In January 1999, the Food Safety and Inspection Service (FSIS) finalized performance standards for the cooking and chilling of meat and poultry products in federally inspected establishments. More restrictive chilling (stabilization) requirements were adopted despite the lack of strong evidence of a public health risk posed by industry practices employing the original May 1988 guidelines (U.S. Department of Agriculture FSIS Directive 7110.3). Baseline data led the FSIS to estimate a "worst case" of 10(4) Clostridium perfringens cells per g in raw meat products. The rationale for the FSIS performance standards was based on this estimate and the assumption that the numbers detected in the baseline study were spores that could survive cooking. The assumptions underlying the regulation stimulated work in our laboratory to help address why there have been so few documented outbreaks of C. perfringens illness associated with the consumption of commercially processed cooked meat and poultry products. Our research took into account the numbers of C. perfringens spores in both raw and cooked products. One hundred ninety-seven raw comminuted meat samples were cooked to 73.9 degrees C and analyzed for C. perfringens levels. All but two samples had undetectable levels (<3 spores per g). Two ground pork samples contained 3.3 and 66 spores per g. Research was also conducted to determine the effect of chilling on the outgrowth of C. perfringens spores in cured and uncured turkey. Raw meat blends inoculated with C. perfringens spores, cooked to 73.9 degrees C, and chilled according to current guidelines or under abuse conditions yielded increases of 2.25 and 2.44 log10 CFU/g for uncured turkey chilled for 6 h and an increase of 3.07 log10 CFU/g for cured turkey chilled for 24 h. No growth occurred in cured turkey during a 6-h cooling period. Furthermore, the fate of C. perfringens in cooked cured and uncured turkey held at refrigeration temperatures was investigated. C. perfringens levels decreased by 2.52, 2.54, and 2.75 log10 CFU/g in cured turkey held at 0.6, 4.4, and 10 degrees C, respectively, for 7 days. Finally, 48 production lots of ready-to-eat meat products that had deviated from FSIS guidelines were analyzed for C. perfringens levels. To date, 456 samples have been tested, and all but 25 (ranging from 100 to 710 CFU/g) of the samples contained C. perfringens at levels of <100 CFU/g. These results further support historical food safety data that suggest a very low public health risk associated with C. perfringens in commercially processed ready-to-eat meat and poultry products.

A ground beef patty processor detected Escherichia coli O157:H7 in five production lots during routine testing with polymerase chain reaction (PCR) technology. This finding stimulated research to determine the incidence and potential entry points of the pathogen during processing. One of these lots (53,960 kg) was divided into 71 pallets (760 kg each) of food service ground beef patties. Ten cartons (19 kg each) were removed from each pallet, for a total of 710 cartons. Four patties were taken from each carton and subdivided to provide comparable samples for E. coli O157:H7 analyses by three different laboratories. Two laboratories employed different immunoassay tests, and one used PCR to screen samples. One sample set was analyzed for aerobic plate, coliform, and E coli Biotype I counts to determine if any relationship existed between these microbial groups and the incidence of E. coli O157:H7. For 73 samples, presumptive positive results for E. coli O157:H7 were obtained by one or more methods. For 48 of these 73 samples, positive results for the pathogen were culture confirmed. The largest number (29) of culture-confirmed positive E. coli O157:H7 results were detected by PCR. Most positive results were obtained during a short segment of processing. All culture-confirmed E. coli O157:H7 strains were further characterized by two genetic subtyping techniques, resulting in two to four different patterns, depending on the subtyping procedure employed. For any sample tested, the aerobic plate count was < 3.0 log CFU/g, and coliform and E. coli Biotype I counts were < or = 1.00 log CFU/g. The results of this study suggest that most positive samples were associated with a contaminated batch of raw material introduced just before the 1725- to 1844-h processing segment. These results also indicate that more aggressive sampling plans and genetic screening technologies such as PCR may be used to better detect low levels of E. coli O157:H7 in ground beef products.

Fifteen laboratories took part in a collaborative study to validate a method for enumerating Escherichia coli 0157:H7. The method is based on use of a hydrophobic grid membrane filter and consists of 24 h presumptive enumeration on SD-39 Agar and serological confirmation to yield a confirmed E. coli 0157:H7 count. Six food products were analyzed: pasteurized apple cider, pasteurized 2% milk, cottage cheese, cooked ground pork, raw ground beef, and frozen whole egg. The test method produced significantly higher confirmed count results than did the reference method for milk, pork, and beef. Test method results were numerically higher than but statistically equivalent to reference method results for cheese, cider, and egg. The test method produced lower repeatability and reproducibility values than did the reference method for most food/inoculation level combinations and values very similar to those of the reference method for the remaining combinations. Overall, 94% of presumptive positive isolates from the test method were confirmed serologically as E. coli 0157:H7, and 98% of these were also biochemically typical of E. coli 0157:H7 (completed test). Corresponding rates for the reference method were 69 and 98%, respectively. On the basis of the results of this collaborative study and the precollaborative study that preceded it, it is recommended that this method be adopted official first action for enumeration of E. coli 0157:H7 in meats, poultry, dairy foods, infant formula, liquid eggs, mayonnaise, and apple cider

No abstract

Standardized microbiological sampling and testing procedures were developed that can be used throughout the beef slaughter and processing industry to facilitate the collection and any desired compilation of comparative data. Twenty samples each from carcasses (brisket, flank, and rump areas combined); subprimal cuts (clods); lean trim; and cutting and/or conveyor surfaces were collected in three slaughter and processing operations, with the first operation being a preliminary trial and resulting in no reported data. Microbiological analyses for Clostridium perfringens, Escherichia coli O157:H7, Listeria monocytogenes, Salmonella spp., Staphylococcus aureus, Campylobacter jejuni/coli, total coliforms, E. coli Biotype I, and aerobic mesophilic bacteria (aerobic plate count, APC) were performed on all samples by an outside laboratory. The procedures developed were effective in allowing samples to be collected, shipped, and analyzed in the same manner for all operations. From a logistical standpoint, approximately 20 samples each of carcasses, clods, lean trim, and surfaces could be taken within 4 to 6 h by five people. Forty samples each of carcass, clod, lean trim, and conveyor surfaces from two plants tested negative for E. coli O157:H7, Salmonella spp., and Listeria spp., with the exception of L. monocytogenes being isolated from one carcass and one clod sample. APCs and total coliform counts were between 103 to 105 and 102 to 103 CFU/cm2 or CFU/g, respectively, for the 40 samples each of carcasses, clods, and lean trim. APCs for surface swab counts ranged from ≤ 10 to 103 CFU/cm2.