Assumptions of Acceptance Sampling and the Implications for Lot Contamination: Escherichia coli O157 in Lots of Australian Manufacturing Beef

Abstract

The aims of this work were to determine the distribution and concentration of Escherichia coli O157 in lots of beef destined for grinding (manufacturing beef) that failed to meet Australian requirements for export, to use these data to better understand the performance of sampling plans based on the binomial distribution, and to consider alternative approaches for evaluating sampling plans. For each of five lots from which E. coli O157 had been detected, 900 samples from the external carcass surface were tested. E. coli O157 was not detected in three lots, whereas in two lots E. coli O157 was detected in 2 and 74 samples. For lots in which E. coli O157 was not detected in the present study, the E. coli O157 level was estimated to be <12 cells per 27.2-kg carton. For the most contaminated carton, the total number of E. coli O157 cells was estimated to be 813. In the two lots in which E. coli O157 was detected, the pathogen was detected in 1 of 12 and 2 of 12 cartons. The use of acceptance sampling plans based on a binomial distribution can provide a falsely optimistic view of the value of sampling as a control measure when applied to assessment of E. coli O157 contamination in manufacturing beef. Alternative approaches to understanding sampling plans, which do not assume homogeneous contamination throughout the lot, appear more realistic. These results indicate that despite the application of stringent sampling plans, sampling and testing approaches are inefficient for controlling microbiological quality.

Risk Assessment of Escherichia coli O157 Illness from Consumption of Hamburgers in the United States Made from Australian Manufacturing Beef

“…However, for the purpose of this risk assessment it was assumed that contaminated lots were not removed from the export supply chain due to N-60 testing at the processing plant and were shipped to the United States for grinding into patties. In addition, it was assumed that the contaminated lots that were detected during routine export certification testing (16) are representative of all other contaminated lots.…”

“…Previous work indicated that only a small fraction (0.5%) of Australian lots are likely contaminated with E. coli O157. (16) Hence, the efficiency of simulation could be improved by only simulating contaminated lots. This was done because uncontaminated lots do not add to the risk (constant risk that equals zero); however, the fraction of contaminated lots is taken into account for the calculation of final risk estimates.…”

“…The distribution and concentration (cfu/cm 2 ) of E. coli O157 in five lots of manufacturing beef that had failed to meet Australian requirements for export to the United States have previously been estimated. (16) There were four lots in which one carton was contaminated and one lot that had two contaminated cartons out of the 12 cartons that were sampled. However, due to the small number of cartons sampled per lot the subsequent carton prevalence estimates are uncertain.…”

“…Recently, the opportunity arose to estimate this risk, and also to estimate it using a novel approach based on the prevalence and concentration of E. coli O157 in lots of beef that had been withdrawn from the export chain following the detection of the pathogen by the N-60 testing regime used in testing of all export lots destined for grinding in the United States. (16) The present article reports the results of a risk assessment of E. coli O157 from the consumption of beef burgers made only from lots of Australian manufacturing beef. Starting with the prevalence and concentration of E. coli O157 in lots of Australian manufacturing beef, contamination was modeled from grinding and patty forming through to cooking and consumption.…”

Performance standards and meat safety — Developments and direction

“…Investigation of Australian manufacturing beef in which E. coli O157 had been detected (and removed from the food chain) indicates that, when E. coli O157 is detected it is not widespread through the lot and is at low concentration (Kiermeier, Mellor, Barlow, & Jenson, 2011). From these data, one would suspect a very low risk of consumers being affected.…”