Traceability & Control post Elliott Review
After a delay for the summer parliamentary recess the Elliott Review Final Report has now been published and Professor Elliott will present his findings at the Association of Public Analysts annual meeting this month. A Zero tolerance of food crime is one of eight integrated recommendations of the report, which also highlights the important role of analysis. For example, the Government should be encouraging industry to conduct sampling, testing and supervision of food supplies at all stages of the supply chain. Within the Laboratory Services recommendation the need to have standardised, validated approaches is also stated. Additionally, the distinction between sampling and testing is made within the Audit recommendation and a call for product sampling to be conducted by auditors even if samples are only retained and not analysed. The Government response and discussion of these topics has just begun; meanwhile, the Food Standards Agency has a proposal under consideration for a threshold for meat species contamination separate to its established 1% Action Limit for threats. When a content of undeclared meat species is between the equivalent of 0.1% and 1.0% (w/w) it is proposed that the causes should be investigated and corrected by the producer. In a recent study published in the Journal of the Association of Public Analysts the Limit of Detection (LOD) of three methods used by official control laboratories during the 2013 Horsemeat surveys was verified. All three methods were based on PCR and were able to detect 0.1% w/w raw horse meat in raw beef meat. But PCR is not the only choice of method. During this time, industry and official laboratories also used microplate ELISA among their testing methods. In our laboratories, when analysing raw meat samples the estimated LOD of microplate ELISA (using a standardised Cut-Off factor of 2.5) is significantly below 0.1% (w/w). In order to verify equivalence between these two different approaches a direct comparison under standardised conditions – preferably using Reference Materials at an appropriate level – would help to provide assurance that the probability of detecting a species is the same in both.