Part 3 in our series about food fraud testing
Part 1: Food testing checklist
Part 2: How to design a food testing plan
Guidance on test frequencies:
When developing a food fraud testing plan, one of the biggest challenges is choosing the frequency of testing. Food Fraud Advisors has developed these guidelines to use as a starting point. They are not a definitive guide, always follow the advice of your analytical laboratory and specialists.
For any given material, the frequency of testing should be based on its susceptibility to food fraud. Suggestions for five different types of materials (A, B, C, D) are described below. Test frequencies should be reviewed and adjusted on a regular basis. Frequencies can be decreased if results are consistently acceptable and increased if problems have occurred or if there are changes to the product or the supply chain.
Type A: a material known to be very susceptible to food fraud (eg. honey, white fish fillets) but for which no specific incidences have occurred within your supply chain and for which no intelligence related to your supply chain has been received.
Phase 1: test a sample from 5 out of 5 incoming batches from every supplier and/or geographical source. For agricultural materials, this sampling protocol should be repeated on a seasonal basis across the year. The aim is to build a picture of authenticity across the year for your suppliers. Progress to phase 2 if all the results from phase 1 have been satisfactory. If test results have revealed problems with authenticity, phase 1 should continue for the affected suppliers or places of origin. Suppliers with satisfactory results can progress to phase 2.
Phase 2: reduce the frequency of testing to one sample per 5 to 10 incoming batches, with the frequency based on your estimate of the ongoing risk. As in phase 1, materials with seasonal supply fluctuations should be tested over various seasons. Use the results from phase 2 to continue to add to the overall picture of authenticity for this material type within your supply chain. Continue until you have at least 10 results. Progress to phase 3 if all results are satisfactory.
Phase 3: is an ongoing maintenance phase. Aim to test one sample per 10 to 20 incoming batches as an ongoing protocol.
Type B: Material for which you have received specific intelligence related to authenticity that pertains to your supply chain.
Phase 1: focussing on the implicated products or ingredients, take samples from at least 5 incoming batches. If the material is non-homogenous – for example if it is in individual packages, or individual pieces rather than being a bulk liquid or powder – consider taking duplicate or triplicate samples from each incoming batch, if budget allows. If all of these samples ‘pass’ the tests, consider trying alternative test methods or test types. Move to phase 2 if all test results are satisfactory.
Phase 2: reduce the frequency of testing to one sample per 5 incoming batches. Continue this frequency of testing until your intelligence information and your test results are sufficient to suggest that the risks within your supply chain have passed. If the risk has abated, reduce the frequency to that of phase 3 for type A.
Type C: Moderately susceptible materials
Phase 1: test a sample from 1 of 5 to 10 incoming batches from every supplier and/or geographical source. For agricultural materials, this sampling protocol should be repeated on a seasonal basis across the year. The aim is to build a picture of authenticity across the year. Continue until you have at least 5 results per supplier or geographical source. Progress to phase 2 if all the results from phase 1 have been satisfactory. If test results have revealed problems with authenticity, phase 1 should continue for the affected suppliers or places of origin. Suppliers with satisfactory results can progress to phase 2.
Phase 2: test one sample per 10 to 20 incoming batches with the frequency based on your estimate of risk. As in phase 1, materials with seasonal supply fluctuations should be tested over various seasons. Use the results from phase 2 to continue to add to the overall picture of authenticity for this material type. Continue until you have 5 results. Progress to phase 3 if all results are satisfactory.
Phase 3: If all results from phases 1 and 2 are satisfactory, the frequency of testing can be significantly reduced. One test per year per supplier should be adequate for most materials and suppliers.
Type D: Food products at risk of counterfeiting or suspected of being from illegitimate sources
A market surveillance sampling plan should be devised. Retail and/or wholesale sources in high risk markets should be targeted in addition to online marketplaces. Anonymous market ‘buyers’ could be engaged to obtain the samples. Buyers can be instructed to purchase a specific number of items per batch per outlet, or they can be instructed to inspect or photograph samples in situ. The amount of surveillance will depend on the brand owner’s tolerance of counterfeiting and the testing budget. Products at high risk of counterfeiting are likely to need ongoing surveillance. Food products that have been found to have entered marketplaces through theft, diversion or grey market channels can have their risk reduced by implementing controls within distribution chains.
Type E: low risk materials. If a food ingredient or product has been shown to be low risk in a food fraud vulnerability assessment then authenticity testing should not be required. If the material or supply chain has changed such that the material requires authenticity testing then its food fraud vulnerability should be re-classified to medium or high risk.