ABSTRACT
Inoculated beef trim containing a cocktail of green fluorescent protein-marked Escherichia coli biotype I cultures as surrogates for E. coli O157:H7 was introduced into two large, commercial grinding facilities capable of producing 180,000 kg of ground product in 1 day. Three repetitions were performed over 3 days. Sampling occurred at three different points within the process: postprimary grind, postsecondary grind-blender, and postpackaging. Resulting data show that, as the inoculated meat passes through the system, the presence of the marked surrogate quickly diminishes. The depletion rates are directly related to the amount of product in kilograms (represented by time) that has passed through the system, but these rates vary with each step of the process. The primary grinder appears to rid itself of the contaminant the most quickly; in all repetitions, the contaminant was not detected within 5 min of introduction of the contaminated combo bin into the system, which in all cases, was prior to the introduction of a second combo bin and within 1,800 kg of product. After the blending step and subsequent secondary grinding, the contaminant was detected in product produced from both the parent combo and the combo bin added directly after the parent combo bin; however, for those days on which three combo bins (approximately 2,700 kg) were available for sampling, the contaminant was not detected from product representing the third combo bin. Similarly, at the packaging step, the contaminant was detected in the product produced by both the parent and second combo bins; however, on those days when a third combo bin was available for sampling (repetitions 2 and 3), the contaminant was not detected from product produced from the third combo bin.