Click here to get the PDF

In late 2005, officials at the USDA’s Food Safety and Inspection Service told consumer advocates that the agency was initiating changes to the way meat and poultry products are inspected. Referred to as “risk based inspection,” the agency’s plan is supposed to re-deploy its food safety resources (7600 inspectors who are stationed in meat and poultry slaughter and processing facilities) according to the food safety risk posed by each individual plant. 

On its face, the plan seems reasonable; thinking about the best way to use resources seems like a responsible thing to do. But when considered in context, the proposed risk-based inspection system represents a profound reduction in food safety protection for consumers – even as meat and poultry products continue to be stamped “USDA Inspected and Passed.” 

There are numerous problems with USDA’s plan to implement risk-based inspection, ranging from the practical – a lack of data necessary to calculate the risk that is supposed to be driving the assignment of inspection – to concerns about the legal basis for making such a dramatic change and a failure to adequately involve all the parties affected by this change – especially consumer groups and meat inspectors themselves.

Not Enough Data to Calculate Risk

The agency does not have enough data to properly calculate the risk posed by a particular product or plant. The plan for risk based inspection calls for several pools of data to be used to calculate risk, including results from microbial testing in plants, plants’ record of past violations of food safety rules, reviews of plants’ food safety plans, and reports of consumer illnesses attributed to specific plants. Unfortunately, none of these pools of data is complete or accurate enough to give a realistic picture of how plants are performing.

Microbial Testing

A major source of data the agency says it will use to calculate risk is generated by the agency’s microbial testing program. The agency conducts a series of tests under its Salmonella program in processing establishments that produce ground beef, chicken or turkey; E. coli O157:H7 tests in establishments that produce ground beef; and Listeria sampling in establishments that make ready-to-eat products.

Some processing establishments are subject to no agency microbial testing at all. According to the agency, approximately a quarter to a third of establishments will not be tested for any pathogen, and consequently the agency will have no information about their pathogen control record. 

But even at establishments that are tested, the testing occurs only sporadically and the most recent results in the database may not accurately reflect the current conditions in the plant. With infrequent testing, a public health threat could exist at a plant for quite some time, while the agency is still operating on the assumption that conditions have not deteriorated because previous testing was acceptable.

The agency conducts Salmonella compliance sets in plants approximately once per year.  A September 2006 USDA Inspector General report clearly illustrated that the Agency has an incomplete database for its Salmonella testing program, identifying as many as 865 establishments nationwide that have no testing data for Salmonella, a possible undersampling rate of 58 percent. The agency has recently increased its E. coli O157:H7 testing at ground beef plants, but still does not average one test a month.

Past Performance

The agency has said it will use data it already has in its computer system on individual plants’ compliance with agency regulations to calculate the risk for each plant, including instances where USDA inspectors saw violations and documented them with forms called noncompliance records (NRs). But there are numerous gaps in the data concerning plant performance, because many violations are either not witnessed by government inspectors, or even when they are witnessed, are not documented. 

Many plants are part of “patrol” assignments – a group of plants that will be covered by one inspector who travels between them during the shift. The majority of these patrols include more than two establishments so inspectors will be there for less than half of the production day. Obviously, any violations that occur during the inspector’s absence are not recorded.  Additionally, even violations that are observed may not be recorded because of the time it takes to accurately file a report. 

Food Safety Plans

The agency also plans to use periodic reviews of plants’ food safety plans as part of the calculation of a plant’s risk. But these reviews happen only every three years, on average, and don’t include plans for all of the different products a plant makes, which can vary widely in terms of what food safety risks are present.

Consumer Illness

Another source of data for calculating the risk posed by a particular plant should be records of illnesses caused by products produced there. Yet, despite repeated questioning by consumer groups at a public meeting on the topic, the agency could not articulate a plan for including data on reported cases of food-borne illness into the data pool that will be used to calculate a meat plant’s risk profile.  

Shaky Legal Foundations

The agency’s plan to shift to a risk-based inspection system is being done without the involvement of Congress, and seems to ignore several of the key mandates of the laws governing meat and poultry inspection.

“Continuous” Inspection

A change to risk-based inspection does not meet the standard of “continuous” government inspection set by the Federal Meat Inspection Act and the Poultry Products Inspection Act, the federal laws that established the meat inspection system, and which are still in effect.  

The admission that slaughter operations could be subject to risk based inspection makes the legal basis for this change even more questionable, since the laws establishing meat and poultry inspection call for continuous government inspection of carcasses and birds.

Wholesomeness

Risk based inspection ignores a host of issues about the “wholesomeness” of meat and poultry products. The USDA considers defects like tumors, vomit, feathers, bruises, cuts, and other damage as a separate category from bacterial contamination problems that it classifies as “food safety” violations. The calculations of risk that will drive the assignment of inspectors under RBI will not include these wholesomeness concerns, potentially leaving consumers unprotected from these important conditions.

Putting Politics Ahead of Consumer Safety

The agency is rushing through a process that would radically reduce consumer protection. 

Contrary to agency promises of a systematic process for implementing a risk based inspection system, there seems to be a sudden sense of urgency to put changes in place. We have heard repeatedly from both industry and the Under Secretary for Food Safety that changes have to be made before the next presidential election in 2008. In September, the USDA’s Under Secretary for Food Safety was quoted in the media as saying that he wanted to roll out the new inspection program by the first quarter of 2007. 

Already, the agency has told the media that they are considering allowing “virtual inspection” of some meat plants – allowing companies to email records so that agency personnel can examine them without ever coming to the plant. This is a dramatic departure from the legal requirement of “continuous” government inspection, and effectively removes the protection offered to consumers by having USDA personnel on-site while meat and poultry are produced.

Budget

Far from a minor adjustment intended to maximize food safety, this plan is really being used as a way to reduce the USDA’s budget. The changes in the way inspectors are assigned to meat and poultry plants would make current inspector shortages permanent, effectively shrinking the size of the agency’s frontline inspection workforce.

For several years, the agency has experienced chronic inspector shortages around the country. During the summer of 2006, agency records revealed significant vacancies -- 9 percent for the Jackson (MS) District, 10 percent for the Atlanta District, 11 percent for the Raleigh District, and 13 percent for the Denver District. One agency official recently admitted that the New York City area typically has a 25 percent vacancy rate.  

When one inspection position is vacant, other inspectors have to assume responsibility for those plants, resulting in inspectors being “doubled” or “tripled up". This summer we learned of one inspector in the Albany District who was covering 18 plants for many weeks. Recently we learned of an inspector in the Philadelphia District who has been covering 26 plants for several weeks, at least. Obviously, when inspectors are doubled or tripled up they don’t have enough time even to visit all of the plants they are to cover that day.

Transparency

The agency is not including the “resources” it hopes to re-deploy – its own employees – in this process. USDA has conducted “employee feedback sessions” that used questionable methodology, held employee “town hall” meetings at times when most inspection personnel were working and could not participate, and provided very limited opportunity for employees to participate in the lone public meeting on this issue.

To make matters worse, the agency has not been truthful in its public statements about this program.  

Agency officials continually assert that every processing plant is visited once per shift by an FSIS inspector. 

Yet, numerous processing plants are not inspected daily and some operate for days at a time without a visit by an inspector, because of chronic vacancies in the inspectors ranks. This is confirmed by the Agency’s own documents and comments made by inspectors during the “employee feedback sessions” conducted earlier this year.

Agency management officials also assert that non-compliance reports (NRs) are written for every regulatory violation; therefore NRs represent a good source of data for calculating a plant’s risk. 

But the actual number of NRs written are understated due to inspector shortages, which result in inspectors not being present in a plant to witness the violation or not having time to document it. Additionally, many inspectors report being told by their supervisors not to write NRs, but to deal with a violation verbally. 

Agency officials initially told consumer advocates that the shift to risk-based inspection would only happen in meat and poultry processing plants, not slaughter plants.

At an October public meeting on this topic, agency officials confirmed that the changes brought about by a risk based inspection system would not be isolated to certain segments of meat production (such as processing), but that all aspects of meat regulation might be subject to risk based inspection, including inspection of slaughter procedures and pre-operational sanitation. 

Conclusion

Taken together, these flaws make risk based inspection something that consumers cannot afford.  Far from being a sound use of science, or an efficient use of limited resources, risk based inspection means one thing for consumers – more risk and less protection.

 

For more information see:

• Food & Water Watch’s comment to the USDA on their proposed Risk Based Inspection plan
  
• Foul Fowl:  An Analysis of Salmonella Contamination in Broiler Chickens