This is a followup on my response to Brianna Leach’s opinion essay published by Food Safety News on May 26. My initial response focused on HACCP. In order for Part Two to make sense, readers might consider reading Part One, entitled “Is contributing author HACCP trained?” published by Food Safety News on May 28.
The two legs upon which HACCP was designed were Prevention, and Corrective Actions. If unsafe meat is produced, a plant must implement Corrective Actions to prevent future recurrences. This is foundationally important! How can microbial test results aid in Corrective Actions?
When FSIS mandated HACCP implementation by the meat and poultry industries, the agency justified the mandate by explaining the new system would be “science-based”. A major component of “science” would entail required microbial sampling, both by the agency and the plants.
In Part One, I defined HACCP as a preventative system designed to control significant identified hazards by means of validated Process control measures. Prior to selling my plant in 2005, my Slaughter hazard analysis contained fifteen Process Steps. My Raw Ground hazard analysis contained seventeen Process Steps. Thus, if my plant produced unsafe meat, which of these 32 Process Steps experienced failures? How can a plant identify the Process Step(s) which were deficient? The guilty culprit can be identified by validating each Process Control step in the plant’s production line. Under True HACCP, plants validated each Process Step to prove the entire system qualified for HACCP. Plants focused testing at those Process Steps most likely to contaminate food. Testing finished product is fully incapable of identifying which Process(es) failed.
E.coli and Salmonella are “Enteric” bacteria, which is defined as “Of or within the intestine”. When intestines empty themselves in the form of fecal manure, and the manure attaches to hides, we see that these bacteria will spread not only from perforated intestines, but also from manure-covered hides. The primary origin of E.coli and Salmonella is the slaughter operation. Granted, contamination can also occur from other origins, but its occurrence is insignificant. Most further processing plants do not slaughter, nor do retail meat markets. Neither entity has intestines or manure-covered hides on their premises. Both entities purchase all their meat from outside source slaughter providers. When meat is sampled from these entities, adverse test results would seem at first blush to indicate that the entity has a failure in its production line, intimating sanitation failures.
USDA/FSIS refuses to acknowledge the GIGO (Garbage In, Garbage Out) effect in such situations. Both of these entities rely entirely on the safety of meat purchased from their up line slaughter providers. But, FSIS endorses assessing full liability against the downstream further processor. As such, FSIS expects the downstream entities to (1) detect the presence of invisible pathogens in meat purchased from the outside, and (2) remove the invisible bugs. Ironically, neither the source originating slaughter plant nor its FSIS inspectors were capable of accomplishing either (1) or (2), allowing the product to be shipped into commerce in containers proudly labeled “USDA Inspected and Passed”. PASSED pathogens!
Therefore, what kind of Corrective Actions can the downstream entities perform to prevent future recurrences of contaminated meat?
Years ago, FSIS piously proclaimed that it preferred to perform sampling “As close to the consumer as possible”. This preference may still exist today? This statement portrays the agency as being a premiere pro-consumer food safety activist; conveniently ignoring that such testing is performed as far away from the source of contamination as possible.
Under HACCP, FSIS mandated a “Zero Tolerance” for fecal, ingesta, and milk. Thus, the agency allegedly prohibits the existence of any E.coli in meat products. More on this later.
Testing finished product, such as ground beef, circumvents the possibility of corrective actions since in many cases the true SOURCE of contamination is not identified, by intentional agency design, and which of the dozens of Process Steps was/were deficient. Indeed, if finished product testing reveals the presence of pathogens, and consumers were notified, fewer sicknesses would occur. But, what can be done to promote HACCP’s foundational Corrective Action mandate? Reducing future recurrences is diplomatically ignored. Dr. William Sperber, prominently mentioned in Part One, states “Nothing fails like finished-product testing to assure food safety. It was the failure of conventional QA finished product testing in the 1960’s that did not detect Salmonella in many dried foods and ingredients that led to the current HACCP system in the food processing industry.”
When FSIS introduced its version of HACCP in the 90’s, the agency explained that microbial sampling would be an integral and required part of this improved system, heretofore mostly ignored. In fact, in my plant’s fifty-four years of operating prior to HACCP implementation, neither the agency nor the plant conducted any microbial sampling. After HACCP’s implementation, both FSIS and my plant collected multiple dozens of samples for ongoing lab analysis. As a result of agency sampling of my ground beef, I recalled 270 lbs of product in January, 2002. In February, we experienced three consecutive days of lab positives, seemingly indicating my plant was filthy! More of this later.
The largest plants implemented HACCP on or before January 26, 1998. A mere six days later, on February first, FSIS issued Directive 10,010.1 which exempted some plants from agency-conducted sampling. The Directive stated that plants which already do their own sampling, and had experienced zero positives in the previous six months, were exempted from FSIS sampling. We will see how that exemption worked out.
During my recall in January, 2002, the FSIS inspector who collected the sample immediately declared that the meat ground had originated from another plant, NOT from my kill floor. Thus, agency enforcement actions were quickly concluded. Before releasing its recall to the media, the FSIS recall committee called me to confirm I was willing to conduct a recall. The committee was a multi-disciplinary group of ten agency veterans, top notch folks. After agreeing to the recall, a committee member informed me that I must conduct a HACCP Reassessment, and implement corrective actions to prevent recurrences. I asked the committee what corrective actions were available to me, realizing the meat was Coarse Ground Beef I had purchased from an outside slaughter provider. I asked how I could force my suppliers to ship me strictly safe meat. The committee’s reply was a veritable watershed moment for me. They replied………nothing! Bingo, my education had begun. Now angry, I repeated the question, indignantly. Finally, one of its members replied that I should obtain a copy of my supplier’s HACCP Plan. Now what good would that have done? But, this gathering of the best technicians the agency could gather had to provide me some kind of reply, however inane.
At the time, agency policy required the collection of fifteen additional samples. In the midst of the fifteen, we experienced three consecutive days of lab positives, on February 19, 20 and 21. Now however, we had copiously documented all related evidence regarding the source of the unsafe meat, and all our grinding protocol on every day. Upon our review of grinding records, we discovered that the source of all three positives was plant # 969, ConAgra’s plant in Greeley, Colorado, all from one grinding lot on one production day. A document hand-written and signed by the inspector and his veterinarian supervisor which referred to # 969 was featured in a subsequent NBC Nightly News segment. How did the Minneapolis FSIS District Office respond to such smoking gun evidence? The DO stated I had no proof that the unsafe meat did not originate from my own kill floor (why document anything?), shut down my grinder for four months, while allowing #969 to continue operations with no corrective actions.
What happened after four months? A similar conundrum occurred at a meat plant in Denver, again producing several positives after grinding meat from # 969. No longer able to hide its head in the sand, FSIS intervened, and #969 conducted two recalls totaling 19.1 million pounds. The agency then graciously allowed me to commence grinding again.
Smoking gun evidence from two plants, emanating from this 15-sample protocol would have been worth its weight in gold, had the agency responded in an ethical, scientific fashion. Because the protocol was embarrassingly effective, FSIS had to do something. So, it eliminated the 15-sample protocol! The agency substituted a different protocol, including a dramatic reduction in the number of samples. This change was announced in Notice 70-10, dated November 30, 2010.
This story is not about my plant, but about systemic attitude problems within FSIS which effectively insulate the large source plants from pathogen liability, while passing all liability downstream to victimized further processors. Corrective actions were ignored at the Source. Compared to original, True HACCP, FSIS sampling policies emasculated the food safety value of FSIS-style HACCP. Similar problems have occurred all across America at small FSIS-inspected facilities.
Following ConAgra’s 19.1 million lb recall, USDA’s OIG performed an investigation. OIG’s findings included numerous scathing revelations, some of which were:
“USDA had reduced its oversight short of what was prudent and necessary for the protection of the consumer”
“Although animal feces on product was repeatedly observed during production at ConAgra, USDA took no enforcement action”
“From April 12, 2002 through October 21, 2002, OIG noted at least 115 E.coli positives at the ConAgra plant”
Likewise, an article in the local Greeley Tribune newspaper stated that in the 100 days prior to ConAgra’s recall, the plant had experienced 34 lab positives, in the absence of effective Corrective Actions.
How could ConAgra have qualified for the exemption required in Dir 10,010.1, requiring zero positives in the previous six months? Also, plants were not required to share lab results with the agency, directly benefitting FSIS comfort. If FSIS had been aware of results, the agency would have been required to step in and attempt some enforcement actions, which it was loathe to do. The agency now has access to plant-conducted pathogen testing results.
And whatever happened to the agency’s “Zero Tolerance” mandate? Well, the mandate only pertains to small plants, not to the industry’s hegemons. Adding insult to injury were callous but truthful remarks made by then-USDA Under Secretary for FSIS, Elsa Murano. In a July 25, 2002 article in the World-Herald Bureau, Murano was quoted as saying:
“Because Est. #969 slaughters 5,000 head of cattle and produces about 800,000 pounds of beef daily at the Greeley plant, one or several positive E.coli tests on some days wasn’t cause for alarm. That is not that high a number”.
I might add that one positive test at a small plant results in aggressive FSIS enforcement actions. As stated in Part One, FSIS-style HACCP has conveniently deregulated the large meat plants, while hyper-regulating small plants, forcing many out of business. This is FSIS-style “science”.
When my plant experienced three consecutive days of adverse lab results, the broker who sold me the Coarse Ground Beef provided me a PR pamphlet produced by ConAgra. The pamphlet explained the multiple hurdle intervention steps implemented by the plant, including some scientific work provided by Colorado State University. It’s conclusion was that the wide portfolio of ConAgra’s interventions produced a “99.99999% (7-log) reduction in pathogens, virtually sterilizing carcasses”. It’s subsequent 19.1 million lb recall reveals that science can sometimes be stranger than fiction.
FSIS had the authority, but lacked the will, to quickly change its sampling policies. For example, on July 26, 2002, the agency’s Office of Field Operations sent an email to all District Offices stating:
:At the time the [ground beef] sample is taken, the IIC will obtain from the establishment, the name, point of contact, and phone number for the establishments supplying the source materials for the lot of ground beef being sampled.”
Unfortunately, this remarkable agency about face policy change was rescinded two months later. An agency official stated at a public meeting that the agency rescinded it “For Legal Reasons”.
Three months later, a Senate Field Hearing was conducted in Billings, MT on December 11, 2002 to discuss these issues. At the Hearing, I suggested that FSIS inspection personnel document all source evidence at the time of sample collection. The Hearing moderator, Senator Conrad Burns, asked William Smith, then-FSIS Asst Administrator Office of Field Operations “Mr. Smith, do you want to respond to that?” Mr. Smith replied “…..we could have inspectors doing more important things”. Perhaps food safety is not important? Nor are meaningful Corrective Actions. Mr. Smith also stated “…..a number of packers would also be very upset about us collecting information on negative findings”. Mr. Smith was unusually candid, because the largest packers would indeed be upset if inspectors would document the origin of meat being sampled, imperiling large source slaughter plants.
As previously stated, FSIS policies did not permit inspectors to document the origin of sampled meat at the time of sample collection. A full eight years later, on October 8, 2010, FSIS issued Notice 58-10 which mandated source documentation when agency inspectors collected meat samples for lab analysis. The next year, on August 3, 2011, Ag Secretary Tom Vilsack directed FSIS to announce, in 90 days, the first step in transforming the agency’s traceback policy. One year later, as related in Helena Bottemiller’s article in Food Safety News dated May 3, 2012, Under Sec for Food Safety Dr. Elisabeth Hagen said that FSIS would soon begin tracing meat contaminated with E.coli O157:H7 back to the source before illnesses are linked to the product and in some cases even before test results are confirmed positive. Dr. Hagen stated that FSIS would “move quickly to identify the supplier of the product”, and then if the test is confirmed positive, notify any processors who received contaminated beef from the implicated supplier to recall product. Helena Bottemiller also stated that at the time, FSIS did not conduct a traceback investigation unless there were illnesses linked to the product. Yet two years later, on August 14, 2014, Dan Flynn’s article in Food Safety News stated “On August 13, USDA announced that FSIS is adopting new procedures to trace contaminated ground beef back to the source more quickly. One tool will be grinder logs that small retail meat shops will now have to maintain.” Ironic that a full twelve – fifteen years after the industry implemented HACCP, only then did FSIS begin developing traceback policies. Without irrefragable source evidence, the industry cannot implement meaningful corrective actions, a truism lost on FSIS.
After we had sent a ground beef sample to an outside lab years ago, the lab subsequently called me to state that the sample was potentially positive, and did I want the lab to continue the test to determine if the sample was confirmed positive. I was incredulous. The lab rep explained that if we terminated the lab’s testing of that sample, we would be able to state that we had not experienced a confirmed positive. I then asked what was the value of microbial sampling if I prematurely terminated lab findings? He merely responded that plants have the right to stop additional lab tests. I instructed him to finish the test, which fortunately resulted in a confirmed negative.
So when Brianna Leach referred to microbial sampling of product, she needs to be fully aware of FSIS-style sampling protocol which reduces the ability to determine the true source of contamination. Plus, she and all consumers deserve to understand lab protocol.
In conclusion, microbial lab results can be useful in promoting food safety and public health, if they allow a scientific identification of the source of contamination. Until FSIS develops the courage to force the source plants to implement meaningful corrective actions, the production of unsafe products will continue, and consumers will be sickened. Meaningful corrective action steps are short-circuited by the lack of efficacious critical control points (CCP’s) for raw meat and poultry.
Standing behind the skirt of true HACCP, FSIS has lulled consumers to sleep. Explaining that the agency adheres to a “Zero Tolerance” mandate for E.coli, ostensibly possible because of HACCP’s advent, consumers perceive that all raw meat and poultry should be safe to consume. When the meat industry, as well as USDA suggest that consumers are responsible to cook raw meat to kill any resident bacteria, consumers and consumer organizations howl and accuse the industry of shirking its responsibility. It is necessary to admit, although embarrassing to FSIS, that all raw meat and poultry carries inherent risk, because it has not been subjected to a legitimate CCP kill step. How can that be? Frankly, because FSIS-style HACCP is not true HACCP, which as I previously stated was established for products intended to be ready-to-eat. As I stated in Part One, the top agency official in 1995 admitted that FSIS had changed HACCP. The problem is not with the meat industry! The primary culprit here is FSIS, which falsely portrayed its version of HACCP as being capable of producing consistently safe raw meat and poultry because of HACCP. My hope is that this Part Two clearly elucidates for everyone multiple problems inherent with FSIS-style HACCP — for raw products.
The primary proteins my wife and I purchase are raw ground beef and boneless chicken breasts, both high risk products. But we treat both items as highly lethal, because they are! The mere fact that both products have carried the official USDA Mark of Inspection means nothing to us. We accept the reality that the only means available for us to consume safe meat and poultry is to accept rigorous cooking and handling protocol in our kitchen.
Perhaps two years ago, Seattle food safety attorney Bill Marler authored a blog article in which he complimented the meat industry for its progress in producing safe meat. Marler stated that when he got his start defending consumers who had been sickened in the Jack In The Box outbreak, his primary workload was litigation against meat plants. In his relatively recent article, Marler admitted that his business had very few cases involving meat plants. The meat industry has indeed made substantial progress in its goal of producing safe meat, while yet acknowledging the futility of producing consistently safe raw meat and poultry in the absence of valid Critical Control Points such as full cooking and irradiation.
Thus, we can deduce that HACCP’s implementation in the meat and poultry industry has resulted in safer food. Would the industry have unilaterally implemented these changes without the FSIS mandate? No. The fact still remains that raw meat and poultry does not qualify for true HACCP, in the absence of a legitimate CCP. So when contaminated meat is produced, FSIS cannot in good conscience claim the affected plant has a failure in its HACCP Plan in the total absence of a kill step. This awkward conundrum spells trouble for FSIS, which has an insatiable appetite for pointing fingers at allegedly insanitary plants. Typically, these downstream further processing plants are but innocent victims which unwittingly and legally purchased source meat from upline suppliers, meat laced with invisible pathogens. Coupled with the agency’s historical ambivalence towards tracing back to the source of contamination, American consumers need to handle and cook raw meat and poultry with the greatest of caution.
The December 2009/January 2010 edition of Food Safety Magazine included an insightful article authored by afore-mentioned microbiologist Dr. William Sperber and Richard Stier. The article was entitled “Happy 50th Birthday to HACCP”. The article’s concluding statements were as erudite now as they were back then, and provide a fitting capstone to this Part Two. The statements were:
“Whatever the ultimate configuration of HACCP system details, it is comforting to note that the HACCP pioneers, beginning 50 years ago, established features of HACCP that have not changed.”
The authors then identified three features, one of which was:
“It [HACCP] does not rely on product testing or lot acceptance criteria”, and “These features are the timeless essence of HACCP, serving as a permanent testament to the vision of the HACCP pioneers”.
FSIS-style HACCP, lacking consistent scientific integrity, has given true HACCP a blemished reputation, while misleading consumers.
(To sign up for a free subscription to Food Safety News, click here.)