MEMO TO USDA: You think you are tough on food safety in poultry. But too often your policies are just punitive and counterproductive. And, by the way, the so-called science behind many of your policies is not very scientific. Those are the lessons, in fact, from the human illness outbreaks of Salmonella Heidelberg and Cargill’s ground turkey recalls in 2011, the first of which occurred one year ago this month (August).

Who says so? The poultry industry that you regulate, Mr. Food Safety and Inspection Service Administrator.

OK ... you might say the poultry industry is biased in those assessments and doesn’t like regulation. Fair enough, but that doesn’t necessarily mean their criticisms of the food safety system are wrong. Let’s be fair; neither the poultry industry nor the Food Safety and Inspection Service should get by on self-grading without scrutiny.

What the salmonellosis outbreaks and product recalls reveal  

The outbreaks of human illnesses from Salmonella Heidelberg beginning in February of 2011, Cargill’s recall of 36 million pounds of ground turkey on August 3 and a second recall by the company on September 11 of 185,000 pounds provide lessons about the performance of the food safety system. The disease outbreaks and recalls, in fact, reveal fault lines in USDA policy and its real-world implementation. They include the following:


  • Pathogen performance standards as established by the Food Safety and Inspection Service are neither accurate predictors of the ultimate food safety of poultry products nor dependable indicators of the success of a plant’s process control and should not be relied upon for those purposes.
  • True, scientifically sound, Hazard Analysis Critical Control Point systems are not the same thing as regulatory HACCP as established in USDA policy. The absence of a requirement for pathogen kill steps in regulatory HACCP makes it an unreliable blueprint for predictable food safety outcomes.
  • While the Food Safety and Inspection Service focuses on the level of incidence of foodborne pathogens in its regulatory programs, a truer, more realistic indication of food safety would be the number of foodborne pathogen cells.
  • Attribution data for human illnesses from foodborne pathogens as developed by CDC and state health authorities are open to interpretation in establishing associations with categories of food and are unreliable for tracing human illnesses to specific sources of food.


Given this lack of scientific coherence, it is no wonder that controversy exists about the USDA’s food safety policies. A legitimate question to consider is whether these policies, on balance, promote or detract from food safety.

Something wicked this way comes  

The 2011 outbreaks of human illness due to Salmonella Heidelberg and Cargill’s product recalls are also a case study in how food safety might not be going as well as a poultry processor may think at any given time. It further shows that information collected by CDC and USDA wasn’t shared as early as it should have been to reduce the risk to public health.
Mike Robach, vice president, corporate food safety and regulatory affairs, Cargill, presented information about the disease outbreaks and the recalls at U.S. Poultry and Egg Association’s 2012 Poultry Processor Workshop. He raised issues at the workshop involving the Food Safety and Inspection Service's conduct of the case.

In the dark about links to Salmonella outbreaks  

What you don’t know can hurt you in the food safety arena. Cargill was meeting Food Safety and Inspection Service food safety performance standards at its Springdale, Ark., plant, according to Robach, when the company received a call from the service July 15 requesting information about its distribution of turkey to a Cincinnati retail distribution center. Unknown to Cargill at the time, human illnesses due to Salmonella Heidelberg had begun in February 2011, and had been tracked and investigated by CDC and state health authorities for five months at that point. The company also was not told, then, that its ground turkey was suspected by the Food Safety and Inspection Service of being linked to human illnesses.

Also unknown to Cargill at the time was that retail samples of ground turkey originating from its Springdale plant had tested positive in April and in May for the Salmonella Heidelberg outbreak. The samples were taken in routine monitoring by the National Antimicrobial Resistance Monitoring System, which is run jointly by CDC, FDA and USDA.

Referring to the service's request for distribution information, Robach said, “On July 15, we figured something was up, but they would not give us any information,” he told listeners at the workshop.

Cargill investigates and tweaks food safety interventions  

Cargill wouldn’t hear from the Food Safety and Inspection Service again for another two weeks, but the company began examining the food safety data for its Springdale plant. The plant was meeting Food Safety and Inspection Service food safety performance standards, but there had been some episodic peaks of Salmonella in 85 percent lean ground turkey produced by a mechanical (bone-in) deboning process. It was also found that a breeder flock in late 2010 had tested positive for Salmonella Heidelberg (not an unusual serotype to occur in turkeys). While the analysis indicated the Springdale plant was in compliance with Food Safety and Inspection Service food safety standards, some tweaks were made to the plant’s interventions, according to Robach.

Epidemiology revealed to Cargill  

On Friday, July 29, the USDA issued a public health advisory warning consumers not to consume ground turkey. But it wasn’t until August 1 that the epidemiology behind the advisory was revealed to Cargill, according to Robach, and the revelation came from state health authorities, not the USDA.

“On Monday, August 1, we got a call from a state public health epidemiologist who shared the epidemiological information that the public health authorities had,” Robach said. “At the time, the information had only been shared among the public health community and [the Food Safety and Inspection Service].”

One of the things that the epidemiology revealed to Cargill was that the human illnesses were due to Salmonella Heidelberg, the same serotype that had been found earlier in its breeder flock.

Cargill recalls 36 million pounds of ground turkey  

Cargill acted on the information August 3, 2011, voluntarily recalling 36 million pounds of 85 percent lean ground turkey produced in its Springdale mechanical deboning process from February through August 3, 2011. The company also suspended use of the plant’s grinding operation.


The Food Safety and Inspection Service responded by sending a Food Safety Assessment team to the Springdale plant and issued a Notice of Intended Enforcement Action. Cargill responded with a corrective action plan, which was accepted by the service.

Corrective actions before August 10 restart  

Following are the food safety program changes implemented before production of the grinding operation was resumed August 10, 2011:


  • Strengthened control of chiller intervention (peroxyacetic acid) and made it a Critical Control Point
  • Modified Salmonella control program
  • Increased whole-bird sponge testing
  • Increased wing-rinse studies to evaluate process improvements 
  • Increased raw parts monitoring for Salmonella 
  • Implemented acidified calcium hypochlorite intervention


Eliminate Salmonella ... or else  

A continuing point of contention was the service's insistence that Cargill eliminate Salmonella from its operation, something Cargill said would not be feasible. This was a major concern as the service seemed to be formulating new, groundbreaking policy—specifically that the continued presence of a Salmonella strain linked to human illness would constitute insanitary conditions.

Cargill’s second recall  

Three weeks into August Cargill was notified by the Food Safety and Inspection Service that the agency would write a second Notice of Intended Enforcement Action because the outbreak strain had been found in the Springdale facility.
Seven isolates of the outbreak strain had been found as follows:


  • Two in birds at rehang
  • Four in mechanically separated turkey, which was bound for cooking
  • One in 85 percent lean ground turkey


According to Robach, the highest Most Probable Number of cells per gram—in one of the birds at rehang—was extremely low at only four cells. The Most Probably Number in the 85 percent lean ground turkey was 0.092, which is less than one-tenth of a cell per gram.

Cargill recalled approximately 185,000 pounds of ground turkey on September 11.

Cargill takes more aggressive action  

Robach calls Cargill’s food safety actions “the most aggressive in the industry.” Following are some of the changes implemented before production was resumed on December 15, 2011:

  • Separated HACCP plans (debone vs. bone-in)
  • Implemented high pressure processing for 85 percent lean ground turkey produced by the bone-in process
  • Established statistical process control charts, along with interventions including daily wing-rinse testing
  • Established daily testing for Salmonella, including triggers for investigation actions
  • Established test-and-hold procedure
  • Rejection of day’s production if Salmonella incidence above 40 percent
  • Rejection of day’s production if two days above 25 percent


He credits his company with staying focused on realistic approaches that result in improvements in food safety and not being distracted by disagreements with the Food Safety and Inspection Service over policy incongruities. In the end, USDA acknowledged that the mere finding of the outbreak strain was not grounds for a recall, he said.

Learning, change needed by poultry industry and regulators  

Cargill and other poultry companies learned from the outbreaks and recalls, they made changes, and their approaches are continuing to evolve. But what was the Food Safety and Inspection Service's takeaway lesson? Carry an ever bigger enforcement stick?

How would the mission of the Food Safety and Inspection Service best be described in a few words? It is a regulatory agency with enforcement powers. But is the enforcement of regulations its ultimate and highest mission? Questions come to mind concerning its mission.


  • Is the service's first and most immediate responsibility one of enforcement and building a case for applying regulatory sanctions, or the timely sharing of information with a company to achieve the quickest possible reduction in foodborne illness?
  • Which course of action by the Food Safety and Inspection Service would have the greatest likelihood of protecting the public health in an ongoing foodborne illness outbreak: Issuing an early public health advisory about the preparation and consumption of categories/sources of contaminated food, or overseeing the issuance of recall notices later when little, or none, of the product could be retrieved from the marketplace?


Does the Food Safety and Inspection Service have its priorities out of balance? Does it reflexively take the enforcement-oriented path to the exclusion of ones that might be more protective of public health immediately and/or in the longer term? Some of its choices in the Cargill case were questionable, if you accept the notion that the agency’s mission is more than enforcement of regulations and ultimately is serving and protecting the public health.

Ever bigger regulatory stick not the answer  

A big frustration for poultry producers is dealing with what seems to be a pervasive belief at the Food Safety and Inspection Service that pathogen-free poultry could be produced if companies were just willing to do so. Presently, this is not possible to consistently achieve, and experience bears this out. What’s more, the misconception that poultry processors simply aren’t willing to produce safe poultry is not only frustrating for producers it is counterproductive to achieving safe food.

If you work at the Food Safety and Inspection Service and think all the change needs to be made by poultry producers and no big changes are needed at your agency, this illustrates the challenge. Not only is the idea misguided—leading to misjudgments about the realism of regulatory approaches—it undercuts the trust needed in the industry-agency relationship. For those who would say that trust isn’t necessary—or worse, that it is not even useful—the reflex is probably to just ask for a bigger regulatory stick.

Laying responsibility all on one side, or the other, is counterproductive to making needed progress in food safety of poultry. There is more to be gained for regulators to work with producers who are committed to food safety than in fostering a system that stifles initiative and communication.