There’s a lot of chatter about the draft compliance guidelines for HACCP validation circulating. Here’s what’s up (at least at this juncture) with validation in, what I hope, is an understandable dialogue. There are technical aspects that I won’t touch upon, but just want to outline the major theme before you get to parts of a note sent out by Dr. Jay Wenther, Executive Director of the American Meat Processors (AAMP). I’ll post a bit more about AAMP stuff a few lines down the page…
Validation. What is that? Fair question. For a meat processor to operate under USDA or state inspection, it must employ a Hazard Analysis & Critical Control Point(s) (HACCP) plan(s) for various processes and products. In that plan, the operation identifies potential physical, chemical, and biological (microbes) hazards and the means by which they intend to control them. Typically, those make up the CCP’s of HACCP. I’m going to focus on biological hazards, since that’s the primary focus of this.
There may be various lethality steps (translation = steps at which microbes are killed) in any given process. It’s how things are “kept under control.” Examples of these steps might include application of mild acids to carcasses or fresh meat cuts, heating/cooking, or even irradiation. (By now, I hope we all understand that its impossible to produce a sterile carcass without any form of intervention. Another translation: intervention = step to reduce microbial load.) And, there are also other steps that may promote the outgrowth of certain spore-forming species (i.e., Clostridium species) during chilling (after cooking). So, getting products to the right cold temperature within a given timeframe is needed. A deviation from the required temperatures or other parameters is deemed “out of control.” Sounds bad, right? Something could be “out of control” if, for example, it is cooked to 168F and not 170F (if the processor specifies 170F as the required temperature and 165F is the “safe” temperature…).
In the past, laboratories have tested the efficacy of something as an intervention or other way to control microbial growth. These are called “challenge studies.” Based on that information, an intervention could be considered “validated.” AKA, it works. That method is then “scaled-up” to a processing plant, both very small and very large. Then, the plant would do occasional test to verify that everything was going according to plan.
Now, the idea is that a plant is not the same as a lab (true enough) and therefore, validation needs to occur in the plant. So, basically redo the micro experiment under “uncontrolled” conditions. I understand this idea, that because each plant is different (etc.), we need to make sure everything is operating up to snuff. And so, FSIS has issued a “draft” of what plants need to do to validate their operations. I get that.
But, some have a few concerns about the plan. The first is cost to very small processors, especially those to make a variety of products. This may be a considerable investment for them. Second, there will likely be a fairly short fuse in which all this work needs to be done. Perhaps this can be managed over a longer period of time and if less tests are required of very small processors. It will take some serious “counseling,” but in the end, it could be doable.
Here’s what some other folks have to say…
The remainder of this message was sent by Dr. Wenther to AAMP affiliates:
I email you today with an issue that has the means to impact the inspected (state and federal) meat industry and your membership significantly. When I say “significantly,” you should know that this is just as significant as HACCP implementation was 10 years ago! So, read the email and the attached document closely. In my opinion, if we all (including your members individually) don’t step up to fight this the very small and small independent inspected meat industry will be efficiently and effectively run of out inspection and potentially out of business for good. That is how serious I believe this issue to be.
As you may recall, AAMP reported on the issue of HACCP validation in the October 1, 2008 edition of the AAMPlifier. I have attached it if you need to refer to it. At that time, the issue was under discussion with the Agency (the Food Safety and Inspection Service – FSIS). Over the past months, AAMP and other meat trade associations collectively have been having discussions with the Agency regarding this issue. This was kept somewhat private as we all thought it would help improve the working relationship on this issue so we all could come to an agreeable solution. I, personally, was all in favor of validating (with testing) E. coli O157:H7 interventions as long as it was reasonable. Unfortunately, everyone was blind-sided by the complete disregard for our reasonable request of what should require validation and what should be generally accepted. Not to say that the past 6 months have been a complete waste of time, but it is now obvious that “working” with FSIS had absolutely no impact on their thought process.
At this point, I don’t know how the other organizations are responding to this guidance material. With the publication of this guidance material the Agency has definitely put the nail in the coffin of the very small and small meat industry. Unless I am misinterpreting what was printed in the guidance material, I don’t see how processors will be able to comply with this…regardless of the time given to comply. The industry observed a decrease in plants when HACCP was implemented and I see this action as systematic approach of killing the remaining inspected (state and federal) independent processors that make a wide variety of products. Basically, I see the Agency wanting establishments to produce only a few products in each facility. If not…these processors that are currently producing a wide variety of products under multiple HACCP plans and inspection will be forced out of inspection and/or out of business. I guess they still have opportunity to go custom exempt and retail exempt. So much for “Know Your Farmer, Know Your Food” that has been so widely publicized by the USDA Secretary of Agriculture.
The guidance material contains some specific information that you should focus on:
Validation will be required for critical control points
Validation will be required for pre-requisite programs if these programs are used by the establishment to support a determination that a food safety hazard is not reasonably likely to occur.
Implication: Validation will be required for ALL HACCP plans, ALL prerequisite programs (e.g., Listeria monocytogenes, E. coli O157:H7, etc.), ALL CCPS, ALL products (essentially)
In-plant validation includes two aspects:
Supporting documentation (e.g., published processing guidelines, scientific articles from peer-reviewed journals, a challenge or inoculated pack study, data gathered in-house, or regulatory performance standards)
In-plant validation (e.g., in-plant observations, measurements, microbiological test results, or other information demonstrating that the control measures can be implemented within the particular establishment).
Essentially, FSIS is wanting proof whether the establishment is meeting the critical operational parameters and that “the overall objective of the system to produce safe products is being achieved.”
Implication: If a plant has the supporting documentation that plant will meet the first requirement of validation. Although the establishment may be collecting the monitoring and verification activities over the years that only demonstrates that the procedure was followed as dictated in the supporting documentation, but does not ensure that the activity was sufficient enough to control the pathogens identified in the hazard analysis. Therefore, microbial sampling would most likely be required for ALL HACCP plans, ALL prerequisite programs (e.g., Listeria monocytogenes, E. coli O157:H7, etc.), ALL CCPS, ALL products (essentially).
The establishment should develop these data during the initial 90 days of implementing a new HACCP system
The establishment should develop these data whenever a new or modified food safety hazard control is introduced into an existing HACCP system
Establishments using existing HACCP systems developed prior to the issuance of this document that do not have the documents from their initial validation on file will need to gather data according to the timeline [ that the Agency will set out in the Federal Register notice that it issues clarifying the validation requirement]. A prudent establishment would continue sampling at an alternative frequency beyond the initial 90 day period as part of on-going verification to ensure that the HACCP system continues to be effective in controlling the identified hazards.
Implication: For new plants starting out under HACCP, they would need to collect “data” over the first 90 days through their probationary period before the official USDA grant of inspection is granted. For existing state inspected and federal inspected establishments, they would have to provide evidence that initial validation occurred. If that proof was not available they would be required to conduct initial validation in some dictated amount of time (Federal Register notice will specify this time). After the initial validation is conducted, FSIS has the “expectation” that a prudent establishment would continue sampling at an alternative frequency beyond the initial 90 day. Therefore, the expectation would most likely be to perform microbial sampling for ALL HACCP plans, ALL prerequisite programs (e.g., Listeria monocytogenes, E. coli O157:H7, etc.), ALL CCPS, ALL products (essentially) FOREVER!
Validation to demonstrate effectiveness:
In-plant validation also includes gathering data to demonstrate that the collection of interventions and process steps together in sequence produce a safe, wholesome unadulterated product.
FSIS believes that microbiological testing that combines enumeration of indicators with the presence/absence of an identified pathogen in conjunction with monitoring critical parameters plays an important role in the initial validation of many interventions for biological food safety hazards.
“A prudent establishment” would be expected to have quantifiable data, such as indicator organisms, to show the effectiveness of the process.
Implication: For most HACCP plans, prerequisite programs, CCPS, and products…the collecting the monitoring and verification for these activities will not be enough to satisfy inspection personnel. There, FSIS “expectations” would be to conduct side-by-side indicator and pathogen testing. For a ground beef example, it may include testing for indicator organisms (e.g., aerobic plate count, total plate count, generic E. coli) along with testing for the specific pathogen (e.g., E. coli O157:H7). This could occur at points along the process that could include testing of the raw materials, testing of the product throughout processing, and testing of the final package product. For an RTE product, it may include testing for indicator organisms (e.g., aerobic plate count, total plate count, etc.) along with testing for the specific pathogen (e.g., E. coli O157:H7, Salmonella, Listeria monocytogenes, Clostridium botulinum, Clostridium perfringens, etc.). The testing may be determined by the pathogens the establishment identified within their HACCP plan, but understand what the establishment is trying to validate. With lethality, the establishment would have to validate the destruction of E. coli O157:H7 (if the product contained beef), Salmonella, and Listeria monocytogenes. If the establishment identified cooling as a CCP, they may also be required to test for Clostridium botulinum and Clostridium perfringens. Refer to sections below regarding how much sampling would be required, but FSIS is alluding to testing throughout the process (raw and finished product).
A disturbing aspect of this is that the Agency has demonstrated a disregard to science with this guidance material. To date, absolutely no Agency person has provided me with a peer-reviewed, scientific supporting document that provides conclusive evidence that the “indicator organisms” (i.e., APC, , TPC, generic e. coli) is directly and specifically correlated to E. coli O157:H7, Salmonella, Listeria monocytogenes, Clostridium botulinum, Clostridium perfringens, etc. As far as I know…it does not exist. For beef, if E. coli O157:H7 did not exist at dehiding it most likely will not exist later. To find true validation, it must exist. The Agency example only shows 3 out of 13 samples that would be useful. I could envision establishments applying fecal material on carcasses immediately after dehiding just so the data that is gathered is actually credible and useful. The same principal hold true for RTE products. If the pathogen wasn’t present as a raw material then the data collected would be somewhat inconclusive.
FSIS has provided generalized scenarios to illustrate expectations as to in-plant validation and each scenario reviews the collection of multiple microbial indicators as well as pathogens.
Implication: The scenarios provided took a very simplistic approach to validation and would be easy to achieve if an establishment was only conducting a few activities. The scenarios fail to account for the fact that multiple pathogens (e.g., E. coli O157:H7, Salmonella, Listeria monocytogenes, Clostridium botulinum, Clostridium perfringens, etc.) would most likely have to be addressed since most establishments have clearly identified these pathogens within their HACCP plans.
When is in-plant process effectiveness data is required:
For regulatory specified interventions, all that is necessary is to demonstrate the specified end point is achieved.
For Appendix A and B, these documents can satisfy the first part of validation (scientific support), but the establishment must have data to show the process is effective in the plant. From the scenarios given within the guidance document it does not appear that existing HACCP monitoring and verification records may be enough to provide this support. Therefore, microbial testing would be required.
Implication: Although FSIS Appendix A and FSIS Appendix B both clearly state “FSIS considers these guidelines, if followed precisely, to be validated process schedules, since they contain processing methods already accepted by the Agency as effective” the Agency is now saying that validation at each establishment and of ALL products (essentially) must occur. Refer to the implications section of “Validation to demonstrate effectiveness.”
Sample collection for testing
FSIS believes that collecting samples at a point in the beginning of the process is necessary to establish the process’ initial microbial load
FSIS believes that collecting samples at a point after all interventions or ideally from finished and packaged products is necessary to determine whether the HACCP system, as designed, is capable of producing safe, unadulterated products
Implication: This would mean that establishments would have to take a series of microbial samplings throughout the process, or at least at two points in the process (raw material point and finished product point) for ALL HACCP plans, ALL prerequisite programs (e.g., Listeria monocytogenes, E. coli O157:H7, etc.), ALL CCPS, ALL products (essentially). Refer to the implications section of “Validation to demonstrate effectiveness.”
What to test for
Indicator organisms should be used with additional side-by-side pathogen positive/negative detection testing to gather data about the identified organisms of concern in the hazard analysis
Implication: Throughout the guidance material they mention several times that “FSIS does not advocate the introduction of pathogens in the plant environment.” Instead, FSIS advocates and has “expectations” that establishment would conduct side-by-side indicator and pathogen testing. Refer to the implications section of “Validation to demonstrate effectiveness.”
How many samples to collect
Slaughter – Collect samples to statistically represent the HACCP system’s production volume
Other HACCP processes – Collect samples to statistically represent the HACCP system’s production volume
Very small and small establishments
Slaughter – Collect samples to statistically represent the HACCP system’s production volume OR
Slaughter – the regulations for the mandatory generic E. coli testing can be used – 13 carcasses per year (with each species)
Other HACCP processes – Collect samples to statistically represent the HACCP system’s production volume
Other HACCP processes – Sample collection is determined by daily average production of 1,000 pounds or less per process category (see the ground beef sampling guidelines)
Implication: The 13 carcass guidance had at least some logic behind it. This was taken from the generic E. coli sampling method that is used to evaluate and determine sanitary dressing producer during slaughter. As for the guidance on how much to test for other process categories…the Agency just picks numbers out of the air. As for the pound specification provided…that was randomly picked out of the air for the E. coli O157:H7 testing guidelines not based on science or statistics. The Agency allows the plants to use statistics, but the very small and small industry which barely grasps statistical process control (SPC) for the generic E. coli testing currently conducted I don’t see how they can use statistics to determine appropriate testing levels and back that up with some supporting documentation unless further FSIS guidance is provided. No FSIS guidance has been given as for the amount of samples to initially collect for the initial portion of validation. If the pound limitation is set for RTE products to mimic the E. coli O157:H7 testing guidelines, it should be understood that ongoing testing would be tremendous!. FYI…one smokehouse truck can hold approximately 400 pounds of smoked sausage.
How many types of products to collect
Establishments should collect microbial data for at least one product from each HACCP category utilized. Products can be grouped, but the similarities and differences in species, process, product public health risk, and food safety hazards should be considered. This means that if your products slightly vary, you may be required to perform validation on all of them.
Implication: This could amount to the fact that most all products produced within an establishment would require validation. The Agency alludes to the fact that some products could be grouped together if the species and the product are similar, but for many establishments that utilize a variety of formulations this could be a huge determining factor. An all beef snack stick may be considered different from and beef and pork snack stick. The Agency gives an example that a whole product may be different that a sliced product, thus requiring validation for both. Take for example a whole boneless ham vs. a boneless ham sliced in half vs. a boneless ham sliced for deli sales. Due to the fact that the processing may be different, all products may require validation. Initial microbial sampling may be used for all products, but each final product may require different microbial sampling. For the establishment that is producing a wide variety of RTE products this is going to be a huge problem and an even bigger expense!
It is my belief that this initiative will completely remove a majority of the very small and small meat industry establishment from inspection. It will hinder commerce. It will stifle the meat industry by removing the variety of products currently available. It will obstruct the production of any new products from being produced and be commercially available in commerce. It will raise the cost of the products being produced, thus the overall cost the consumers will pay. It will cause establishments to cut jobs as they downsize due to the lack of inspected meat product production. It will force the meat industry to put more products out of the reach of inspection and into retail exemption. It will also put state and federal inspectors out of a job as well….what is left to inspect if these establishments cease operations?
AAMP is developing a few testing scenarios that mimic a more realistic and most likely accepted scenario to send to microbial labs. I would like to find out what the testing would cost initially for a few products. Don’t forget that this testing would have to continue every year. At this point, I don’t know if the Agency ever calculated this massive testing requirement into the cost of HACCP implementation in the beginning of HACCP. In the initial economic impact analysis, the Agency determined that the HACCP implementation would cost slightly more than one-tenth of a cent per pound of meat and poultry. I strongly feel that this initiative will increase that cost tremendously specifically for the independent very small and small meat processors…
Thank you, Jay, for letting me post portions of the note to this blog.