2007 & E.coli: What Happened?

The tremendous upsurge in E.coli 0157:H7 public health outbreaks and recalls in 2007 have not only been detrimental to public health, but a source of embarrassment to FSIS and an industry which had in recent years experienced laudable success in reducing the incidence of the pathogen.  A plethora of possible reasons for the upsurge has been bantered about, accentuated by the closure of Topps, the most recent plant victimized by processing previously-contaminated meat emanating from a source supplier slaughter plant.  While a Canadian slaughter plant has been tentatively identified as the probable source slaughter supplier to Topps, we cannot blame Canada for all our domestic recalls and outbreaks this year. 

To restore credibility to our industry, to stanch the diminished consumer confidence in the safety of our meat, as well as promoting public health, it behooves us to identify potential causes of this year’s multiple failures and take aggressive actions to remedy these causes.  The Food Marketing Institute recently endorsed FSIS’ proposal to release retailer names where meat subject to a recall had been shipped.  FMI’s reversal of its previous stance was greatly influenced by a study which revealed that consumer confidence in the safety of our food supply has fallen from its traditional 80% rating down to 66%.  Most of us share the goal of increasing consumer per capita consumption of meat products, a goal which has been sabotaged by events in 2007.

I sent a request asking for reasons for this conundrum to a variety of individuals closely aligned to the industry:  current and former meat inspection personnel, current and former plant management, and consumer groups.  Their replies follow below, not in any particular order of importance.  If indeed one solution raced to the top, the industry and FSIS would have already noticed it and taken steps to mediate the problem.  As with other national problems, we most likely are dealing with a multi-faceted enemy which can only be attacked via a variety of inter-related interventions. 

A common response was that E.coli has mutated, creating a need for development of new interventions (a new mouse trap) to address a new enemy.  Another plant operator preferred the word “adaptation”, in that E.coli may have adapted itself to anti-microbial interventions, and may have developed an antibiotic resistance.  The suggestion was made that the industry’s reliance on lactic acid is now shown to have constituted merely a temporary palliative which the bacteria has successfully circumvented.  Assuming bacteria continue to evolve, it was suggested that public health imperatives are better served by effective prevention of carcass contamination on the kill floor, rather than focusing primarily on interventions at downline further processing establishments.  As E.coli continue to adapt, it will continue to “immunize” itself against interventions yet to be invented.  That being the case, our best protection is to prevent contamination at the source, rather than playing catch up with newly variant bacteria after they’ve been shipped into commerce.  An ounce of prevention may be worth a ton of outbreaks.

Epidemiological investigations centered on the Topps recall revealed three distinct DNA patterns in impacted consumers.  PFGE sample analyses of the three DNA samples proved that each was unique, which sends shivers down the spine of industry personnel who had been lulled to sleep with visions of supposed success from our multiple hurdle pathogen intervention systems.  E.coli has indeed adapted, the degree of which we most likely have yet to discover. 

One slaughter plant operator made the following statement:  “I know you have attributed contamination to line speeds that are too fast but we are nearly at a crawl dressing our cattle and still find it difficult to keep a carcass clean when the animals are covered in manure”.  This erudite admission points us towards two related comments. 

Number one, the vast majority of respondents agree that a successful accomplishment of “Zero Tolerance” policies has a zero probability, and is an exercise in futility.  The Zero Tolerance Policy may have had positive PR benefits for FSIS, but in all honesty is impossible.  Elsa Murano was precisely correct when she stated that there is no such thing as risk-free raw meat.  Furthermore, annual certifications provided by the mega slaughter facilities always include a disclaimer which properly states that there is no magic bullet allowing for a 100% guarantee of safe meat.  Never a truer word said.  The total elimination of E.coli is a laudable goal, yet not realistic at this time.  The idea of developing animal vaccines to reduce or eliminate the incidence of e.coli in live animals has many devotees.  While this may be a long-term possibility, we are still faced with unanswered problems for present day pathogen prevalence. 

A recent article in the Minneapolis Star Tribune indirectly throws cold water on the Zero Tolerance mentality.  Part of the article stated “In the sterile suites of our ultra modern hospitals, almost 3 percent of all ‘clean surgeries’ still result in a post-surgical site infection.  This means bacterial contamination from the patient’s skin or from someone else on the surgical team infected that incision.  If surgeons can’t do any better under ideal sterile conditions, how can we expect a meat processing plant to guarantee that the carcass coming off the line doesn’t have some hidden microscopic E.coli?”  End quote.  Well, I bet that was written by a slaughter house CEO.  Guess again.  It was written by Michael Osterholm, director of the Center for Infectious Disease Research and Policy and professor in the School of Public Health at the University of Minnesota.  If the medical community admits to a 3% failure rate in post-surgical bacterial infections, kill floors should be a multiple of 3%, in spite of FSIS’ mandate for Zero Tolerance.  Once FSIS admits the impossibility of Zero Tolerance, the question remains as how to proceed. 

Secondly, once we admit that e.coli will be with us pending the creation of the silver bullet, we must develop a way of dealing with those occasions where the pathogen is detected.  No slaughter plant will ever succeed in totally removing this bug utilizing methods available today. Two correspondents recommended that the occasional detection of E.coli should not justify aggressive agency enforcement actions at a plant, but that the agency should place more emphasis on the TREND of microbial sample analyses at the plant.  If lab results reveal that the plant is experiencing slow but steady progress in reducing microbial load on its products, the plant must be classified as succeeding in reducing microbes.  Unfortunately, FSIS’ classifying E.coli as an adulterant, as well as stating that having the bug in a plant is a felony, plants are unwilling to reveal test results to the agency for fear of becoming felons.  For the same reason, plants which experience preliminary “Potential” or “Presumptive” test results regularly truncate the tests for fear of detecting “Confirmed” positives which could trigger agency enforcement actions.  Instead, such potentially positive meat is diverted to plants which fully cook the meat, implementing an effective kill step.  The dilemma of declaring the existence of a felony is especially counter productive when confirmed results are detected at downline further processing plants which unwittingly purchase previously contaminated meat.  You might ask what relevance this has to determine why 2007 has been such a tumultuous year.

If FSIS would sensibly conclude that E.coli is a contaminant, and depart from its adversarial attitude that the presence of E.coli at a plant could shutter the plant (either temporarily or permanently), progress WILL be made to (a) detect the pathogen and (b) determine the true origin of contamination.  To wit:  all respondents who discussed this issue steadfastly argue in favor of an immense increase in agency-conducted microbial testing at all slaughter plants, regardless of size and chain speed.  If the possession of E.coli-contaminated meat would no longer be considered a felony, a starting point could then be constructed from which open-minded discussions about potential improvements could be made.  FSIS sincerity to scientifically discuss cause/effect relationships would encourage plants to be more willing to participate in and release microbial test results.  Rather than maintaining its finger pointing “Gotcha” mentality and removing inspectors from victimized grinding plants, FSIS should immediately gather all the facts, find the true source of contamination, and force the source to design corrective actions which will successfully prevent recurrences.  One common conclusion surfacing from the respondents was that 2007 problems were the natural consequence of FSIS consistently finding fault with downline destination plants, while ignoring problems at the source plants.  Another respondent suggested that USDA is figuratively leaving the barn gate open at the slaughter establishments, allowing E.coli to escape, and subsequently focusing its scrutiny at downstream destination plants after the pathogens have already entered commerce.  His suggestion was to detect E.coli before it gets out of the “farm gate” (i.e. the slaughter house), and force the kill plant to initiate policies to prevent recurrences. 

One meat inspection official pilloried Notice 65-07 which continues to place an inordinate focus on E.coli interventions at downline further processing plants which do not slaughter.  Her claim was that focusing at plants which are merely the destination of previously-contaminated meat adroitly avoids meaningful agency oversight at the slaughter house source of contamination. 

No one denies that E.coli is an enteric bacterium, almost always emanating from sloppy kill floor dressing procedures.  It stands to reason then that microbial testing protocol should be primarily focused at the slaughter facilities, which is contrary to current agency methodologies.  As long as FSIS demands that E.coli be followed downstream and confronted at the destination plants, the source plants will have no incentive to implement corrective actions.  But again, why has this culprit exploded in such magnificent fashion in 2007?  Why did E.coli wait until 2007 to expose itself, while remaining in relative seclusion the previous few years? 

One idea presented was that public health officials (outside FSIS) have become better educated in identifying E.coli to be the culprit causing sicknesses in their patients.  They’ve also become more proficient in communicating with related entities such as city, county, state and national health organizations, as well as CDC.  They likewise are accessing and inputting information on PulseNet.  FSIS this year announced a closer working relationship with CDC in sharing public health outbreak data.  2007 may simply have been the “break out” year when public health officials reached critical mass empowerment, and are now issuing correct prognoses of e.coli sicknesses, rather than intestinal flu or other misdiagnoses.

One respondent suggested that following several years of news stories reporting remarkable pathogen prevention success over E.coli, consumers may have become carelessly apathetic in their food handling procedures and cooking methods.  Perhaps Americans had presumed ground beef to no longer be dangerous, and failed to see the potential for cross contamination.  While this is an interesting thought, its validity diminishes whenever outbreaks occur impacting multiple consumers.  These wide-spread outbreaks reveal that the pathogen had permeated a wide swath of meat, impacting numerous households.  The possibility was also suggested that many of this year’s sicknesses were not caused by eating meat that was not fully cooked.  Rather, kitchen cross contamination of other foods served at the meal may have been the culprit.  While contaminated meat was the originating culprit, perhaps fruit, bread, salads etc may have been the carrier to the consumer.  A follow-up suggestion was made that FSIS, our industry, and consumers should jointly produce an educational video which shows the multitude of ways that household consumers and restaurant operators can unwittingly cross contaminate non-meat items during food handling and meal preparation.

The thought also surfaced that multiple problems in 2007 reveal that the foundational underpinnings of HACCP have failed.  When FSIS introduced HACCP theory in the mid-90’s, the agency hailed (a) Prevention and (b) Corrective Actions as foundational corner stones for its new “science based” HACCP system.  Prevention included several components such as the initial hazard analysis, establishment of critical limits, critical control points, SSOP’s, SOP’s, pre-requisite programs and so forth.  Beyond this, on those infrequent occasions when pathogen problems would arise, plants under the HACCP umbrella would then initiate Corrective actions to prevent recurrences.  Events in 2007 prove that we as an industry have failed in implementing efficacious Corrective actions.  Several respondents forcefully concluded that FSIS is almost exclusively requiring corrective actions at downline further processing plants, the majority of which do not slaughter.  Although such downstream plants do not introduce E.coli into the food chain, FSIS demands that the downstream plants prevent the bug’s recurring detection.  As such, corrective actions to eliminate E.coli will continue to be a dismal failure, for obvious reasons.  One person suggested that HACCP has devolved into a system of theoretical Prevention steps, devoid of Corrective actions.  While this does not fully explain why 2007 exploded, it prophesies subsequent flare-ups based on our failed Corrective actions. 

Another meat inspection official provided several thought-provoking comments, which I will repeat verbatim:

“Laboratory tests for 0157:H7 are much more advanced and sophisticated than they ever were before”.  If this is true, the incidence of E.coli may not have increased, but we are better at detecting the bug. 

“If Inspection takes a sample for E.coli 0157:H7 from a grinding operation (not a slaughter plant) and it comes back positive, then an NR will be written and the grinding plant must reply either on the NR or separately to address the Corrective Actions as per 417.3.  How can the plant truthfully respond to the Corrective Actions such as:

  1. The cause of the positive sample came from the Packing House (how can the grinder identify the cause of this deviation when it happened in another plant?)
  2. The CCP will be under control (how can the grinder control something he is not responsible for?)
  3. What measures to prevent recurrence can the grinder do other than sample each and every lot when the product comes in with the USDA Inspected & Passed Legend and Purchasing Specs which state that an E.coli 0157:H7 intervention step was taken, such as a Lactic Acid Rinse, hot water rinse, etc?
  4. No adulterated product is shipped to commerce (the grinder does not know this unless he samples each and every box).

End of quotation.  Questions posed by this meat inspection official reveal the professional and ethical dilemmas faced by meat inspection personnel when confronted with this conundrum.  The decision is always the same:  force the downstream plant to reassess their HACCP plan, concoct some meaningless dialogue, and continue to insulate the noncompliant source plant from any corrective action.  As such, events occurring in 2007 are guaranteed to duplicate themselves in future years.  Corrective Actions, one of the original legs on which HACCP was constructed, have been short circuited by misdirected agency enforcement actions.  Because of this, we should not be surprised at the ugly conditions revealed in 2007. 

I will next quote a reply provided by a plant operator.  “Have new innovations been deployed at the slaughter facilities in an attempt to lower e.coli contamination on carcasses, only to result in higher levels due to some overlooked factor.  To give you an example, we once started using an antimicrobial conveyor belt from a new manufacturer because initial testing showed promising results only to find out that the belt was made of a somewhat softer material and in a short period of time was so beat up that it became difficult to clean and in the end was more of a harborage for bacteria than our regular belts”.  This operator’s questions can only be answered via validation studies on each new piece of equipment.  Just like our pell mell rush twenty years ago to implement the metric system which resulted in unanticipated problems, our industry may also be reaping problems stemming from a knee-jerk attempt to overcome FSIS regulatory actions by deploying equipment and procedures which have not been adequately test fired.  Again, if FSIS were to engage in a meaningful microbial testing protocol downline from these much-ballyhooed interventions, the success or failure of process changes would be expeditiously revealed.

A common idea posed by meat inspection personnel was that staffing shortages within inspection ranks have allowed contaminants to evade detection on kill floors.  Also, an overstretched and thinly-staffed inspection force simply lacks the time required to (a) observe production lines and (b) to adequately review company-generated paperwork, including microbial sampling data.  I was reminded that FSIS was embarrassingly forced to admit earlier this year that their inspectors had NOT been inspecting every plant on a daily basis as the agency had earlier claimed.  And, even though inspectors on patrol assignment frequently do visit each plant daily, their visits merely constitute a cursory, window-dressing plant visit, lacking adequate time to detect problems.  When situations occurred justifying issuance of NR’s, the lack of time frequently resulted in NR’s not being written, or delayed to a subsequent day when details were no longer fresh; plant management were thereby justified in their denunciation of NR’s which depended on fleeting past memories.  This is yet another reason why 2007 has provided multiple outbreaks, because FSIS lacks personnel required to detect problems and secondly to implement appropriate enforcement actions.  Therefore, noncompliant plants continue to operate as is.  Even though such plants may have theoretically plausible preventive actions incorporated into their HACCP plans, corrective actions are not being implemented due in part to FSIS’ inability to (1) perceive problems, and (2) document these problems.

Another respondent referred to two implications of an industry which is concentrated in the hands of an increasingly small oligarchy, both wholesale and retail.  While most attention is devoted to the small number of slaughter behemoths, the idea was posed that retail concentration may exert an equally adverse role.  The mega retail operators bid lower prices for beef, forcing the slaughter plants to become more efficient, whether by increasing chain speeds, by reducing kill floor personnel, or whatever means are required to lower operating costs.  Furthermore, when a large slaughter plant produces e.coli-laced meat, the potential market of this contaminated meat may be nationwide in scope.  Therefore, one innocent error on the kill floor can easily sicken dozens of victims in several states.  Therefore, we should not be surprised when we experience simultaneous outbreaks over a wide geographical area.  As the slaughter industry continues to consolidate, and the number of small and very small slaughter plants continues to shrink, we should anticipate additional outbreaks increasing in size.

Exacerbating this problem is the high turnover of unskilled workers, increasing the possibility of carcass contamination.  Furthermore, ICE raids have accentuated the difficult worker turnover problem.  The same respondent states and I quote “The fact that packers know they will not be identified as the source for contamination hasn’t helped their concern for food safety”. 

For the variety of reasons above, 2007 was the “perfect storm” year which precipitated the dozens of recalls and outbreaks.  We can expect more of the same, until the truly noncompliant source plants are forced to clean up their act.

In addition to the above ideas, numerous others were suggested which I failed to enclose because my personal perception is that the above statements more closely defined the base reasons for the untimely events witnessed in 2007.  There will not be one silver bullet which FSIS and our industry can implement to fully eradicate this increasingly ubiquitous bacterium.  Delineating problems is relatively easy, but proposing common sense solutions is arguably more difficult.  While a recitation of industry problems in 2007 is easy, offering plausible solutions is more difficult.  Based on input received from respondents, I am utilizing executive privilege and will make the following suggestions which I feel will reduce the incidence of this elusive pathogen and promote public health: 

  1. Government entities should proactively work with the pharmaceutical and livestock industries to develop a vaccine which will lower the incidence of E.coli 0157:H7 in live animals, as well as confronting newly-emerging variant E.coli mutations just now rearing their ugly heads.
  2. FSIS should reclassify E.coli as a contaminant, admit that it may never be totally eliminated, and work with plants to ensure that plant microbial results continue to validate decreased trends in the incidence of microbes.
  3. FSIS should implement a dramatic increase in microbial testing for E.coli in slaughter plants of all sizes. 
  4. FSIS must place primary emphasis at the source of enteric contamination, that is, slaughter plants.  When E.coli is detected at downline further processing plants, FSIS must leave no stone unturned to determine the source of E.coli contamination.
  5. FSIS must proactively partner with all public health entities to immediately post epidemiological PFGE/DNA data.  The failure to do so, and on a timely basis, was revealed subsequent to the Topps recall and closure.  FSIS, FDA, CDC as well as local and state health authorities should release all lab test results to the public on a real-time basis.
  6. Corrective action at the source of contamination must again become a corner stone of HACCP.  Only when the true source of contamination is required to implement corrective actions will we experience fewer outbreaks and recalls. 
  7. Since consumers and the HRI industry are part of the farm-to-fork continuum, USDA and FDA should redouble their efforts to train end users how cross contamination can occur, as well as the need to fully cook meat. 
  8. FSIS and FDA should request sizeable budgetary increases to hire and train more inspection personnel.  Furthermore, agency field force should be authorized to closely scrutinize meat production lines, in spite of the fact that HACCP theory has forced the agency into a “Hands Off” deregulated status.


One last thought:  in the last 2 weeks I’ve received numerous comments from plant management and agency field personnel who are expecting a deluge of FSA’s resulting from Notice 65-07.  As the agency knew full well in advance of inspectors’ compiling the checklists, the small downline further processing plants do not hire third party auditors, do not collect microbial samples on incoming products from all suppliers every month, or even every quarter.  They also do not have the veritable array of costly interventions utilized at the large slaughter plants.  It is no secret that small plants cannot afford to implement these actions.  As a result, FSIS will play havoc with these downline plants which are merely the destination of meat emanating at the mega slaughter plants.  If the agency can successfully shutter a goodly number of these allegedly noncompliant downstream plants, FSIS could then proudly claim success to the news media, as though the culprits causing the 2007 disasters have been identified, and removed from inspection.  May sound good, but what happens next year when the outbreaks continue?  Answer:  I’ll provide you another report next fall, and list Notice 65-07 as the primary cause, because the Notice continues to reveal that FSIS refuses to address the SOURCE of contamination, while greatly increasing its scrutiny at the small downstream DESTINATION plants.  A commonly shared opinion is that these destination plants are expected to validate the efficacy of their slaughter supplier HACCP plans, because FSIS refuses to closely scrutinize these slaughter plants, gratis HACCP.  The large slaughter plants must validate their HACCP plans themselves, and neither the slaughter plants nor FSIS are legitimate in their demand that the destination plants do the validations.  As long as this mentality exists, we can expect future outbreaks.

Resolutions to the 2007 disasters are available, and Notice 65-07 is not one of them.  The now-obvious consequences to 65-07 will merely continue and exacerbate this public health problem. 

Dated this 3rd day of December, 2007.
John W. Munsell, Manager
Foundation for Accountability in Regulatory Enforcement (FARE)