Carrying Out and Reporting Surveys and Results of Studies on the Occurrence of Mycotoxins

Aspects to be considered when carrying out and reporting surveys and results of studies on the occurrence of mycotoxins

Introduction

Because the presence of moulds in foods or commodities is not necessarily indicative of mycotoxins being present, it is necessary to carry out analysis to determine the levels and the incidence of mycotoxins present. Information obtained from surveys and occurrence studies is used to provide a sound scientific basis for a number of subsequent studies, estimations and decisions. It is the main tool to determine the exposure of humans and animals to mycotoxins when considered together with information on dietary intake. Exposure assessment forms part of the process of risk characterization, risk assessment and risk analysis. The ultimate applications of results from surveys and occurrence studies are to protect consumer safety by drawing attention to contamination problems, to provide data for the process of risk management, and to support action to reduce consumer exposure to mycotoxins as necessary. These data can be used to guide authorities on the priorities for control activities and enforcement of mycotoxin regulations and to provide input for the introduction of regulatory limits (1). These data can also help to evaluate the efficiency of agricultural practices and processing in detoxifying or removing mycotoxins, and to develop guidelines for Good Manufacturing Practices that will reduce the overall mycotoxin contamination of food and animal feedingstuffs. This information is also used in epidemiological studies.

Unfortunately there is often insufficient information in reports, abstracts or published articles about surveys, or the analyses are not supported by proper quality assurance. This can result in having to reject findings required for risk assessment or other important evaluations. Quantitative trace analysis requires considerable resources and it is important that surveys and other studies are effectively designed and reported so that maximum cost benefit is obtained.

To this end it is important to follow recommendations about planning and performing this type of study agreed by international experts. Thus maximum benefit and reliable conclusions and recommendations can be obtained from survey data.

Recommended Information Required for Reporting Surveys and Occurrence Studies

Surveys and studies of the occurrence of mycotoxins can address many different situations making the provision of general recommendations a complex task. The Joint FAO/WHO Evaluation Committee on Contaminants and Food Additives (JECFA) developed recommendations which were reflected in the summary and conclusions that followed the 56th session held in Geneva from the 6th to the 15^th of February 2001, where the primary task was the assessment and characterization of risks associated with consumption of foods that may be contaminated by certain mycotoxins. This Committee made a clear statement of the objectives of any study and specifically recommended the following:

a) Surveillance data be accompanied by a clear description of the analytical method used and an indication as to whether the method has been formally validated.

b) Limits of detection and quantification should be provided, with the definitions used to derive them.

c) Recoveries determined by use of spiked samples or reference materials should be given with analytical results, and it should be specified which levels of spiking were applied, and whether the analytical data reported were corrected for recovery.

d) An estimate of the measurement uncertainty should be given, derived from repeatability measurements (RSD) or from calculation.

e) The source of the calibrant(s) should be provided, the procedure used to verify its identity and concentration should be described, and the method of preparation of calibration solutions should be given.

f) It should be indicated whether the laboratory that reported the results was accredited and, if so, for which analyte-matrix combinations.

g) It should be indicated whether the laboratory that reported the results takes part in inter-laboratory comparison studies and, if so, for which analyte-matrix combination".

Sampling

Because of the heterogeneous way mycotoxins are distributed in bulk commodities it is vital that special attention should be devoted to sampling. Hence, when the target mycotoxins and products to be included have been agreed it is important that the samples be collected using a clearly defined sampling plan agreed by all authorities responsible for interpreting and using the information to be obtained.

Results should include information about the number, origin (country, place, etc) and types of samples, the size of the samples, collection dates (e.g. years, months, season, etc), points of sampling, sub-sampling regime as necessary, random or targeted, treatment of samples (e.g. preparation for analysis, storage, mixing) and any other relevant details which can assist in the interpretation of the final results. The number of samples should be large enough to ensure that the incidence and level detected are representative of the real situation and the average content in the item whether it be a food product, animal feedstuffs, milk, alcohol, blood, urine etc.

The FAO/WHO document also highlights the following "most studies of sampling have focused on the development of sampling plans for regulatory purposes, and little work has addressed the need for specific sampling plans for surveillance purposes", thus "further studies on sampling variability are urgently required so that practicable and economically feasible sampling plans can be developed for convenient and accurate determination of the occurrence of mycotoxins". Finally, clear labelling of the food products or items with details, as appropriate, of variety, local name, commercial trademark, etc. is vital in order to ensure that they are correctly identified and classified.

Treatment of data and presentation of results

• The main numerical data that are needed for each category of food product or item that has been analysed are:
• The number of total and positive samples (also in percentage).
• The maximum level detected.
• The range of concentrations detected in the positive samples.
• The overall arithmetic mean value (arithmetic mean value of all samples – both positive and negative samples).
• The mean value of all the positive samples.
• The median value of the positive samples: 50 % percentile (in many studies the statistical parameter that allows a more adjusted valuation of the incidence is the median and not the mean, because the value of the mean is immediately skewed in the presence of isolated results that may appear in the limit of the high concentrations).
• The values for the Limit of Detection/Limit of Quantification (LOD/LOQ).

Given the impractability, in most cases, of presenting all the individual results for all samples analysed, it is important, whenever possible, to include at least the following; the number of samples analysed grouped according to the levels detected and breaking them down into suitable concentration ranges depending on the type of the study. As an example, the following format is proposed by the European Commission for reporting the results of a coordinated programme for the official control of foodstuffs relating to aflatoxins in spices (2):

	Number of samples
Aflatoxin (microgram/kg)	Not detected	< 2	2 - < 10	10 - 50	> 50

The method used to calculate arithmetic means from results obtained is important. Within the European Commission the following guide represents a uniform approach to the calculation of mean values which was adopted by all participants in the tasks for scientific cooperation (SCOOP) relating to the risk assessment of aflatoxins ( 3 ) and assessment of dietary intake of ochratoxin A ( 4 ) :

If an LOD and an LOQ are available, calculate mean level using LOD/2 for results less than the LOD. For results between LOD and LOQ use numerical values if available.

2) If only LOQ is available, use LOQ/2 for values below the LOQ.

Reporting survey results

In view of the considerable resources required to carry out surveys and other studies on the occurrence of mycotoxins it is vital to obtain the maximum benefit by providing comprehensive and well presented data.

Reports, whether they be official results, scientific papers, summaries or abstracts must be informative enough for the reader to make a quick judgement of whether the article provides the information sought and must highlight the main findings of the study. As a minimum a summary should name the mycotoxin, the kind of matrices examined, the number and origin of the samples, the analytical method used and a synopsis of the results to include the number of positive samples, the maximum level, the mean and median values detected, and the main conclusions.

In view of the considerable resources necessary for this type of work it is essential that scientists present the data in a standardized and fully validated manner so that maximum benefit is obtained from the results and correct deductions are made.

References

(1) Scudamore, K., Mycotoxins, an independent assessment of MAFF - funded applied research and surveillance 1993 - 1996. Edited by MAFF - UK. (1999)

(2) European Commission. Commission Recommendation of 8 January 1997 concerning a coordinated programme for the official control of foodstuffs for 1997. Official Journal of the European Communities. Nº L 022 (1997), 27 – 34

(3) European Commission. Assessment of dietary intake of ochratoxin A by the population of EU Member States. Reports on tasks for scientific cooperation. EUR 17523. (1997)

(4) European Commission. Risk assessment of aflatoxins. Reports on tasks for scientific cooperation. EUR 17526. (1997)