Food Protection Trends - July 2009 - 429

released, the fleece is pulled downwards, over the rump. The sequence of slaughter and dressing procedures is fully described in process flow diagram form by Bell and Hathaway (3). The industry collects systematic data on Escherichia coli and Salmonella at export establishments, which are collated via the E. coli and Salmonella Monitoring (ESAM) program by the Australian Quarantine and Inspection Service (AQIS). Typically, the ESAM database accrues around 13,000 test results for ovines each year; because the program has been in operation since 1998, a great many test data are available. The database has proven useful for setting performance standards for each livestock category, including ovines (17). The database can also provide individual establishments with temporal summaries, together with an industry average over the same period. For those companies close to, or “better” than, the industry average, the data provide confidence in their processing system. By contrast, those that have summary data “worse” than the industry average may find it difficult to explain their apparent inability to meet industry norms. To assist establishments in evaluating their process, a software tool has been developed similar to one that was developed to assist beef processors (11). The modus operandi quantifies a “Problem Score” associated with the live animal and then combines this with a “Process Score”, by assessing unit operations in the slaughter plant. This article describes the scientific underpinning for the tool and its validation against microbial populations on carcasses in six processing plants. MATERIALS AND METHODS The tool consists of a series of questions to be answered by the user. It was initially developed as a Microsoft Excel spreadsheet and later turned into a standalone application based on the Microsoft .NET framework. The Problem Score is calculated by assigning a score of 0 for a “good” practice and a score of 1 for a “bad” practice. An exception relates to the proportion of animals slaughtered with short wool (< 5 cm), for which the score is based on the proportion of animals processed with short fleece. The Problem Score is calculated by summing the individual problem question scores, and the higher the Problem Score, the larger the likely incoming problem. For questions concerning the process, a score of 0 is given for each “good” practice and a score of 1 for each “bad” practice; some questions contain more than two options, in which cases additional possible scores of 2, 3, etc. can accrue. The Process Score is calculated by summing all the individual process question scores. Again, the higher the Process Score, the less likely the processing operations will be able to cope with a serious incoming problem. Practices have not been weighted proportionally, since any constant of proportionality would likely depend on what other practices are also in use. For each question, the possible answers and associated score are given, along with reasons for the choice. These questions have been organized for this manuscript in seven roughly related groups; the tool asks questions in an order that largely reflects the order of operations during slaughter and dressing. It should be emphasized that, while the questions and their order relate primarily to inverted dressing, which is the method by which the majority of sheep are processed in Australia, the tool can also be used to inform on conventional dressing by selecting the “not applicable” option where appropriate. In order to develop appropriate questions, six sheep abattoirs were visited twice by the authors and observations made on unit operations which, while basically common to all plants, show substantial between-plant differences. The abattoirs were selected (non-randomly) because of their participation in the third national baseline (16), the differences in line speed and dressing technology and for sourcing stock from the same geographic region. The abattoirs produced between 2,000 and 8,000 animals per day. Quality Assurance and Production Managers were present during these visits and provided information about the process. An iterative process then followed, in which drafts of the spreadsheet were presented for comment to quality assurance staff of at least twenty sheep processing abattoirs. Based on their comments and suggestions, the tool was modified to its present form. While the tool was being developed, the third national baseline study (16) was in progress, enabling up-to-date information on carcass microbiology to be gathered for each of the six plants selected for the study. The database from the baseline study was interrogated to generate prevalence and concentration of indicator organisms (aerobic plate count (APC) and Escherichia coli) from carcass sponges taken at each plant on at least two occasions. Problem Score incoming livestock questions Since 1880, when sheep meat was first exported aboard the SS Strathleven from Australia to the United Kingdom (1), New Zealand and Australia have supplied virtually the entire sheep export market. This may explain the fact that most published work on microbiology of sheep slaughter and dressing emanates from these countries, beginning with Grau and Smith (10) and Nottingham (14), who described sources of carcass contamination. More recently, Biss and Hathaway (5) and Bell and Hathaway (3) have provided microbiological assessment of the two current dressing systems used by the industries of Australia and New Zealand: conventional and inverted. The publications of these authors form the basis for much of the scientific underpinning of the elements of the software tool described below. The pelt is considered the primary source of microbial contamination transferred to the carcass during slaughter and dressing. Length of fleece, whether it is wet/dry, and its degree of contamination, particularly with feces/soil (termed “tag”), are major influences on carcass microbiology (3, 5). The foregoing forms the basis for four questions that estimate the Problem Score — contamination likely to enter the plant on livestock. The first question is: What percentage of sheep slaughtered have short (< 5 cm) wool? Animals with long wool are harder to dress because of ‘roll-back’ of wool (where the fleece has the potential to flap onto the freshly exposed carcass surface) and a greater chance of contaminating knives and workers’ hands, arms, aprons etc. (3). In the tool, the percentage, p, which is entered without decimals, is JULY 2009 | FOOD PROTECTION TRENDS 429

Food Protection Trends - July 2009

Table of Contents for the Digital Edition of Food Protection Trends - July 2009