Jump to content

  • Quick Navigation
Photo
- - - - -

EMP in a RTE facility that makes vegetable filled hand pies


  • You cannot start a new topic
  • Please log in to reply
6 replies to this topic

#1 drquinn1964

drquinn1964

    Grade - Active

  • IFSQN Associate
  • 6 posts
  • 3 thanks
2
Neutral

  • United States
    United States
  • Gender:Male
  • Location:Galt, CA

Posted 18 March 2021 - 05:52 AM

I am veterinarian by trade that after 15+ years of mixed animal practice worked for 10+ years as a Supervisory Public Health Veterinarian and Enforcement Investigations and Analysis Officer with USDA-FSIS before starting my own company providing food safety and humane handling consultation to USDA-FSIS regulated facilities 3+ years ago. I have recently branched out into FDA food manufacturing facilities and recently completed PCQI training from Alchemy Academy. I am now helping a small FDA RTE facility that makes vegetable filled hand pies to rewrite their entire food safety system so that it is FSMA Part 117 compliant.  While developing their EMP I have reviewed multiple guidance documents and presentations so that I am comfortable with the EMP I have developed and am recommending to the facility owner.  This facility has never conducted sampling and testing and they do not have a lot of financial resources to conduct large scale upfront sampling and testing.  Dr. Marshall from Eurofins suggests in his EMP presentations that when first starting facilities should sample numerous sites, but that is as far as I can see in all of the guidance documents and presentations regarding how many sites and how often in the beginning, i.e. to build a baseline.  To begin with I am having them sample and test 5 Zone 1 FCS twice weekly for L. species, APC, EB, and Coliforms; 5 Zone Non-FCS twice weekly for Salmonella, L. species, APC, EB, and Coliforms; 5 Zone 2 Non-FCS twice weekly for Salmonella, L. species, APC, EB, and Coliforms; and the same for 5 Non-FCS in Zones 3 and 4; all for 4 weeks.  This is about $15,000 of sampling and testing.  So, I am trying to determine if this sampling scheme has a chance to provide me adequate information to further determine the following so that I can reduce the ongoing sampling and testing expenses.

  1. frequency and number of sample locations for ongoing sampling in each zone; and
  2. if APC, EB, and/or Coliforms can be correlated adequately to Salmonella so that one or more of the indicators can used instead of Salmonella on a regular basis with perhaps bi-annual or quarterly Salmonella verification.

I have some basic statistics knowledge, and have used an online sample size calculator, http://powerandsamplesize.com, previously, but was burned by FSIS-RIMS for using it in a poultry slaughter establishment validation study, so I am not confident with using it, unless someone can help me understand how to support its use or offer a better suggestion.  I have seen references to Microorganisms in Foods 8 as a resource that may offer some further guidance but I am not yet ready to spend a couple of hundred dollars on a possibility unless someone can say that this reference text will really provide guidance. Any guidance would be greatly appreciated. Thank you to all that take the time to read my novella and provide assistance!

 

Sincerely,

 

Quinn H.



#2 Charles.C

Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 18,311 posts
  • 5129 thanks
1,129
Excellent

  • Earth
    Earth
  • Gender:Male
  • Interests:SF
    TV
    Movies

Posted 18 March 2021 - 01:58 PM

I am veterinarian by trade that after 15+ years of mixed animal practice worked for 10+ years as a Supervisory Public Health Veterinarian and Enforcement Investigations and Analysis Officer with USDA-FSIS before starting my own company providing food safety and humane handling consultation to USDA-FSIS regulated facilities 3+ years ago. I have recently branched out into FDA food manufacturing facilities and recently completed PCQI training from Alchemy Academy. I am now helping a small FDA RTE facility that makes vegetable filled hand pies to rewrite their entire food safety system so that it is FSMA Part 117 compliant.  While developing their EMP I have reviewed multiple guidance documents and presentations so that I am comfortable with the EMP I have developed and am recommending to the facility owner.  This facility has never conducted sampling and testing and they do not have a lot of financial resources to conduct large scale upfront sampling and testing.  Dr. Marshall from Eurofins suggests in his EMP presentations that when first starting facilities should sample numerous sites, but that is as far as I can see in all of the guidance documents and presentations regarding how many sites and how often in the beginning, i.e. to build a baseline.  To begin with I am having them sample and test 5 Zone 1 FCS twice weekly for L. species, APC, EB, and Coliforms; 5 Zone Non-FCS twice weekly for Salmonella, L. species, APC, EB, and Coliforms; 5 Zone 2 Non-FCS twice weekly for Salmonella, L. species, APC, EB, and Coliforms; and the same for 5 Non-FCS in Zones 3 and 4; all for 4 weeks.  This is about $15,000 of sampling and testing.  So, I am trying to determine if this sampling scheme has a chance to provide me adequate information to further determine the following so that I can reduce the ongoing sampling and testing expenses.

  1. frequency and number of sample locations for ongoing sampling in each zone; and
  2. if APC, EB, and/or Coliforms can be correlated adequately to Salmonella so that one or more of the indicators can used instead of Salmonella on a regular basis with perhaps bi-annual or quarterly Salmonella verification.

I have some basic statistics knowledge, and have used an online sample size calculator, http://powerandsamplesize.com, previously, but was burned by FSIS-RIMS for using it in a poultry slaughter establishment validation study, so I am not confident with using it, unless someone can help me understand how to support its use or offer a better suggestion.  I have seen references to Microorganisms in Foods 8 as a resource that may offer some further guidance but I am not yet ready to spend a couple of hundred dollars on a possibility unless someone can say that this reference text will really provide guidance. Any guidance would be greatly appreciated. Thank you to all that take the time to read my novella and provide assistance!

 

Sincerely,

 

Quinn H.

 

Hi drquinn,

 

afaik many zonal systems are related to categories like wet/dry, RTE/NRTE(with "killing" step) so as to focus target priorities.

 

The choice may also relate to whom has Regulatory jurisdiction of your process ?. FDA ?

 

IMEX most reference texts to Sampling focus on evaluations related to Lot/Batch or Continuous Production Scenarios, current situation appears to be neither ?.

 

A variety of zonal procedures are illustrated here -

 

http://www.ifsqn.com...ls/#entry100060

http://www.ifsqn.com...am/#entry119334

 

Offhand, I anticipate that all of the Plans in the above examples are "ad-hoc". Personally I do not recall seeing any analytically derived zonal sampling schemes although FDA's methodology may have such a basis but I have only seen implementation articles.

 

I appreciate you are attempting to determine baselines but it may be relevant to assess how may baselines you actually need ? Yr sampling density is, predictably, astronomic compared to most schemes in above links. Offhand, only FDA maybe semi-competes.

 

Micro.orgs 8 is an updated/expanded version of, IIRC, Micro.orgs 2 and is very useful but, from memory, has limited content related to the current  situation.

 

PS - The FSIS Listeria Sampling System (which you are no doubt familiar with) is probably the most intricately detailed EMPG food sampling scheme I have encountered but IIRC is focused on Processes which contain a "killing" stage which may not be the case in yr Process unless perhaps a baking RTE set-up ?.

 

PPS - Re ^^^(red) - afaik Correlations between Salmonella/L.mono/Listeria to indicator species cannot be assumed unless you can (somehow?) Validate such from yr own data. Hence their frequent direct appearance in zonal analytical schemes.


Kind Regards,

 

Charles.C


#3 kfromNE

kfromNE

    Grade - PIFSQN

  • IFSQN Principal
  • 501 posts
  • 167 thanks
122
Excellent

  • United States
    United States
  • Gender:Female
  • Interests:Bicycling, reading, nutrition, trivia

Posted 18 March 2021 - 04:21 PM

For the FDA and environmental monitoring. The guidance document about listeria will be your best bet. Personally - you are already sampling listeria and salmonella, why test for indicator organisms as well. I don't see the need. What's your purpose to test them?

 

https://www.fda.gov/...ready-eat-foods



#4 drquinn1964

drquinn1964

    Grade - Active

  • IFSQN Associate
  • 6 posts
  • 3 thanks
2
Neutral

  • United States
    United States
  • Gender:Male
  • Location:Galt, CA

Posted 18 March 2021 - 07:02 PM

Thank you Charles and kfromNE.  The product I listed is baked and is frozen during the stabilization process and maintained in a frozen state throughout finished product storage and distribution as well as being labeled "Keep Frozen".  While reviewing multiple post FSMA EMP guidance documents and powerpoint type presentations I found that many of them described ongoing sampling in each zone containing a certain percentage of the total number of samples.  Yet none provided any particular guidance on how to determine the total number of samples needed when starting an EMP, when no sampling and testing occurred previously, so that one could use that data in a statistically supportable way to determine ongoing, not initial, sampling frequency and number of samples taken each time.

 

Reading your comments it sounds like I don't need the initial data to determine ongoing sampling frequency and number of samples taken, I should instead just use the guidance provided in the FDA 2017 Control of Listeria Monocytogenes in Ready-To-Eat Foods: Guidance for Industry on pages 36 and 37 to collect at least 5 FCS and 5 non-FCS samples using the recommended lowest frequency, monthly, of routine sample collection because the food produced is frozen and does not support the growth of L. mono.  For Zones 2, 3, and 4 I would then sample at least 5 non-FCS locations from each zone as well.

 

Regarding why I have designed the initial sampling plan to include both Salmonella and indicator organisms such as TPC, EB, and coliform, that comes from my personal experience in FSIS, from Dr. Marshal at Eurofins, and from the CA Almond Board PEM Guidance.  Based upon the risk of producing a RTE product for which preventive controls, including sanitation and process controls, are required to produce a safe product, the facility is required to include sampling and testing for the hazard analysis identified pathogens at a minimum.  My experience in FSIS also informs me that if one only tests for pathogens but leaves out indicators of sanitation practices (cleaning and employee hygiene), then your program is just like an underground coal mine without canries, just waiting for people to succumb to dangerous gases.  Granted Eurofins and Dr. Marshall are in the business of selling testing, but as microbiologists with many yers of experience investigating and eradicating pathogens from products and FCS they do have credibility.  The Eurofins EMP White Paper states the following goals. Primary: to find pathogens in the environment before they contaminate product. Secondary: to find spoilage microorganism in the environment before they affect product.  Tertiary: to assess effectiveness of cleaning, sanitation, and employee hygiene practices.  The CA Almond board explains that although indicators like coliform and EB do not directly correlate with Salmonella presence because Salmonella can be present when these indicators are negative, they can loosely demonstrate that when their counts are increasing then the percentage of Salmonella positive samples in the environment can dramatically increase.  So I am looking for whether coliform or EB counts provides the most consistent data in this particular facility for the purpose of reducing routine testing costs associated with Salmonella testing and to keep a canary present.  I am using TPC as an indicator of sanitation effectiveness.

 

Since my initial sampling plan for the first four weeks uses the same number of samples per zone as the FDA recommended routine sampling number, but samples at twice per week rather than at the FDA recommended routine sampling frequency of monthly, then my initial sampling plan should yield enough information so that I can correlate either increasing coliform or EB counts with increased Salmonella prevalence.  I do understand that in order to maintain that correlation I will have to consider including Salmonella testing as well as the chose indicator at some less frequent but ongoing level, such as every six months or annually.
 
Thank you again for your guidance.
 
Quinn H.


#5 Charles.C

Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 18,311 posts
  • 5129 thanks
1,129
Excellent

  • Earth
    Earth
  • Gender:Male
  • Interests:SF
    TV
    Movies

Posted 18 March 2021 - 10:22 PM

 

Thank you Charles and kfromNE.  The product I listed is baked and is frozen during the stabilization process and maintained in a frozen state throughout finished product storage and distribution as well as being labeled "Keep Frozen".  While reviewing multiple post FSMA EMP guidance documents and powerpoint type presentations I found that many of them described ongoing sampling in each zone containing a certain percentage of the total number of samples.  Yet none provided any particular guidance on how to determine the total number of samples needed when starting an EMP, when no sampling and testing occurred previously, so that one could use that data in a statistically supportable way to determine ongoing, not initial, sampling frequency and number of samples taken each time.

 

Reading your comments it sounds like I don't need the initial data to determine ongoing sampling frequency and number of samples taken, I should instead just use the guidance provided in the FDA 2017 Control of Listeria Monocytogenes in Ready-To-Eat Foods: Guidance for Industry on pages 36 and 37 to collect at least 5 FCS and 5 non-FCS samples using the recommended lowest frequency, monthly, of routine sample collection because the food produced is frozen and does not support the growth of L. mono.  For Zones 2, 3, and 4 I would then sample at least 5 non-FCS locations from each zone as well.

 

Regarding why I have designed the initial sampling plan to include both Salmonella and indicator organisms such as TPC, EB, and coliform, that comes from my personal experience in FSIS, from Dr. Marshal at Eurofins, and from the CA Almond Board PEM Guidance.  Based upon the risk of producing a RTE product for which preventive controls, including sanitation and process controls, are required to produce a safe product, the facility is required to include sampling and testing for the hazard analysis identified pathogens at a minimum.  My experience in FSIS also informs me that if one only tests for pathogens but leaves out indicators of sanitation practices (cleaning and employee hygiene), then your program is just like an underground coal mine without canries, just waiting for people to succumb to dangerous gases.  Granted Eurofins and Dr. Marshall are in the business of selling testing, but as microbiologists with many yers of experience investigating and eradicating pathogens from products and FCS they do have credibility.  The Eurofins EMP White Paper states the following goals. Primary: to find pathogens in the environment before they contaminate product. Secondary: to find spoilage microorganism in the environment before they affect product.  Tertiary: to assess effectiveness of cleaning, sanitation, and employee hygiene practices.  The CA Almond board explains that although indicators like coliform and EB do not directly correlate with Salmonella presence because Salmonella can be present when these indicators are negative, they can loosely demonstrate that when their counts are increasing then the percentage of Salmonella positive samples in the environment can dramatically increase.  So I am looking for whether coliform or EB counts provides the most consistent data in this particular facility for the purpose of reducing routine testing costs associated with Salmonella testing and to keep a canary present.  I am using TPC as an indicator of sanitation effectiveness.

 

Since my initial sampling plan for the first four weeks uses the same number of samples per zone as the FDA recommended routine sampling number, but samples at twice per week rather than at the FDA recommended routine sampling frequency of monthly, then my initial sampling plan should yield enough information so that I can correlate either increasing coliform or EB counts with increased Salmonella prevalence.  I do understand that in order to maintain that correlation I will have to consider including Salmonella testing as well as the chose indicator at some less frequent but ongoing level, such as every six months or annually.
 
Thank you again for your guidance.
 
Quinn H.

 

 

Hi Quinn,

 

You must have a client with extremely deep pockets.

 

Process/Consumer handling of yr pie product is undetailed and may mitigate comments in following sentence. Some  possible reservations regarding various assumptions in yr 2nd/3rd/4th paragraphs are exampled below  -

 

This [FDA Draft] guidance is not directed to processors of RTE foods that receive a listericidal control measure applied to the food in the final package, or applied to the food just prior to packaging in a system that adequately shields the product and food contact surfaces of the packaging from contamination from the food processing environment.

Product categories considered to be at-risk foods generally allow for the growth of Listeria spp. at some point prior to consumption and generally include:

1.  Refrigerated, perishable foods that are exposed to the plant environment after the final

lethality step. 

2.  Frozen foods exposed to the plant environment after the final lethality step and intended to  be  thawed  for  an  extended  time  prior  to  consumption  (e.g.,  deli  sandwiches,  baked goods, salad ingredients).

3.  Foods produced with no lethality step (e.g., dips, spreads, salads, fresh produce).

 

While the coliform/E. coli and the total Enterobacteriaceae groups are not perfect [!] indicators of the presence of Salmonella spp. in the processing environment, they nevertheless are good indicators of cleaning and sanitation practices. Although there is a lack of data correlating TEB counts and Salmonellae in environmental samples from almond and nut processing operations, data does exist showing the existence of a loose [!] correlation between TEB counts and the occurrence of Salmonellae in environmental samples from a dried-milk processing plant [!] (60):

Attached File  EB vs Salmonella.png   11.86KB   0 downloads

 


Kind Regards,

 

Charles.C


#6 drquinn1964

drquinn1964

    Grade - Active

  • IFSQN Associate
  • 6 posts
  • 3 thanks
2
Neutral

  • United States
    United States
  • Gender:Male
  • Location:Galt, CA

Posted 19 March 2021 - 03:02 AM

Charles,

 

Thank you for the information.  No, my client does not have deep pockets, and that is why I have been looking for guidance on how to develop an initial sampling and testing scheme for the EMP based upon all of the guidance that you have previously posted and others that I have found and read.  If I could find reliable guidance that supports a lower number of samples and frequency both initially and for routine purposes I would use it, but I have not seen it.

 

The product does not receive any listericidal control measure applied to the food in the final package, or applied to the food just prior to packaging in a system that adequately shields the product and food contact surfaces of the packaging from contamination from the food processing environment.  So the FDA Guidance does apply.  Because the product is an at-risk-food is why I am initially testing Non-FCS for both Listeria spp. and Salmonella as well as the three indicator organisms.  I understand that APC is mainly for sanitation effectiveness verification.

 

So the choice that is left will be to determine if either Coliforms or EB correlates better with Salmonella in my client's facility to determine which of those will be used as the Salmonella indicator during routine sampling and testing probably at the once monthly frequency.  I have previously read the information you provided in your 3rd insert and having read Nestle's Micro Specifications it makes sense to me why EB would more closely indicate Salmonella in that environment.  But alas, even though Coliforms and EB have not proven to be reliable indicators of Salmonella presence in a variety of RTE food processing environments, what else is there to use that is less expensive than Salmonella testing itself?

 

Any further ideas are greatly appreciated.

 

Quinn H



#7 Evans X.

Evans X.

    Grade - AIFSQN

  • IFSQN Associate
  • 46 posts
  • 21 thanks
12
Good

  • Greece
    Greece
  • Gender:Male

Posted 19 March 2021 - 10:29 AM

Greetings Quinn H,

 

I think all of the above mentioned, though they are an extremely nice biblio review, are a bit of an overkill and overthinking. Theoretically all are good, but in practice is were your vast experience has to kick-in. One step that I think has been overlooked a bit is that enviromental sampling is not the only thing happening in the facilities. All data deriving from micro analyses should also include the analyses you do in your products, raw materials and all of them together give you the information you need for setting up the sampling plan. Having these in mind, here is a suggestion:

 

For pathogens in the environment your baseline should be formed by a one time all sampling to check if the facilities are free from them, cause if they are not zero, then you already start with a problem. Afterwards you set a lets say monthly or bi-monthly rotation check on your different sampling points, while also sampling (again according to a plan which I guess will be more frequent than the environmental) your product/raw materials. If you have a finding in these then you can go back to a whole zones/areas testing to determine if the contamination derived from them or if they got affected.

Same goes for indicator m/o cause when you set the baseline for them, thought they do not directly correlate with Salmonella, they still act as indicators for bad hygiene or some other "problem" (other m/o induction in the environment) if you find trends forming through time. These also could be tested-for more frequently (bi-weekly or monthly).

To sum it up, all the analyses from environment, product, raw material, water, personnel, air(!) and the trends they produce, will give you the overall sampling plan and its frequency and even the specific frequency of sampling points that are more susceptible to having a problem than others (eg if sampling point X has a higher possibility of contamination, you can test it every time you gather your samples or at least more frequently than others in your rotation). A sampling plan is not a one time made thing, as it can change (even though possibly not very often) according to the production needs and changes that may occur, the results from an analysis etc.

 

Hope it helps!






1 user(s) are reading this topic

1 members, 0 guests, 0 anonymous users