Compliance with legal requirements: definition of limits, sampling and measurement uncertainty

Abstract 

In many cases compositional requirements for foodstuffs (e.g. limits for the fat, protein, dry matter, or water content) are established by legislation. Adequate compliance testing is possible only if limits are clearly defined, taking measurement and sampling uncertainty into consideration. Furthermore, decisions on compliance must be based on samples which reflect the composition of the quantity to be evaluated. The resulting sample sizes are normally regarded by food inspection authorities as being much larger than acceptable. Consequently, an alternative strategy should be developed. Autocontrol data (i.e. inspection results obtained by the factory) in principle provide an adequate data basis for decisions on compliance. However, they must be reliable and the food inspection authority must have access to these data on request. Using these data and on condition that they show an approximate normal distribution, an inspection strategy based on arithmetic mean and standard deviation can be developed. Reliable and transparent decisions on compliance can thus be made. In many cases an adequate verification of food authenticity requires a comparison of raw material and product composition. Maximum acceptable differences, taking the relevant sources of variation into consideration, have to be defined and should be used instead of limits.