New 'virtual' inspection for meat plants proposed
Dr. Richard Raymond, former Nebraska Chief Medical Officer and current USDA Undersecretary for Food Safety, proposed a system of "virtual" meat inspections, with a focus on more frequent inspections at meat plants with poor safety records, according to an article in the Omaha World-Herald (registration required).
Currently the system works like this:Inspectors are assigned to a specific facility or to regularly patrol several smaller plants.
An automated system spits out a schedule of the inspectors' weekly tasks -- ensuring that a plant is clean, checking the temperature of its ovens and refrigerators and reviewing its anti-contamination measures.
If significant problems develop at a particular plant, additional inspections can be done. But on a day-to-day basis, the system aims to create a uniform level of inspection without regard to a facility's history or the safety measures it uses.
Under the proposed risk-based system, plants would receive a risk rating based on the types of products they handle, safety measures they have in place and their track records.
Plants deemed to present a high risk would receive more intensive inspections while those with low risk would receive more cursory examinations.
The new system proposed by Dr. Raymond is controversial, and consumer groups like the Consumer Federation of America are encouraging the USDA to conduct a pilot study of the new inspection system before fully implementing it.
Like so many agency proposals, Risk-Based Inspection (RBI) is science-based and further promotes Public Health, IN THEORY. A close examination of the underlying assumptions which form the foundation of RBI reveal the potential of intentionally-designed shortcomings within RBI.
One of the prerequisites USDA uses to determine risk at individual plants is volume of production. Erudite reasoning! Very small plants are assigned a relative value of 1.0, small plants 1.5, and large plants 2.0. Therefore, a plant killing 5,000 head per day is assigned only twice the risk of a very small plant which might kill 10 head per week. The disparity is obvious, and suggests that the agency wishes to further insulate the large slaughter plants via RBI.
Another prerequisite of RBI is the number of historical enforcement actions individual plants have experienced. The primary enforcement action is the issuance of Noncompliance Reports (NR's). However, NR's are not assigned relative significance which relate to the production of safe food. For example, one NR is written because a plant employee forgot to initial one portion of one daily plant record. This "noncompliance" has zero relationship to the production of safe food, and is merely an inadvertant clerical error posing no danger to public health. Another NR is written because the agency has detected adulterated food at the plant. Although USDA is considering assigning degrees of seriousness to NR's, this common sense classification system is not currently in use, making relative comparisons of the numbers of NR's at different plants meaningless when determining which plants constitute a threat to public health. At one time, noncompliances were rated Critical, Major or Minor. Today, all alleged discrepancies are thrown into one pot with no relative categorization, minimizing the agency's ability to determine which plants truly need increased scrutiny. Simultaneously, new USDA field supervisors oftentimes engage in unilateral demands that inspectors dramatically increase their issuance of NR's, although the plants are not guilty of more noncompliances than in the past. This biased system creates artificial perceptions that individual plants and/or districts have a poorer track record in their attempt to produce safe food.
Another agency building block in RBI is USDA lab results of microbial testing. Most plants do no slaughter now, meaning that their entire production emanates from meat purchased from outside large slaughter plants. The GIGO Effect (gabage in, garbage out) is of major concern to these small, downline, further processing non-slaughter grinding plants. To compound the problem, USDA prohibits inspectors from documenting the origin of meat being sampled at the time of sample collection. Instead, inspectors are to document source verification data only if the sample is determined to be "confirmed positive" at the USDA lab, which requires a minimum of four additional days. This unnecessary delay creates numerous scientific problems which won't be addressed in these comments. Because of the agency's biased sampling protocol, tracebacks to the TRUE ORIGIN of contamination are effectively prevented, with full agency endorsement. Therefore, large slaughter plants are rarely assigned liability for the pathogen-laced meat detected at downline further processing plants. Unfortunately, the agency's implementation of RBI focuses at the downline plants which unwittingly purchase previously-contaminated meat from the large slaughter facilities. When agency testing results reveal contaminated meat at these further processing plants, NR's and other enforcement actions are initiated by the agency, but strictly at the further processing plant, not at the true origin of contamination. Therefore, the agency's desire to utilize previous enforcement actions to justify its RBI goal of reassigning personnel inappropriately concludes that increased scrutiny is necessary at the downline plants, while justifying decreased surveillance at the large slaughter plants where the contamination originated. Although the agency pays lip service to the concept of "Tracebacks" in numerous agency publications, actual agency policies are intentionally designed to prevent tracebacks to the true origin.
Another prerequisite agency building block in RBI is "In Commerce Findings", such as recalls of contaminated meat. This is directly related to the previous discussion of the agency's monolithic unwillingness to traceback to the TRUE SOURCE of contamination. Since May of this year, 8 recalls of E.coli 0157:H7 contaminated meat have occurred. Six of these eight recalls were from plants which do not slaughter! Realizing that E.coli 0157:H7 originates within animal intestines (and adhere to dirty hides), and realizing that non-slaughter further processing plants have no intestines or hides on their premises, how can these downline plants be held guilty for introducing E.coli pathogens? Close conversations with these victimized further processing plants subsequent to agency enforcement actions reveal that USDA consistently assigns full liability to these downline plants for the presence of pathogens, and the agency has no interest in tracing back to the plants where the contaminants were indeed introduced. Therefore, the agency's use of historical "In commerce findings" to determine plant risk for RBI purposes again focuses on the Destination of contaminants, rather than at the Origin of contamination. This plan design virtually guarantees that increased surveillance will be imposed at the wrong plant........with tacit agency approval.
Risk Based Inspection has great potential merit, and would properly enable the agency to reassign personnel to plants which constitute the highest risk. However, USDA definitions of risk criteria are greatly flawed, and requires substantial revision if RBI is to increase surveillance at plants which truly are noncompliant.
John W. Munsell
President, Montana Quality Foods & Processing
Manager, Foundation for Accountability in Regulatory Enforcement (FARE)
Miles City, MT
Before