DETERMINATION OF THE PERFORMANCE ATTRIBUTES OF POSTMORTEM MEAT INSPECTION PROCEDURES

Traditional codes of post-mortem meat inspection are often inappropriate to the spectrum and prevalence of diseases and defects present in a particular class of livestock slaughtered in a specific geographical region. When a change in inspection procedures is proposed, a risk assessment model can provide a scientific basis for evaluation and generate important information for risk management decisions. A risk assessment of proposed new inspection procedures requires that all the grossly identifiable hazards that could be present in the tissues of interest be identified, and a trial be conducted to compare the performance characteristics of the inspection procedures. A confidence interval for the difference between the non-detection rates per N tissues that pass inspection provides the most meaningful comparison. A decision on the acceptability of a proposed new procedure is based on the worst case included in the confidence interval. Full matching and negative matching trial designs can be used, and each requires different statistical methods to determine confidence intervals for non-detection rates. Political and economic considerations may be involved in the determination of a level of acceptable risk, but determination of performance characteristics is independent of such considerations.