Food Safety Magazine, April/May 2009

would, ideally, lead rapidly to other farms, potentially affected lots and back up the supply chain to all receivers. With rare exceptions, this may be a significant challenge today.

In contrast to the desire to execute such a rapid identification of common inputs or proximity, a frequently expressed concern is that such rapid traceability data-sharing technologies will result in the arbitrary and capricious removal of all products even remotely or tangentially connected with an implicated lot. Due to the often highly perishable nature of produce, it is likely that insufficient time would be available to determine whether the concern about co-contamination is substantiated.

Industry Needs for Traceability
Systems

Traceability systems may be designed to include varying levels and approaches to improving breadth, depth and precision. A longstanding function of traceability has been to differentiate one

supplier from another by providing their buyers and prospective buyers the assurances of a detailed and well-imple-mented program. Exactly how much breadth, the degree of information regarding the product’s production attributes, is needed in a particular timeframe is a matter of debate. In general, the consensus is that the first flood of information should be well controlled and limited by design to surgically isolate one potential source from all others. In the heat of a crisis, such as an outbreak investigation, it is easy to overload the investigators with too much information. Delving into additional layers, such as a common seed lot or specific harvest equipment, would only come later, once the legs of the trace-back have been confirmed to a reasonable degree.

Depth of traceability is related to the logic of requiring one step back, one step forward at each point in the supply chain; this point is illustrated by the ability of the FDA, for example, to laboriously walk back, one leg at a time, along a marketing channel for peppers

that led to specific farms in Mexico that were reported to have supplied the original samples with a genetic match to the case isolates of S. Saintpaul. ¹¹Additional genetically matched strains were obtained from environmental samples and pepper samples at a regionally related farm. In some long-existing traceability systems as described above, the carton-associated lot codes would have allowed that step to occur in one giant leap, right back to the farm block. The availability of individual consumer units in domestic refrigerators, though not generally the case, was instrumental in facilitating the rapid trace-back of implicated bagged spinach to a processor and a narrow group of farm sources in 2006.¹⁰One objective of the PTI would be to bring that degree of depth identification transfer of implicated product to all sectors. A significant but manageable barrier to a one-step concept are those commodities, such as mature green tomatoes, that may be commingled at various points from initial packing to repack operations.