Extent of beef carcass contamination with Escherichia coli and probabilities of passing US regulatory criteria

In the 1996 U.S. Meat and Poultry Inspection Regulations, Escherichia coli biotype I counts were included as "performance criteria"' of the slaughteri ng process. The criteria were based on a three-class attributes sampling pl an applied in a moving window. The values for m and M and c and n were set at 5 and 100 CFU/cm(2), and 3 and 13 samples, respectively, for beef carcas ses after overnight chilling following slaughter. In this study, beef carca sses were analyzed for counts of E. coli, and the results were expressed ac cording to the above criteria. Furthermore, probabilities of passing E. col i performance criteria were determined. Carcasses were sampled in seven sla ughtering plants (four steer and heifer; three cow and bull), during two se asons, and at three plant locations (pre-evisceration, after final carcass washing, and after 24 h of carcass chilling). Each entire carcass sample (1 00 cm(2) from the brisket, flank, and Nmp) was analyzed individually for E. coli counts. Compared with the regulation, which set the value of m and th e acceptable range based on the 80th percentile of E. coli contamination da ta from U. S. Food Safety and Inspection Service nationwide baseline studie s, our results showed that, on the average and depending on plant and seaso n, 84.2 to 100% of the chilled carcass samples were in the acceptable range . The average percentages of chilled samples in the unacceptable range, set at the 98th percentile, were 0 to 6.7%. Depending on plant and season, the overall probabilities of chilled carcasses passing the regulatory requirem ent were 0.597 to 1.0 (brisket), 0.471 to 1.0 (flank), and 0.485 to 1.0 (ru mp). The results indicated substantial variation among plants and between s easons in ability to meet the E. coli performance criteria.