To avoid excessive cost, in larger herds DEFRA has indicated its acceptance of sampling a statistically derived proportion of the herd. The number of animals tested is based on the inputs discussed above: individual test sensitivity, individual test specificity, expected prevalence in (known) diseased herds, and the required herd-level sensitivity and specificity of the testing strategy.
Not sampling all animals in a herd involves accepting an increase in risk that an infection may be missed relative to testing the whole herd: where targeted sampling occurs, the owner will be required to sign to indicate acceptance of this risk.
Given the inherent difficulties in predicting where bovine tuberculosis might occur in a herd, there are no specific guidelines on how to determine which animals to test in cases where whole-herd sampling is not performed. General advice is that the sampling of animals should be designed to target cohorts in the herd by age and group. That is, samples should be taken from all distinct groups of animals and within each group should reflect the age structure of that group.
Finally, because this is a voluntary scheme, there is no compulsion on the owner to present specific groups for testing. It is ultimately in owners’ best interests to ensure that freedom from disease is tested for as thoroughly as possible.