendazim) to 97% (azoxystrobin, penconazole, pyrimethanil) of the participating laboratories. However, only 30% of the participants managed to analyze all pesticides satisfactorily. Overall, the performance of the participants in this PT was good but there is room for improvement in the development of multi-residue methods for the simultaneous analysis of a large number of pesticides with an increased accuracy.

Feed premix: a difficult matrix for the accurate determination of trace elements - the outcome of IMEP-114 and IMEP-36.

This paper presents the outcome of two proficiency tests (IMEP-114/36) running in parallel, which focused on the determination of total As, Cd, Pb, Hg and Sn in feed premixes. Both exercises aimed to test the performance of laboratories measuring trace elements in feed, in compliance with Directive 2002/32/EC of the European Parliament and of the Council on undesirable substances in animal feed. IMEP-114 was run for the European Union National Reference Laboratories, while IMEP-36 was open to all food control laboratories that wished to participate. In total 80 laboratories from 37 countries registered to both proficiency tests, from which 75 laboratories reported results. The test material used in this exercise was commercially available feed premix, which was processed, bottled, labelled and dispatched to participants. The performance of the participating laboratories was similar in both proficiency tests and was very good in the case of total Cd and less satisfactory for total As and total Pb. For total Sn only one-third of the participants reported results, from which 33% performed satisfactorily. Thirty-four participants reported results for total Hg, although the expert laboratories stated that the mass fraction for that measurand was below the limit of detection of the method used. An evaluation of the reasons that could be at the basis of unsatisfactory results has been carried out and is presented here.