41 replies to this topic

[jaygil]

We also have a statement from our flour mill that describes how they control for mycotoxins by having their grain suppliers take responsibility for ensuring that the grains are free of mycotoxins. Our recent BRC auditor was satisfied with that. I have a feeling that mycotoxins are a new concept for FDA inspectors, prior to FSMA and HARPC they never asked about it. So in my opinion they have a learning curve here. Those of us who have lived with HACCP have learned about mycotoxins and their hazard analysis and controls.

[mgourley]

We also have a statement from our flour mill that describes how they control for mycotoxins by having their grain suppliers take responsibility for ensuring that the grains are free of mycotoxins. Our recent BRC auditor was satisfied with that. I have a feeling that mycotoxins are a new concept for FDA inspectors, prior to FSMA and HARPC they never asked about it. So in my opinion they have a learning curve here. Those of us who have lived with HACCP have learned about mycotoxins and their hazard analysis and controls.

 

It's my understanding that, taken to the extreme, a statement from your flour mill that the grain supplier ensures the grains are free of mycotoxins is not sufficient. You would have to have documentation from the grain producers showing test results.

 

The logic being if the grain producer did not ensure absence of mycotoxins, and the mill did not test for them, and you do not test for them, there "could be" mycotoxins in the flour.

 

It's a rabbit hole.

 

Marshall

[Charles.C]

Charles,

 

Indeed. But to FFF's post, there is something to say about the FDA's guidance document re cookie baking. While the example they give in Section 6.13 of the document FFF links to, it assumes one product and one temperature and time.

We make several dozens of different types of cookies with various extrusion weights, various bake times and temperature and various finished product moisture.

 

To follow the FDA guidance, I would have to create a hundred or so Food Safety Plans with different parameters. Or, I could just create generic bake step parameters, which seem to defeat the goal of the Preventive Controls FDA want.

 

It's a Catch-22.

 

In the end, I will simply run an oven mole through each cookie oven and determine Log reduction and then put up chart recorders or probes that feed a database contained on a PC. Then have people duly write down oven temps on some schedule.

 

It's really a pointless exercise because if you are making a crunchy cookie, and the finished product after cooling is not crunchy, it goes in the bin. That's a bit of a simplification, but you get the point.

 

Marshall

 

Hi Marshall,

 

Sorry delay, only just saw above post.

 

Somehow i intuitively doubt that the FDA will go along with the visual rejection approach as a standard protcol. Because it looks unscientific unless there is a generic published  validation study somewhere which afaik there isn't. I take yr points about (a) batch versus continuous and (b) the potentially unlimited number of "variable" cookies but qualitative generalisations such as the present one are likely to be automatically regarded as an attempted zigzag unless a credible validatory reference can be pulled out of the hat. Can it ?

 

I don't have access to the FDA's Lathrop paper quoted in 3F's link but i found another analogous study for hamburger buns (2016) which has this comment -

 

In a similar study, Lathrop et al. (25) inoculated peanut butter cookie  dough  with  a  five-serovar  cocktail  of  Salmonella (Tennessee FSL-R8-5221, Tornow FSL-R8-5222, Hartford FSL-R8-5223, Typhimurium FSL-WI-030, and Agona FSL- S5-517)  and  baked  it  at  1778C  for  up  to  15  min.  They reported  that  Salmonella  serovars  were  detectable  in  the peanut butter cookies (after the enrichment) when baked for 14 min; however, Salmonella serovars were not detected in the cookies after 15 min of baking. The longer Salmonella serovar inactivation time observed by Lathrop et al. (25) compared  with  this  study  (9  min)  is  likely  owing  to differences  in  the  food  matrices  used  in  the  respective studies (peanut butter cookies versus bread dough; specifically, differences in fat content and aw values).

The question is how sensitive are the process requirements with changes in composition. And Validation thereof.

 

I also noticed one curious feature in the bun study, the bun internal temperature was reported as an (internal) mean value, not a core datum. This goes against all the text books afaik. Don't know how the FDA reference does it.

 

I haven't played with the AIB tool but I suspect it must be using some "auto" D values for any calculation. Offhand that sounds like an interesting validatory challenge.

 

I presume by mole you mean a datalogger in the cookie. The calculation of log reduction then must presumably assume a D-value again. Hmmmm.

 

@Graingirl,

 

Further details of Marshall's Vulnerability analysis exist elsewhere on this forum. But note that IIRC this procedure was oriented to BRC, not SQF. There are several others also for BRC but SQF have predictably managed to well-confuse the original concept (which IMO was itself a fudge).

 

PS - this cookie paper may also be of some interest -

 

 Cookie composition versus baking temperature profile, 2014.pdf   613.76KB   29 downloads

[mgourley]

Charles,

 

This is the output of the Kill Step Calculator.

 

Marshall

Attached Files

•  Bun Kill Step Calculator.xls   99.5KB   37 downloads

[Charles.C]

Charles,

 

This is the output of the Kill Step Calculator.

 

Marshall

 

Hi Marshall,

 

Thanks for the Example.

 

It's a very well presented/readable website and has expanded since i last looked at it. The Procedures/Examples are very coherenty written IMO.

 

I daresay the integrative methodology / output sheets, given the appropriate z/D values, is/are identical to the excel sheets posted elsewhere on this Forum (mostly borrowed from the meathaccp website). Another visual on-line calculator is on the milk/dairy Organisation's website.

 

However, as per my Post 28 I was unable to find any validatory link/backup to the selection of z/D. Perhaps i missed it. IMEX with seafood, one has to provide the referential support for such values.

I noticed the AIB's proud claim that their methodology is "FSMA - Compliant".  Have all the numerous (I presume)  z/D values "built-in" been officially approved by FDA ?

 

PS - the data logger capability is amazing. I have previously used very long thermocouple lines / manual integration which is operationally exhausting.

I noticed the caveat in the bun procedure regarding the variation allowed for the various temperatures to reach 170degF. In some processes of "small" items I suspect this would not be so easy to get compliance.

 

PPS - I'm also surprised that BRC would accept a basic concept based on Salmonella  inasmuch as I thought UK/FSA are totally locked into   L.monocytogenes as a target micro. criterion for processes such as this. Certainly the case IIRC for hamburgers where US contrarily opted for pathogenic E.coli as the target.

[FurFarmandFork]

I do a Raw Material Vulnerability Assessment that asks some questions and assigns a point value based on the answer. The final point value indicates the level of risk for IA/EMA.

Point values are subjective, but should be meaningful and logical to come up with a realistic risk level. (That's the complicated part).

 

It takes a bit of tweaking, but if you run the model on an ingredient that has a well documented history of IA/EMA, you should come up with a result of at least a Medium, and probably High level of risk.

 

For example;

1. Ingredient Fraud History 

Yes (10)

Possible (5)

No (0)

 

2. Economic Factors

Likely to Occur (5)

May Occur (2)

Not Likely to Occur (0)

 

3. Geographic Origins

Foreign Supplier (3)

Domestic Supplier (0)

 

4. Complexity of Supply Chain

Complex (2)

Simple (0)

 

5. Ease of Access to Material

Ready Access - Materials are unprotected (5)

Minimal Access - Manufacturer may not have a robust Food Defense Program (3)

Unlikely Access - Nature of material or manufacturing. Manufacturer has a robust Food Defense Program (0)

 

6. Nature of the Material

Liquid/Powder - requires blending (1)

Natural/Bulk (0)

 

7. Cost of Materials

High (2)

Medium (1)

Low(0)

 

8. Supply Chain Confidence

Low (1)

High (0)

 

9. Supplier History

Bad (1)

Good(0)

 

10. Ingredient Testing Methods

None (3)

Occasional (1)

Robust (0)

 

11. Existing Controls

Weak (1)

Strong (0)

 

There are 34 possible points here. This is how I have my risk levels broken down:

Low 1-12

Medium 13-20

High 21-34

 

Let's look at how this would work for cinnamon. It has a fraud history (10), it is from a foreign supplier (3) so it's already Medium risk.

 

You can certainly make this as complicated as you wish, but the FDA guy was fine with my approach.

 

HTH,

Marshall

 

Marshall, this is an awesome system! You should re-post this as a separate named thread so that people searching for ingredient risk assessments can more easily find it in the future.

[sruggero]

Thanks for sharing and good job!

[mgourley]

Marshall, this is an awesome system! You should re-post this as a separate named thread so that people searching for ingredient risk assessments can more easily find it in the future.

I can't really claim it as my own work. There was a fairly long thread a year or so ago that covered various approaches to IA/EMA.

 

Charles C. finally came up with a really nice Excel document that has risks and mitigation factors giving a final risk rating.

 

This is basically a result of that thread input into Safe Food 360.

 

Marshall

[mgourley]

Hi Marshall,

 

Thanks for the Example.

 

It's a very well presented/readable website and has expanded since i last looked at it. The Procedures/Examples are very coherenty written IMO.

 

I daresay the integrative methodology / output sheets, given the appropriate z/D values, is/are identical to the excel sheets posted elsewhere on this Forum (mostly borrowed from the meathaccp website). Another visual on-line calculator is on the milk/dairy Organisation's website.

 

However, as per my Post 28 I was unable to find any validatory link/backup to the selection of z/D. Perhaps i missed it. IMEX with seafood, one has to provide the referential support for such values.

I noticed the AIB's proud claim that their methodology is "FSMA - Compliant".  Have all the numerous (I presume)  z/D values "built-in" been officially approved by FDA ?

 

PS - the data logger capability is amazing. I have previously used very long thermocouple lines / manual integration which is operationally exhausting.

I noticed the caveat in the bun procedure regarding the variation allowed for the various temperatures to reach 170degF. In some processes of "small" items I suspect this would not be so easy to get compliance.

 

PPS - I'm also surprised that BRC would accept a basic concept based on Salmonella  inasmuch as I thought UK/FSA are totally locked into   L.monocytogenes as a target micro. criterion for processes such as this. Certainly the case IIRC for hamburgers where US contrarily opted for pathogenic E.coli as the target.

Charles,

 

I know nothing about the science, but I do know AIB spent a few years on this.

 

As far as Salmonella v. L. monocytogenes, I'd guess that there are some raw materials that are capable of carrying Salmonella, so the efficacy of the "kill step" is important.

I'd guess that L. mono is less of a concern because it's not generally present in raw materials (for a bakery, at least) and considered more of an "environmental" pathogen.

 

Marshall 

[Charles.C]

Charles,

 

I know nothing about the science, but I do know AIB spent a few years on this.

 

As far as Salmonella v. L. monocytogenes, I'd guess that there are some raw materials that are capable of carrying Salmonella, so the efficacy of the "kill step" is important.

I'd guess that L. mono is less of a concern because it's not generally present in raw materials (for a bakery, at least) and considered more of an "environmental" pathogen.

 

Marshall 

 

Hi Marshall,

 

I daresay AIB will make their validations available to FDA so as to support their methodology. No news is probably Good News.

 

The argument as to the most "appropriate" target species has been going on for decades.

IIRC the UK prefers L.mono (except where definitely not suitable) due their belief that in most food matrices it is more heat resistant than the likely Salmonella spp. The USA disputes this (which caused a "dispute" over the safety of "rare" UK hamburgers a few years back resulting in a stand-off), for example -

 

Another example of when the scientific support may not need to address the specific pathogen listed in the hazard analysis is for thermally processed products because the lethality reduction of one pathogen can be used as an indicator for another.  The heat resistance of pathogens is known and can be compared across products.  FSIS recommends that establishments use Salmonella as an indicator of lethality because it tends to be more heat resistant than other pathogens.  For example, the lethality treatment of meat jerky should achieve at least a 5.0 log reduction of Salmonella and should also achieve sufficient reductions in the other bacterial pathogens of public health concern (e.g., at least a 5.0 log reduction for E. coli O157:H7 for products containing beef as recommended in the Salmonella Compliance Guidelines for Small and Very Small Meat and Poultry Establishments that Produce Ready-to-Eat (RTE) Products).  In addition, the lethality treatment of meat and poultry jerky should achieve at least a 3.0 log reduction in Listeria monocytogenes (Lm) although a 5.0 log reduction or greater is desirable because it provides an even greater safety margin for ensuring that Lm doesn’t grow during cold storage to detectable levels.  However, FSIS does not expect establishments to validate that their process achieves reduction in E. coli O157:H7 or Lm if it achieves sufficient reductions in Salmonella because Salmonella is considered an indicator of lethality.  Without further scientific support, establishments should not use pathogens other than Salmonella as indicators of lethality.  For example, establishments should not use reductions in Lm to support similar reductions in Salmonella without support that Lm is at least equally as heat resistant as Salmonella.

 

(FSIS 2013)

 

Some other EC countries use even more exotic target species due their high heat resistance. The AIB referenced paper used a cocktail of species.

[FurFarmandFork]

 AIB's proud claim that their methodology is "FSMA - Compliant".

 

 

Uh huh. We need a separate thread to name and shame all companies droppning FSMA claims, fearmongering, and bad information, it's driving me nuts.

 

@Charles, in the heat treatment guidance, it's worth noting that FDA recommends using the appendix 3 bacterial growth and inactivation information as a starting point, they have only provided values for L. monocytogenes and C.botulinum Type B. I take this as a hint that they would prefer you to use Lmono as the target organism in most cases, especially since the required time/temps they provide exceed those provided by FSIS for salmonella.

 

Or if it's not a hint, then inspectors will be looking for it since that's the only guidance they have to inspect against. 

 

I tend to agree with mgourley on the use of salmonella in any low Aw foods though. While I consider listeria ubiquitious in all incoming ingredients (flour included), the heat resistance of salmonella in dry environments is well documented and a good/more likely target for your validated cook. Also a good indicator that they used salmonella in the cookie example.

[Charles.C]

Uh huh. We need a separate thread to name and shame all companies droppning FSMA claims, fearmongering, and bad information, it's driving me nuts.

 

@Charles, in the heat treatment guidance, it's worth noting that FDA recommends using the appendix 3 bacterial growth and inactivation information as a starting point, they have only provided values for L. monocytogenes and C.botulinum Type B. I take this as a hint that they would prefer you to use Lmono as the target organism in most cases, especially since the required time/temps they provide exceed those provided by FSIS for salmonella.

 

Or if it's not a hint, then inspectors will be looking for it since that's the only guidance they have to inspect against. 

 

I tend to agree with mgourley on the use of salmonella in any low Aw foods though. While I consider listeria ubiquitious in all incoming ingredients (flour included), the heat resistance of salmonella in dry environments is well documented and a good/more likely target for your validated cook. Also a good indicator that they used salmonella in the cookie example.

 

Hi 3F,

 

Thks for the interesting comments.

I note that that yr link states the draft content is borrowed from the well-known Fishery Guide (ca 2011). This contains the following -
 

Goal of cooking for most products

 

One reason for cooking products that will not be reduced oxygen packaged is to eliminate vegetative cells of pathogenic bacteria (or reduce them to an acceptable level) that may have been introduced to the process by raw materials or by processing that occurs before the cooking step.

 

Selection of the target pathogen is critical to the effectiveness of cooking. Generally, L. monocytogenes is selected as the target pathogen because it is regarded as the most heat-tolerant, foodborne bacterial pathogen that does not form spores. Cooking processes are not usually designed to eliminate spores of bacterial pathogens.

 

Determining the degree of destruction of the target pathogen is also critical. Generally, a reduction of six orders of magnitude (six logarithms, e.g., from 10³  to 10^-3 ) in the level of contamination is suitable. This is called a 6D process.

 

However FSIS in my previous Post's quote state (presumably referring to beef/poultry meat) -
 

 

FSIS recommends that establishments use Salmonella as an indicator of lethality because it tends to be more heat resistant than other pathogens.

 

Looks like one needs to look up some (equivalent) D values and compare.

 

This 2015 "meat" paper suggests not much difference in heat resistance between S/LM  at 70degC although the specific species of Salmonella could be relevant. Hence the cocktails. Pathogenic E.coli is somewhat less heat resistant.

 

 D values versus temperature for L.mono, Salmonella, Path.E.coli in raw meat.pdf   1.11MB   23 downloads

 

I anticipate that other compilations may show more divergence.

[mgourley]

No idea who chose what bug for what purpose.

 

All I can say was that the FDA inspector was good with the data I showed him and the resultant output.

 

There was never really the "need" for this type of info in the baking industry before. It was self evident that a piece of dough going through a 450 degree oven for 8-12 minutes would not have anything left alive inside.

 

The inspector never mentioned listeria until he asked if we had an environmental program and what we test for.

 

Marshall

[Charles.C]

Hi Marshall,

 

Thks for above posts.

 

So what is the actual status of the "draft" in respect to FSMA's requirements ? Literally JFI ?

 

PS - I assume products internal temperature is <= ca. 100degC ?

 

It's probably got some Bacillus spores in it.

[FurFarmandFork]

 

So what is the actual status of the "draft" in respect to FSMA's requirements ? Literally JFI ?

 

 

Like every other "draft". Mandatory for you, non-mandatory for FDA if it becomes an issue. I recently spoke with a spice manufacturer who went through his FSMA audit and the inspector literally used the flow charts from the food safety plan and the Appendix 1 draft to come up with his own hazard analysis, then audited the FSP against the hazards identified in appendix 1.

[Cravin' Cajun?]

Wow Marshall...it sounds like you and I may have had the same FDA auditor perform our inspections, because everything you said they did was similar, especially the part about taking one of our products being made on the line that day and doing their own hazard analysis.  They used that Guidance Document appendix like it was their "bible", so I had to chuckle when I read your post!

 

Our FDA inspector (tall gentlemen, seafood expert) showed up with 3 other folks...1 from the state, and 2 other inspectors from the FDA.  He explained that these 3 folks had just received his training on how to conduct FSMA/Preventive Control inspections and were going to learn "on the job" so to speak.  I had 4 people here for 4 full days looking us over backwards, forwards, and side-ways...it was exhausting and very hard to keep them in a tight group when going through the facility.  They focused on the same things you listed in your original post, especially food allergens, even though the only allergen in our facility is wheat.  Most of their time over the 4 day inspection was spent on this subject, which I thought was overkill, considering wheat is not going to send anyone into anaphylactic shock and would only be a class II recall, should cross-contamination occur.

 

It was a LONG 4 days, but they didn't find anything and no 483 was issued (thank goodness!).  It seemed to me and our entire management team that they were basically teaching new inspectors at our expense, so be prepared folks!!!

[mgourley]

Was his name Brian?

 

Marshall