9 replies to this topic

[amberlyda]

Hello, we have been SQF certified for 3 years now.  I have only been the QA manager for the last year.  I have a degree in biology. I have to rewrite our environmental program.  We produce edible cookie dough with no kill step (no eggs and we use heat treated flour) Long story short the previous QA manager wrote our procedures before we got in several different and extended production lines.  The product is exposed to the air for longer periods of time and goes through several machines before being packed.  We have thrown away thousands of dollars of product just based on elevated apc results. I am changing our APC limits since they have increased on a frequent basis.  I have tested all of the equipment and talked to several microbiologist and they concluded that our limits were too low for our product and new production lines.  The FDA gives no guide lines, but many other countries do, so I am citing their guidelines as a basis for ours.  We would like to test our food contact surfaces on a daily basis. I have a hygiena monitor and we currently test ATP for food contact surfaces daily and use their listeria indicator swabs for non food contact surfaces quarterly. I would like to add APC and coliform to this daily morning testing for afternoon results. I do not have a lab on site. This would not replace our 3rd party lab testing on all cookie dough before shipment. I would also send in comparative samples to make sure the hygiena test kit was calibrated correctly on a quarterly basis. I was thinking it would be good to add this testing since we are increasing our allowable apc results on our products. All of my preliminary surface testing for apc/coliform has come back "not detectble" on our surface swabs using the hygenia tester. We test our product before shipping for the following pathogens; APC, coliform and listeria spp.

[Charles.C]

Hello,

 

We have been SQF certified for 3 years now.  I have only been the QA manager for the last year.  I have a degree in biology. I have to rewrite our environmental program.  We produce edible cookie dough with no kill step (no eggs and we use heat treated flour) Long story short the previous QA manager wrote our procedures before we got in several different and extended production lines.  The product is exposed to the air for longer periods of time and goes through several machines before being packed.  We have thrown away thousands of dollars of product just based on elevated apc results. I am changing our APC limits since they have increased on a frequent basis.  I have tested all of the equipment and talked to several microbiologist and they concluded that our limits were too low for our product and new production lines.  The FDA gives no guide lines, but many other countries do, so i am citing their guidelines as a basis for ours.  We would like to test our food contact surfaces on a daily basis.  I have a hygiena monitor and we currently test ATP for food contact surfaces daily and use their listeria indicator swabs for non food contact surfaces quarterly .  I would like to add APC and coliform to this daily morning testing for afternoon results.  I do not have a lab on site.  This would not replace our 3rd party lab testing on all cookie dough before shipment.  I would also send in comparative samples to make sure the hygiena test kit was calibrated correctly on a quarterly basis.

I was thinking it would be good to add this testing since we are increasing our allowable apc results on our products.  All of my preliminary surface testing for apc/coliform has come back "not detectble" on our surface swabs using the hygenia tester.

 

We test our product before shipping for the following pathogens; APC, coliform and listeria spp.

 

Hi amberlyda,

 

Was there a question ?  I guess you are asking for comments.

 

I deduce you are considering use of a microsnap type system.

 

A few comments/queries -

 

(1) Responses will be more likely relevant if you can post some numerical data, eg APC limits / actual results.

 

(2) Source of  guidelines ?

 

(3) Lab results like "not detected" need to be quantitatively interpreted or they may be meaningless.

 

(4)  Current ATP results ? Correlation with micro ?

 

Here are a few microsnap-related articles -

 

 mcs1 - Case_Study-_Environmental_Monitoring_Study_at_Cosmetics_Manufacturer.pdf   298.41KB   47 downloads

 mcs2 - What_do_your_microbiology_test_results_really_mean.pdf   200.95KB   79 downloads

 mcs3 - microbial measurement, the inconvenient truth,2016.pdf   292.38KB   49 downloads

 mcs4 - validation of microsnap for APC.pdf   1.5MB   59 downloads

 

and, more generally -

 

 gen1 - FSMA - Environmental_Monitoring_Self_Assessment.pdf   84.4KB   92 downloads

 gen2 - Env.Hygiene Monitoring, GuideforEHOs.pdf   116.15KB   67 downloads

 gen3 - case study tofu-use of ATP,microbial indicators for hygiene monitoring,2021.pdf   1.95MB   50 downloads

 gen4 - review ATP Monitoring Devices at Measuring Organic Matter.pdf   449.17KB   46 downloads

 

I also thought these reference-included comments were significant -

 

There are no standardized limits for the testing of TVC from surfaces; hence, validations cannot be run using this technology although MicroSnap Total can be used for surface testing but only with guidelines drawn up as conventions in each industry sector.

(mcs4)

 

Overall, the MicroSnap Total method is capable of detecting aerobic bacteria at levels between 100 and >1 000 000 CFU/g in a 7 h incubation period using proper techniques for dilution and by calculating levels of bacteria from dilutions.

(mcs4)

 

Certification bodies such as AOAC typically compare methods over a 3 log cycle within a range of 1000 to 10million cfu per gram of foodstuff and define equivalence as ≤0.5 log cfu difference ( i.e.  a 3 fold difference  in CFU  is acceptable). Validation of  enumeration methods for environmental samples is much more difficult because the samples contain fewer organisms (<10 – 500 cfu) and the variation from all sources above is even greater

(mcs3)

 

PS - Some international micro guidelines for food contact surfaces are compiled/reviewed at link (2012) below. Variations are "Large" .

 

http://www.ifsqn.com...ces/#entry60958

[amberlyda]

Hi amberlyda,

 

Was there a question ?  I guess you are asking for comments.

 

I deduce you are considering use of a microsnap type system.

 

A few comments/queries -

 

(1) Responses will be more likely relevant if you can post some numerical data, eg APC limits / actual results.

 

(2) Source of  guidelines ?

 

(3) Lab results like "not detected" need to be quantitatively interpreted or they may be meaningless.

 

(4)  Current ATP results ? Correlation with micro ?

 

Here are a few microsnap-related articles -

 

mcs1 - Case_Study-_Environmental_Monitoring_Study_at_Cosmetics_Manufacturer.pdf

mcs2 - What_do_your_microbiology_test_results_really_mean.pdf

mcs3 - microbial measurement, the inconvenient truth,2016.pdf

mcs4 - validation of microsnap for APC.pdf

 

and, more generally -

 

gen1 - FSMA - Environmental_Monitoring_Self_Assessment.pdf

gen2 - Env.Hygiene Monitoring, GuideforEHOs.pdf

gen3 - case study tofu-use of ATP,microbial indicators for hygiene monitoring,2021.pdf

gen4 - review ATP Monitoring Devices at Measuring Organic Matter.pdf

 

I also thought these reference-included comments were significant -

 

 

 

 

PS - Some international micro guidelines for food contact surfaces are compiled/reviewed at link (2012) below. Variations are "Large" .

 

http://www.ifsqn.com...ces/#entry60958

Sorry, long weekend.  I could have formed that better.  I have looked up the guidelines for unbaked ready to eat food guide lines from: London HPA, Canada Alberta health services for micro guidelines for ready to eat foods, AU/ New Zealand RTE guidelines,  NSW Food authority in reguard to RTE foods, IFSA world SAfety guidelines for RTE foods,  USA DOD  national advisory committee for micro criteria for foods.  USA FDA does not give any guidelines for APC counts and only loosely guides us on enterobacteria testing criteria.

 

My previous QA manager had set the APC limits at 10,000 cfu/g, and 100 cfu/g for coliform, listeria spp negative.  this is done on our finished product testing before shipping.

 

My usual APC numbers are under 2,000 cfu/g with this new line, but we are jumping all over the place with no identifiable cause.  I have had results at 34,000, 50,000, 77,000.  When i have the samples retested the generally go down.  I have had the same sample test at 77,000cfu/g and the retest came back at 400cfu/g.  Generally with these results coliform has not tested over 10 cfu/g.  So there is no other cofactors.  With this new line the product is exposed to the air for longer periods of time and touches more surfaces.  With our current guidelines i have had to throw away well over $20,000 in finished product since february.

 

We ATP test after cleaning at night. and test and do visual inspections before operating in the morning.

 

I am looking to adjust the APC guidelines based on other countries guidelines up to 100,000 cfu/g (other countries even allow 1,000,000 cfu/g for this type of product).  I am doing a tiered approach of <10,000=ideal, >50,000-<100,000=satisfactory and over 100,000 would be unsatisfactory and retesting and additional testing would need to be done while the product is on hold.

 

Because i am changing the guidelines for our company I was thinking of adding APC and coliform testing to our pre op routine along with our current ATP.  The testing with hygienas unit only gives me a qualitative result for APC and coliform in RLUs.  this daily testing would not take the place of our 3rd party testing on product samples which gives us a quantitative result.  I would have the 3rd party lab verify on a quarterly basis that our hygiena unit matched their qualitative results in their lab

 

My thought for adding the extra testing to our Environmental program was basically to cover us for SQF certification because i am changing the testing limits.  I am being super critical of our program before i go live with it because we do have an RTE product with no kill step before it reaches the consumer.

 

Im just looking for people to pick this apart.  We are growing so fast as a company and Im by myself in this department management wise and have a lot of changes i have to make in reguards to our policies and procedures

[Charles.C]

Sorry, long weekend.  I could have formed that better.  I have looked up the guidelines for unbaked ready to eat food guide lines from: London HPA, Canada Alberta health services for micro guidelines for ready to eat foods, AU/ New Zealand RTE guidelines,  NSW Food authority in reguard to RTE foods, IFSA world SAfety guidelines for RTE foods,  USA DOD  national advisory committee for micro criteria for foods.  USA FDA does not give any guidelines for APC counts and only loosely guides us on enterobacteria testing criteria.

 

My previous QA manager had set the APC limits at 10,000 cfu/g, and 100 cfu/g for coliform, listeria spp negative.  this is done on our finished product testing before shipping.

 

My usual APC numbers are under 2,000 cfu/g with this new line, but we are jumping all over the place with no identifiable cause.  I have had results at 34,000, 50,000, 77,000.  When i have the samples retested the generally go down.  I have had the same sample test at 77,000cfu/g and the retest came back at 400cfu/g.  Generally with these results coliform has not tested over 10 cfu/g.  So there is no other cofactors.  With this new line the product is exposed to the air for longer periods of time and touches more surfaces.  With our current guidelines i have had to throw away well over $20,000 in finished product since february.

 

We ATP test after cleaning at night. and test and do visual inspections before operating in the morning.

 

I am looking to adjust the APC guidelines based on other countries guidelines up to 100,000 cfu/g (other countries even allow 1,000,000 cfu/g for this type of product).  I am doing a tiered approach of <10,000=ideal, >50,000-<100,000=satisfactory and over 100,000 would be unsatisfactory and retesting and additional testing would need to be done while the product is on hold.

 

Because i am changing the guidelines for our company I was thinking of adding APC and coliform testing to our pre op routine along with our current ATP.  The testing with hygienas unit only gives me a qualitative result for APC and coliform in RLUs.  this daily testing would not take the place of our 3rd party testing on product samples which gives us a quantitative result.  I would have the 3rd party lab verify on a quarterly basis that our hygiena unit matched their qualitative results in their lab

 

My thought for adding the extra testing to our Environmental program was basically to cover us for SQF certification because i am changing the testing limits.  I am being super critical of our program before i go live with it because we do have an RTE product with no kill step before it reaches the consumer.

 

Im just looking for people to pick this apart.  We are growing so fast as a company and Im by myself in this department management wise and have a lot of changes i have to make in reguards to our policies and procedures

 

Hi amberlyda,

 

I would have thought that qualitative results for APC/coliform are of limited value.

Microsnap is quantitative and, equipment-wise  essentially only requires an incubator. Although I note that it, for unspecified reasons, incubates at 30degC rather than, afaik, the more usual US range of 35-37degC.

 

Hopefully each lot is assessed with multiple samples. Relying on 1 sample can be a recipe for disaster.

 

^^^(red) Suggests that either (a) lab is unreliable, (b) sample source is highly non-homogenenous. Or both.

 

If you have had a previous extended history of satisfactory results. logically something significant must have changed. Obviously lack of a thermal elimination step accentuates any increased micro. growth. (Assuming main ingredient is heat-processed flour, the limited micro. info I cud find does suggest a target of 10,000-50,000 cfu/gram).

Offhand, I predict that yr unusually "high" values of APC are due to one or more of (a) raw materials, (b) excessive holding periods at elevated temperatures (eg >10degC), (c) cross-contamination. The typical investigative micro. procedure  is to take samples starting from 1st stage and working forward.

I would also consider trying another lab with duplicate samples.

 

APC and Coliform are not pathogens and not directly related to Food Safety. I'm rather surprised that you don't test finished product for (generic/pathogenic)E.coli, Salmonella, S.aureus, B.cereus.

 

I assume yr ATP results for food contact surfaces are satisfactory. Microbial surface assessment will require data other than just Listeria.

[amberlyda]

Hi amberlyda,

 

I would have thought that qualitative results for APC/coliform are of limited value.

Microsnap is quantitative and, equipment-wise  essentially only requires an incubator. Although I note that it, for unspecified reasons, incubates at 30degC rather than, afaik, the more usual US range of 35-37degC.

 

Hopefully each lot is assessed with multiple samples. Relying on 1 sample can be a recipe for disaster.

 

^^^(red) Suggests that either (a) lab is unreliable, (b) sample source is highly non-homogenenous. Or both.

 

If you have had a previous extended history of satisfactory results. logically something significant must have changed. Obviously lack of a thermal elimination step accentuates any increased micro. growth. (Assuming main ingredient is heat-processed flour, the limited micro. info I cud find does suggest a target of 10,000 cfu/gram).

Offhand, I predict that yr unusually "high" values of APC are due to one or more of (a) raw materials, (b) excessive holding periods at elevated temperatures (eg >10degC), (c) cross-contamination. The typical investigative micro. procedure  is to take samples starting from 1st stage and working forward.

I would also consider trying another lab with duplicate samples.

 

APC and Coliform are not pathogens and not directly related to Food Safety. I'm rather surprised that you don't test finished product for (generic/pathogenic)E.coli, Salmonella, S.aureus, B.cereus.

 

I assume yr ATP results for food contact surfaces are satisfactory. Microbial surface assessment will require data other than just Listeria.

Our samples are homogenous.  We batch samples testing to reduce cost.  So roughly 5 batches/bowls are mixed together per sample tested.  12 pieces of dough are taken from each batch and 3 of those pieces are chosen at random and mixed with the other batch samples.  We will have over 20 batches a day per lot. I have been told by the lab that This does lead to inconsistent results.  When we have elevated coliform counts over >100cfu/g we do more specific testing with the lab for ecoli and salmonella (chocolate).  The only thing that has changed has been the introduction of the new line.  Our dough is mixed and held at room temperature for up to 30 min before being cut and the dough is transferred to many other conveyors, scales etc before its final packaging.  It used to be made and immediately go into a cutter/nitrogen tunnel and hand packed immediately after the nitrogen tunnel.

We do get COAs from our raw goods and our chocolate does allow for up to 20,000 cfu/g for APC

We test our food contact surfaces pre and post op.  If anything tests higher than 30RLU it has to be recleaned.

Ive sent multiple swab samples to the lab for each part of the new production line and had it tested for apc and coliform and they keep coming back <10 cfu/g.  So as far as testing it seems that our equipment is clean.  

 

Here is one example of the guidelines I could find for APC counts relative to our product.  Our ingredients are (heat treated flour, margarine, sugar, chocolate chips, powdered milk, vanilla, water, peanut butter, sprinkles, salt, oil, and oreos---not all together).  Products is mixed, frozen and packed.  Guidelines for our specific product dont seem to exist anywhere and I even had one microbiologist say that APC counts would be unreliable because  of the lack of kill steps.  With the product samples listeria has been negative and coliform is less than 10 cfu/g.

 

https://www.cfs.gov...._for_Food_e.pdf (page 14 food category 6)

https://mb-labs.com/...ic-plate-count/

 

Our testing is based on the fact that our raw goods come with COAs, they are low risk, we dont use eggs etc.  These were the guidelines given to me when i took over. its so frustrating and nerve racking as im trying to update things

[kettlecorn]

Have you considered third-party testing representative samples of your raw materials? It is great that you have COAs for them, but a good place to start might be randomly testing some lots of your raw materials to see whether the third-party results conform to the COAs. If they don't, then no amount of fine-tuning on your end will fix them. I've seen third-party test results of raw goods diverge widely (for whatever reason, and the statistics and ethics of this can get complex) from the received COA. 

 

Since most of the problems have emerged since you changed your productions lines and exposure you air, I wonder whether you might try some air testing at various sites around your facility. We produce a RTE product too, and air quality and air flow makes quite a difference in test results. 

 

It appears to me you are heading in a solid direction. 

[Charles.C]

Our samples are homogenous.  We batch samples testing to reduce cost.  So roughly 5 batches/bowls are mixed together per sample tested.  12 pieces of dough are taken from each batch and 3 of those pieces are chosen at random and mixed with the other batch samples.  We will have over 20 batches a day per lot. I have been told by the lab that This does lead to inconsistent results.  When we have elevated coliform counts over >100cfu/g we do more specific testing with the lab for ecoli and salmonella (chocolate).  The only thing that has changed has been the introduction of the new line.  Our dough is mixed and held at room temperature for up to 30 min before being cut and the dough is transferred to many other conveyors, scales etc before its final packaging.  It used to be made and immediately go into a cutter/nitrogen tunnel and hand packed immediately after the nitrogen tunnel.

We do get COAs from our raw goods and our chocolate does allow for up to 20,000 cfu/g for APC

We test our food contact surfaces pre and post op.  If anything tests higher than 30RLU it has to be recleaned.

Ive sent multiple swab samples to the lab for each part of the new production line and had it tested for apc and coliform and they keep coming back <10 cfu/g.  So as far as testing it seems that our equipment is clean.  

 

Here is one example of the guidelines I could find for APC counts relative to our product.  Our ingredients are (heat treated flour, margarine, sugar, chocolate chips, powdered milk, vanilla, water, peanut butter, sprinkles, salt, oil, and oreos---not all together).  Products is mixed, frozen and packed.  Guidelines for our specific product dont seem to exist anywhere and I even had one microbiologist say that APC counts would be unreliable because  of the lack of kill steps.  With the product samples listeria has been negative and coliform is less than 10 cfu/g.

 

https://www.cfs.gov...._for_Food_e.pdf (page 14 food category 6)

https://mb-labs.com/...ic-plate-count/

 

Our testing is based on the fact that our raw goods come with COAs, they are low risk, we dont use eggs etc.  These were the guidelines given to me when i took over. its so frustrating and nerve racking as im trying to update things

 

Hi amberlyde,

 

Suggestions in Post 6 are mostly commendable IMO.

 

I appreciate micro.testing may not be cheap. but neither is rejecting product. Seems to me this is a quite complex and  "sensitive"  Product/Process from a micro. POV. I deduce you regard the finished product as "Low Risk". Such tags are subjective but personally I would  consider it "High Risk."

 

IMO, for APC, it would be preferable to analyse more samples rather than doing the extensive blending/reduction you describe and relying on one result. The former enables an evaluation of variations within the Lot plus micro data is typically considered  to be associated with  considerable inaccuracies. Systems like nmMc were developed/promoted to allow for microbiological data limitations. In contrast, for items like Salmonella where detection is the only criterion, some blending is considered a standard analytical Procedure.

 

I would also do some routine micro. testing of input materials. Trust and verify.

 

For Standards of undocumented products, "quality" items like APC may initially have to be based on what it is operationally achievable( or customer requirements of course). FS-specific parameters can usually be more readily designated via established/Regulatoryl levels.

 

IMO, "coliform" is not a particularly meaningful item to use for indicator purposes in comparison to, say,  (generic) E.coli.

 

It would be useful to do some micro.testing before/after the holding delay you mention.

 

I am not familiar this kind of Product/Process but the category you quoted seemed rather dissimilar to me ? (1st link is mainly a "revamp" of UK"s recommendations).

 

No direct experience of microsnap option but it does seem to offer a useful quantitative,  in-house, rapid,  evaluation of APC of food contact surfaces. And potentially other micro. items. (Whether it is worth initiating a detailed in-house micro option is a more complicated issue, involves investment in equipment, personnel, etc and likely demands substantial throughput to be economically viable).

[amberlyda]

Have you considered third-party testing representative samples of your raw materials? It is great that you have COAs for them, but a good place to start might be randomly testing some lots of your raw materials to see whether the third-party results conform to the COAs. If they don't, then no amount of fine-tuning on your end will fix them. I've seen third-party test results of raw goods diverge widely (for whatever reason, and the statistics and ethics of this can get complex) from the received COA. 

 

Since most of the problems have emerged since you changed your productions lines and exposure you air, I wonder whether you might try some air testing at various sites around your facility. We produce a RTE product too, and air quality and air flow makes quite a difference in test results. 

 

It appears to me you are heading in a solid direction. 

Thank you.  I was considering adding raw goods testing to our more sensitive ingredients like flour, chocolate, cocoa etc.  I was thinking of doing this on a quarterly basis to verify  COAs.  Do you think this would be frequent enough? I have tested the lab air, compressed, fans etc and they have all come back "clean".  Our factory is in TExas and as the humidity rises we see more spikes in APC.  We have air filtration but in an older building its hard to combat when its 90 degrees outside and pouring rain:)

[amberlyda]

Hi amberlyde,

 

Suggestions in Post 6 are mostly commendable IMO.

 

I appreciate micro.testing may not be cheap. but neither is rejecting product. Seems to me this is a quite complex and  "sensitive"  Product/Process from a micro. POV. I deduce you regard the finished product as "Low Risk". Such tags are subjective but personally I would  consider it "High Risk."

 

IMO, for APC, it would be preferable to analyse more samples rather than doing the extensive blending/reduction you describe and relying on one result. The former enables an evaluation of variations within the Lot plus micro data is typically considered  to be associated with  considerable inaccuracies. Systems like nmMc were developed/promoted to allow for microbiological data limitations. In contrast, for items like Salmonella where detection is the only criterion, some blending is considered a standard analytical Procedure.

 

I would also do some routine micro. testing of input materials. Trust and verify.

 

For Standards of undocumented products, "quality" items like APC may initially have to be based on what it is operationally achievable( or customer requirements of course). FS-specific parameters can usually be more readily designated via established/Regulatoryl levels.

 

IMO, "coliform" is not a particularly meaningful item to use for indicator purposes in comparison to, say,  (generic) E.coli.

 

It would be useful to do some micro.testing before/after the holding delay you mention.

 

I am not familiar this kind of Product/Process but the category you quoted seemed rather dissimilar to me ? (1st link is mainly a "revamp" of UK"s recommendations).

 

No direct experience of microsnap option but it does seem to offer a useful quantitative,  in-house, rapid,  evaluation of APC of food contact surfaces. And potentially other micro. items. (Whether it is worth initiating a detailed in-house micro option is a more complicated issue, involves investment in equipment, personnel, etc and likely demands substantial throughput to be economically viable).

You are a wealth of information and i greatly appreciate your feedback.  I agree about doing raw goods testing.  Do you think quarterly sample testing would be enough on higher risk ingredients suck as flour, cocoa, chocolate etc?

[Charles.C]

You are a wealth of information and i greatly appreciate your feedback.  I agree about doing raw goods testing.  Do you think quarterly sample testing would be enough on higher risk ingredients suck as flour, cocoa, chocolate etc?

 

Hi amberlyda,

 

TBH my experience has mainly been with an in-house micro.lab which offers substantial financial/volume flexibility compared to external services.

 

I assume you have micro. specifications for the respective ingredients.

 

The sampling frequency logically relates to feedback results but I would attempt to initially obtain  as much relevant, prioritised, data as quickly as (financially/logistically) possible. An initial oversampling may reduce the subsequent requirements.

 

Just as an idea, select highest 3/4 ingredient percentage items to analyse for major micro parameters. Also submit samples from (at least) 2 different lots for each item but include at least one pair for one item taken from the same bag, eg flour (idea is to assess the lab, a dummy known data sample is better if it exists but probably doesn't) . More samples wud be better but the cost rapidly tends to get prohibitive.

Then review the results.