Jump to content

  • Quick Navigation
Photo

HACCP Likelihood Scoring UHT Milk Product

Share this

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

Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 24 October 2019 - 06:45 AM

Dear All,

 

We're scoring risk of Food Safety hazard for UHT Milk Product.

 

At each process step, we are setting a limit of risk for each hazard identified for semi product. I have two related questions:

 

1. We don't know if we should set the limit before the application of a control measure or after the application of a control measure. 

 

For examples: '

 - Before UHT:   TPC: ≤ 1.000 CFU/mL

 - After UHT: TPC: ≤ 1 CFU/mL

 

2. After we have the test results of hazard of semi product, when giving the score of the likelihood, should we compare the result to the semi-product accepted limit or end-product  accepted limit of hazard?

 

Please help consult.

 

Thank you.

Tom


Edited by Dat Hoang, 24 October 2019 - 06:49 AM.


mahantesh.micro

    Grade - MIFSQN

  • IFSQN Member
  • 166 posts
  • 88 thanks
28
Excellent

  • India
    India
  • Gender:Male
  • Location:Bangalore- Karnataka (India)
  • Interests:Food Safety Standards, Specification developments, Procedures
    Other interests are- cricket and indoor games

Posted 24 October 2019 - 07:14 AM

Dear All,

 

We're scoring risk of Food Safety hazard for UHT Milk Product.

 

At each process step, we are setting a limit of risk for each hazard identified for semi product. I have two related questions:

 

1. We don't know if we should set the limit before the application of a control measure or after the application of a control measure. 

 

For examples: '

 - Before UHT:   TPC: ≤ 1.000 CFU/mL

 - After UHT: TPC: ≤ 1 CFU/mL

 

2. After we have the test results of hazard of semi product, when giving the score of the likelihood, should we compare the result to the semi-product accepted limit or end-product  accepted limit of hazard?

 

Please help consult.

 

Thank you.

Tom

As per HACCP, we always set critical limits considering the final product (end product)

Limit should be set after the application of UHT.



Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 20,542 posts
  • 5662 thanks
1,544
Excellent

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

Posted 24 October 2019 - 07:50 AM

Dear All,

 

We're scoring risk of Food Safety hazard for UHT Milk Product.

 

At each process step, we are setting a limit of risk for each hazard identified for semi product. I have two related questions:

 

1. We don't know if we should set the limit before the application of a control measure or after the application of a control measure. 

 

For examples: '

 - Before UHT:   TPC: ≤ 1.000 CFU/mL

 - After UHT: TPC: ≤ 1 CFU/mL

 

2. After we have the test results of hazard of semi product, when giving the score of the likelihood, should we compare the result to the semi-product accepted limit or end-product  accepted limit of hazard?

 

Please help consult.

 

Thank you.

Tom

 

Context is all-important.

 

Yr queries look to be related to FSMA's vision of haccp regarding when there is (or not) a necessity for preventive controls.??

If so, you are unlikely to get definitive answers. I gave up on this aspect several years ago.

 

If you are referring to traditional codex haccp, may be easier to assist. ??

 

or is this iso-haccp where acceptable values are required ??


Kind Regards,

 

Charles.C


Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 24 October 2019 - 02:00 PM

Context is all-important.

 

Yr queries look to be related to FSMA's vision of haccp regarding when there is (or not) a necessity for preventive controls.??

If so, you are unlikely to get definitive answers. I gave up on this aspect several years ago.

 

If you are referring to traditional codex haccp, may be easier to assist. ??

 

or is this iso-haccp where acceptable values are required ??

Hi Charles, it is ISO 22000 - HACCP, acceptable limit is required. We're having problem when to set the limit (before or after application of a specific control measure) and also the counting of off-limit cases: whether to compare it with process limit or end product limit for the scoring of likelihood.



Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 25 October 2019 - 05:20 AM

Hi Charles, it is ISO 22000 - HACCP, acceptable limit is required. We're having problem when to set the limit (before or after application of a specific control measure) and also the counting of off-limit cases: whether to compare it with process limit or end product limit for the scoring of likelihood.

Hello. Anyone please?



pHruit

    Grade - FIFSQN

  • IFSQN Fellow
  • 2,071 posts
  • 849 thanks
536
Excellent

  • United Kingdom
    United Kingdom
  • Gender:Male
  • Interests:Composing/listening to classical music, electronics, mountain biking, science, sarcasm

Posted 25 October 2019 - 08:05 AM

I'm a bit confused by your question and not the most familiar with ISO22000, but presumably you would set a limit for the processed product (e.g. after the UHT treatment), but would also have a specification in place with appropriate limits for the raw / semi finished product before the process?



Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 25 October 2019 - 09:26 AM

I'm a bit confused by your question and not the most familiar with ISO22000, but presumably you would set a limit for the processed product (e.g. after the UHT treatment), but would also have a specification in place with appropriate limits for the raw / semi finished product before the process?

Yes, we have end spec product, but we're setting semi spec product as well to manage the processes. And we dont know if we can use semi spec as std to score likelihood or have to compare to end product spec for the scoring.


Edited by Dat Hoang, 25 October 2019 - 09:26 AM.


Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 20,542 posts
  • 5662 thanks
1,544
Excellent

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

Posted 25 October 2019 - 12:13 PM

Yes, we have end spec product, but we're setting semi spec product as well to manage the processes. And we dont know if we can use semi spec as std to score likelihood or have to compare to end product spec for the scoring.

 

Hi Tom,

 

afaik, yr concept is atypical for iso22000.

 

Note that TPC is not a safety related micro. characteristic.

 

iso-haccp requires the (safety) risk assessment to be carried out for each process step in respect to the specific safety hazard . The risk (and justification of any associated CCP/OPRP) is assessed on the basis of the hazard's safety consequence to the final consumer at point of consumption.

 

iso22000 also requires the acceptable level related to the specific hazard to be specified.

 

A detailed example (for iso22000:2005) is given here -

 

http://www.ifsqn.com...ge-7#entry50651

 

Not sure if that answers yr question or not ?


Kind Regards,

 

Charles.C


Ryan M.

    Grade - FIFSQN

  • IFSQN Fellow
  • 1,326 posts
  • 479 thanks
290
Excellent

  • United States
    United States
  • Gender:Male
  • Location:Birmingham, AL
  • Interests:Reading, crosswords, passionate discussions, laughing at US politics.

Posted 25 October 2019 - 06:30 PM

I'm not familiar with ISO...but why have a spec for in process material?  This is what it sounds like you are asking because you state "semi-product"?  Why not leave it with raw material spec and the finished product spec?  Is this not acceptable under ISO?

 

Additionally, how in the world are you getting your samples of "semi-product" in a UHT system without jeopardizing the sterility?  Are you taking these before the UHT treatment?



Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 26 October 2019 - 05:34 AM

Hi Tom,

 

afaik, yr concept is atypical for iso22000.

 

Note that TPC is not a safety related micro. characteristic.

 

iso-haccp requires the (safety) risk assessment to be carried out for each process step in respect to the specific safety hazard . The risk (and justification of any associated CCP/OPRP) is assessed on the basis of the hazard's safety consequence to the final consumer at point of consumption.

 

iso22000 also requires the acceptable level related to the specific hazard to be specified.

 

A detailed example (for iso22000:2005) is given here -

 

http://www.ifsqn.com...ge-7#entry50651

 

Not sure if that answers yr question or not ?

 

I'm not familiar with ISO...but why have a spec for in process material?  This is what it sounds like you are asking because you state "semi-product"?  Why not leave it with raw material spec and the finished product spec?  Is this not acceptable under ISO?

 

Additionally, how in the world are you getting your samples of "semi-product" in a UHT system without jeopardizing the sterility?  Are you taking these before the UHT treatment?

Hi Ryan and Charles,

 

We're intending to set specs for semi product, specifically here are micros because we want to control the maximum micro load before a heat treatment to ensure its effectiveness (for e.g. UHT). 

 

However, at the same time, we also want to comply with FSSC/ ISO 22k HACCP methodology. 

 

It seems that Charles' attachment has answered one of my question which is: "Note that "Like.of Occ." refers to the probability of a hazard being present at moment of consumption in the end-product for a manufactured / completed retail product  in the event of temporary non-functioning of the existing (if present) control measure."

 

Our standard has a following sentence.

 

"The likelihood of occurrence of a certain hazard can be linked to the likelihood of occurrence of the situation that is causing the hazard (e.g. metal in end-product caused by breaking equipment parts, likelihood based on historical data that this has taken place and possible occurrence in the future)."

 

According to the std, likelihood can be made on historical data of risk deviation at a process step, that's why we intend to establish spec for semi. But what makes us confused is whether to compare the test result of a specific hazard to the semi-spec or to the end product spec to achieve the likelihood scoring at a specific process step??

 

As Our team is assigned to revamp the HACCP Plan, so there are many points regarding HACCP methodology we need to correct. I also attach our current HACCP UHT Food safety analysis and our corporate's definition for the risk evaluation,  If you see something improper, please help raise.

 

According to the flow chart, we have the following heat treatment of milk:  Thermisation -> Pasteurization  -> UHT which UHT is the final skill step. My question is if test result of micro at Pasteurisation is off, how do we score the likelihood here because we still have the other two heat treatments before it is packaged as the end product. (the likelihood at end product would propably be lower than a single treatment).

Attached Files


Edited by Dat Hoang, 26 October 2019 - 05:45 AM.


Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 27 October 2019 - 10:01 AM

I'm not familiar with ISO...but why have a spec for in process material?  This is what it sounds like you are asking because you state "semi-product"?  Why not leave it with raw material spec and the finished product spec?  Is this not acceptable under ISO?

 

Additionally, how in the world are you getting your samples of "semi-product" in a UHT system without jeopardizing the sterility?  Are you taking these before the UHT treatment?

Hi Ryan, 

 

We need scoring at each step for likelihood, what we do is we compare test result of micro to the spec and give the scores

Yes, we will take the sample before UHT, to test and scoring likelihood of the risk at this step. (specs to compare is the Finished goods specs).



Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 27 October 2019 - 10:05 AM

As per HACCP, we always set critical limits considering the final product (end product)

Limit should be set after the application of UHT.

Hi Mahantesh,

 

Why should we set the limit after the application of UHT (in case we want to control the maximum micro load to enable effective UHT)?



Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 28 October 2019 - 02:18 AM

Hi Tom,

 

afaik, yr concept is atypical for iso22000.

 

Note that TPC is not a safety related micro. characteristic.

 

iso-haccp requires the (safety) risk assessment to be carried out for each process step in respect to the specific safety hazard . The risk (and justification of any associated CCP/OPRP) is assessed on the basis of the hazard's safety consequence to the final consumer at point of consumption.

 

iso22000 also requires the acceptable level related to the specific hazard to be specified.

 

A detailed example (for iso22000:2005) is given here -

 

http://www.ifsqn.com...ge-7#entry50651

 

Not sure if that answers yr question or not ?

Dear Charles,

 

Thanks for the shared file, it is useful. I have some related question.

 

1. The likelihood low medium high of each hazard, what do you base on to give the decision for likelihood? (for e.g. data testing in end product?)

 

2. In our case, we have set accepted limit for semi product, is it ok if we score the likelihood by comparing test result at a specific step with the  semi spec limit at the step as well?

 

3. Why are TPC not a safety related micro characteristic (if ecoli is in them for example)?


Edited by Dat Hoang, 28 October 2019 - 02:20 AM.


Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 20,542 posts
  • 5662 thanks
1,544
Excellent

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

Posted 28 October 2019 - 04:15 PM

Dear Charles,

 

Thanks for the shared file, it is useful. I have some related question.

 

1. The likelihood low medium high of each hazard, what do you base on to give the decision for likelihood? (for e.g. data testing in end product?)

 

2. In our case, we have set accepted limit for semi product, is it ok if we score the likelihood by comparing test result at a specific step with the  semi spec limit at the step as well?

 

3. Why are TPC not a safety related micro characteristic (if ecoli is in them for example)?

 

Hi Dat Hoang,

 

Thks yr posting yr haccp plan.

 

 I guess this is in draft status. I liked the basic hazard analysis layout

 

The methods used in haccp are subjective and many Procedural variations will be auditorially accepted. The iso22000 Standard IIRC requires a "logical" approach.

 

A few criticisms -

 

(a) Many of the Biological  hazards are not safety-related, ie coliforms, (generic)E.coli, Enterobacteriaceae, Yeast. Similarly for "hair".

(b) Various "hazards" are "generalised", eg thermophilic,thermoresistant bacteria, micro/chemical/physical hazards coming from straw.  IMO the haccp plan requires "specifics" or in the absence of such, some sort of cross-reference.

(c) The decision matrix (cols N-T) as presented is IMO not compliant with the iso22000 Standard.

(d) The categories in Severity matrix are undefined and seem ambiguous, eg basic/moderate/average ??

(e) the statement that likelihood is always assessed prior to (any?) application of the control measure is, semantically, debatable, eg cooking, metal detector ?

 

Nonetheless, it is possible that some auditors will accept the, IMO, deficiences (b,c,e). But probably not that in (a,d).

 

Regarding queries in Post 13 -

 

(1)

 

(i) hazards which are controllable by verified PRPs (eg as listed in iso22002-1) are associated with low likelihoods / low risks,

(ii) "low" likelihood takes (some) priority over "high" severity, ( I agreed most of yr numerical matrix although maybe a bit over-quantitative + no need for "extremely rare" "row)

(iii) I have avoided using frequencies for likelihood categories since I found awkward to apply to zero-tolerant pathogens. But it's optional.

(iv) I generally use methods similar to those in 2nd/3rd paragraphs yr excel box (4.2.2.3) as far as possible. Sometimes there will be subjectivity, IMEX auditors tend to be flexible unless the final result is significantly different to that commonly found in their experience of analogous processes.

 

(2) Sorry, I don't really understand this approach. afaik, the iso criterion is that any significant safety hazard must be controlled so as to comply with the relevant acceptable level.

 

(3) This is textbook haccp, ie some microbial species are classified as pathogens, some are not. The reasons for species pathogenicity take you one stage deeper. :smile:

(you are correct that TPC [and coliforms] are undefined  "mixtures"  but are conventionally regarded as  quality parameters based on majority flora).


Edited by Charles.C, 28 October 2019 - 10:25 PM.
edited

Kind Regards,

 

Charles.C


Thanked by 1 Member:

Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 29 October 2019 - 02:34 AM

Hi Dat Hoang,

 

Thks yr posting yr haccp plan.

 

 I guess this is in draft status. I liked the basic hazard analysis layout

 

The methods used in haccp are subjective and many Procedural variations will be auditorially accepted. The iso22000 Standard IIRC requires a "logical" approach.

 

A few criticisms -

 

(a) Many of the Biological  hazards are not safety-related, ie coliforms, (generic)E.coli, Enterobacteriaceae, Yeast. Similarly for "hair".

 

(b) Various "hazards" are "generalised", eg thermophilic,thermoresistant bacteria, micro/chemical/physical hazards coming from straw.  IMO the haccp plan requires "specifics" or in the absence of such, some sort of cross-reference. => What should we mention to be more specific, for thermoresistant bacteria

 

(c) The decision matrix (cols N-T) as presented is IMO not compliant with the iso22000 Standard. => It is the decision tree of our corporate.

(d) The categories in Severity matrix are undefined and seem ambiguous, eg basic/moderate/average ?? I'll share with you the severity list our company is applying, pls see attachment.

(e) the statement that likelihood is always assessed prior to (any?) application of the control measure is, semantically, debatable, eg cooking, metal detector ? => It does have a requirement stating to apply prior to specific control measures (for e.g: UHT)/

 

Nonetheless, it is possible that some auditors will accept the, IMO, deficiences (b,c,e). But probably not that in (a,d).

 

Regarding queries in Post 13 -

 

(1)

 

(i) hazards which are controllable by verified PRPs (eg as listed in iso22002-1) are associated with low likelihoods / low risks,

(ii) "low" likelihood takes (some) priority over "high" severity, ( I agreed most of yr numerical matrix although maybe a bit over-quantitative + no need for "extremely rare" "row)

(iii) I have avoided using frequencies for likelihood categories since I found awkward to apply to zero-tolerant pathogens. But it's optional. => Why do you see it awkward to apply to zero tolerant pathogens?

(iv) I generally use methods similar to those in 2nd/3rd paragraphs yr excel box (4.2.2.3) as far as possible. Sometimes there will be subjectivity, IMEX auditors tend to be flexible unless the final result is significantly different to that commonly found in their experience of analogous processes.

 

(2) Sorry, I don't really understand this approach. afaik, the iso criterion is that any significant safety hazard must be controlled so as to comply with the relevant acceptable level. => It means that we are setting specs for semi product, including micro specs. Because we think with se-mi spec we can better control the processes. The semi specs are developed based on the end spec we expect in end product, and the combinations of control measures esp for micro risk. What do you think?

 

(3) This is textbook haccp, ie some microbial species are classified as pathogens, some are not. The reasons for species pathogenicity take you one stage deeper. :smile:

(you are correct that TPC [and coliforms] are undefined  "mixtures"  but are conventionally regarded as  quality parameters based on majority flora).

Dear Charles,

 

Thanks for your response, pls see my feedback in Blue.

 

Thank you,

Tom

Attached Files


Edited by Dat Hoang, 29 October 2019 - 02:35 AM.


Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 20,542 posts
  • 5662 thanks
1,544
Excellent

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

Posted 29 October 2019 - 05:00 AM

Dear Charles,

 

Thanks for your response, pls see my feedback in Blue.

 

Thank you,

Tom

 

Hi Tom,

 

Thks for yr interesting attachment.

 

The stated qualitative ranking terminologies for severity are IMEX unusual however this is subjective. I deduce that in practice the quantitative interpretations are intended to be used. Such a system is nice in that it enables specificity but it also has disadvantages, for example  many haccp users will not accept the ranking Principle illustrated for allergens since, although technically defensible, it is not afaik legal in most of the World where zero tolerance is applied other than for SO2 and gluten.

 

I note that the thermophilic/thermoresistant entry only refers to an "indicator" so IMO is not relevant to the haccp plan. Similarly Enterobacteriaceae as I previously noted. You may also note the absence of TPC and coliforms, I'm unsure what the terminology "E.coli other" means, perhaps equivalent to my (generic) E.coli so another indicator. Note that E.coli O157 is not the only pathogenic species of E.coli. Additionally certain Vibrio species can be of pathogenic importance, ie V.cholerae, V.parahaemolyticus

 

Most published haccp plans avoid this type of quantitative ranking system since it tends to be arguable but I appreciate you may be obliged to use it.

 

(b) Various "hazards" are "generalised", eg thermophilic,thermoresistant bacteria, micro/chemical/physical hazards coming from straw.  IMO the haccp plan requires "specifics" or in the absence of such, some sort of cross-reference. => What should we mention to be more specific, for thermoresistant bacteria

I'm not familiar with this process so I have no idea, sorry. Maybe it's detailed in the documentation referred. Also see my comment above.

 

(c) The decision matrix (cols N-T) as presented is IMO not compliant with the iso22000 Standard. => It is the decision tree of our corporate. I suggest you study the ISO22000 standard + iso22004 and you will see what I mean.

(d) The categories in Severity matrix are undefined and seem ambiguous, eg basic/moderate/average ?? I'll share with you the severity list our company is applying, pls see attachment.. Thanks.

(e) the statement that likelihood is always assessed prior to (any?) application of the control measure is, semantically, debatable, eg cooking, metal detector ? => It does have a requirement stating to apply prior to specific control measures (for e.g: UHT)/ for example, metal detectors are typically  in operation before you estimate the likelihood ??? Tthe hazard is that the MD fails.  And similarly for cooking. Both are sort of debatable Likelihoods but conventionally accepted to be significant hazards in most haccp plans. Note that a  MD will probably not be classified as a CCP from yr risk matrix

 

Nonetheless, it is possible that some auditors will accept the, IMO, deficiences (b,c,e). But probably not that in (a,d).

 

Regarding queries in Post 13 -

 

(1)

 

(i) hazards which are controllable by verified PRPs (eg as listed in iso22002-1) are associated with low likelihoods / low risks,

(ii) "low" likelihood takes (some) priority over "high" severity, ( I agreed most of yr numerical matrix although maybe a bit over-quantitative + no need for "extremely rare" "row)

(iii) I have avoided using frequencies for likelihood categories since I found awkward to apply to zero-tolerant pathogens. But it's optional. => Why do you see it awkward to apply to zero tolerant pathogens? Consider your risk matrix, a monthly detection of salmonella would probably be financially disastrous, ie equivalent to "very often", not "regular" !  Frequencies need to be interpreted.

(iv) I generally use methods similar to those in 2nd/3rd paragraphs yr excel box (4.2.2.3) as far as possible. Sometimes there will be subjectivity, IMEX auditors tend to be flexible unless the final result is significantly different to that commonly found in their experience of analogous processes.

 

(2) Sorry, I don't really understand this approach. afaik, the iso criterion is that any significant safety hazard must be controlled so as to comply with the relevant acceptable level. => It means that we are setting specs for semi product, including micro specs. Because we think with se-mi spec we can better control the processes. The semi specs are developed based on the end spec we expect in end product, and the combinations of control measures esp for micro risk. What do you think?

I would like to see an example to understand what you mean, the one in Post 1 is not meaningful since TPC is not a safety hazard.

 

 

Interesting discussion. Thks.


Kind Regards,

 

Charles.C


mahantesh.micro

    Grade - MIFSQN

  • IFSQN Member
  • 166 posts
  • 88 thanks
28
Excellent

  • India
    India
  • Gender:Male
  • Location:Bangalore- Karnataka (India)
  • Interests:Food Safety Standards, Specification developments, Procedures
    Other interests are- cricket and indoor games

Posted 30 October 2019 - 04:50 AM

Hi Mahantesh,

 

Why should we set the limit after the application of UHT (in case we want to control the maximum micro load to enable effective UHT)?

Coz UHT is your control measure for biological hazard right??. and also as i said limits applied to your finished product, not the semi finished coz it is finished product that we consume not the semi finished one.

Purpose of UHT is to reduce microbial load, before UHT the microbial load will vary so you cannot fix critical limit, You can fix microbial count limits at receiving stage that bcomes your acceptance criteria. But after the application of UHT microbial load in the product will be almost uniform in count (may vary a bit) and of course you can set critical limit here and it can be made as your CCP



Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 20,542 posts
  • 5662 thanks
1,544
Excellent

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

Posted 30 October 2019 - 05:53 AM

Hi Dat Hoang/Tom,

 

I just noticed that yr excel, 5x6, decision matrix has a criterion for significant hazards of risk > 4, ie includes both "medium" and "high" risk cells.

 

The cut-off point is indeed arbitrary but IMEX "significant" is usually  interpreted as equivalent to "high".


Kind Regards,

 

Charles.C


Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 30 October 2019 - 06:57 AM

Hi Dat Hoang/Tom,

 

I just noticed that yr excel, 5x6, decision matrix has a criterion for significant hazards of risk > 4, ie includes both "medium" and "high" risk cells.

 

The cut-off point is indeed arbitrary but IMEX "significant" is usually  interpreted as equivalent to "high".

Hi Charles,

 

We consider significant hazards with scoring >= 5, not 4 Charles. 

 

See our decision tree attached. The decision tree actually changes the orders of some HACCP Codex's questions in order to enable the categorisation of oPRP as well, besides CCP.

Attached Files



Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 20,542 posts
  • 5662 thanks
1,544
Excellent

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

Posted 30 October 2019 - 08:23 AM

Hi Charles,

 

We consider significant hazards with scoring >= 5, not 4 Charles. 

 

See our decision tree attached. The decision tree actually changes the orders of some HACCP Codex's questions in order to enable the categorisation of oPRP as well, besides CCP.

 

Hi Tom,

 

Please note I wrote > 4.

Yes, I saw the tree (thks). I suggest you compare it to the requirements of the Standard. Unfortunately it has at least one critical omission. Maybe more.

The lengthy explanative text attached (slightly inaccurately quoted) illustrates how difficult it is to present a usable, fully compliant decision tree, particularly for the 2018 version. Unfortunately the tree itself is supposed to provide a direct definitive answer.

(One of the team who developed iso22000 noted in 2005 that the Codex Tree is not an acceptable basis for implementing iso-haccp. Nonetheless, many "fudged" adaptions of it have been accepted by auditors).

 

PS - If you would like to see a fairly intelligible and more (but not 100%) compliant decision tree (for iso22000:2005) can try this post/attachment -

 

https://www.ifsqn.co...18/#entry138212

(a compilation of various proposed alternative trees/procedures for 2005 version is included in post 30 of above thread)

(2 hopefully near 100% compliant, non-tree, Procedures for 2018 version are included in the above thread, Posts 18,22)

 

Of course, the 2018 version has added some further demands/interpretations

 

PPS - the haccp portion of the iso22000 standard has evolved/changed several  times since its original concept, particularly regarding the OPRP factor. ISO-haccp  is IMO probably the most complex/confusing version of haccp currently in use. Even from the original 2005 version  I suspect many auditors found it necessary to accept any reasonably logical haccp presentation which included those elements of the Standard regarded as most "important". The 2018 version of the Standard has, unfortunately, even further increased the complexity IMO.


Kind Regards,

 

Charles.C


Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 31 October 2019 - 02:39 AM

Hi Tom,

 

Please note I wrote > 4.

Yes, I saw the tree (thks). I suggest you compare it to the requirements of the Standard. Unfortunately it has at least one critical omission. Maybe more.

The lengthy explanative text attached (slightly inaccurately quoted) illustrates how difficult it is to present a usable, fully compliant decision tree, particularly for the 2018 version. Unfortunately the tree itself is supposed to provide a direct definitive answer.

(One of the team who developed iso22000 noted in 2005 that the Codex Tree is not an acceptable basis for implementing iso-haccp. Nonetheless, many "fudged" adaptions of it have been accepted by auditors).

 

PS - If you would like to see a fairly intelligible and more (but not 100%) compliant decision tree (for iso22000:2005) can try this post/attachment -

 

https://www.ifsqn.co...18/#entry138212

(a compilation of various proposed alternative trees/procedures for 2005 version is included in post 30 of above thread)

(2 hopefully near 100% compliant, non-tree, Procedures for 2018 version are included in the above thread, Posts 18,22)

 

Of course, the 2018 version has added some further demands/interpretations

 

PPS - the haccp portion of the iso22000 standard has evolved/changed several  times since its original concept, particularly regarding the OPRP factor. ISO-haccp  is IMO probably the most complex/confusing version of haccp currently in use. Even from the original 2005 version  I suspect many auditors found it necessary to accept any reasonably logical haccp presentation which included those elements of the Standard regarded as most "important". The 2018 version of the Standard has, unfortunately, even further increased the complexity IMO.

Hi Charles,

 

Thanks for your shared Coca Decision Tree, its a good reference, Still we often find its very confusing to distinguish between ISO oPRP and CCP definitions. for example the flexibility of the word "Immediately" for actions taken when there is deviation in question 5.

That's why we developed more specific definition (still align with ISO) about oPRP as following. 

 

oPRP: are PRPs that are typically linked to the production process and are identified by the hazard analysis as essential, in order to control the likelihood of the introduction, survival and/or proliferation of food safety hazards in the product(s) or in the processing environment. OPRPs do not reduce hazards, but safeguard conditions to avoid proliferation or (re)contamination.

OPRPs are points in the production process with a smaller food safety risk or where no measurable limits exist. This means that while OPRPs are essential, immediate corrective action towards the product is not always required.
 
CCP: are normally linked with process parameters where 100% measurement can usually be applied and timely reaction upon deviation is needed. For a CCP loss of control implies potential direct food safety risks.
CCPs often ensure reduction or even full elimination of the hazard.
Timely reaction upon deviation is defined as reaction towards the end product in such a time frame that the affected end product batch can be isolated before release.
 
Besides, using the decision tree is not the final step in categorizing specific control measures.
 
After having categorized the control measure via the decision tree as CCP, we will verify, that the CCP is or can be designed in such a way that the following can be applied:
o Measurable critical limits,
o Appropriate ways of monitoring and,
o Timely corrections in case of failure.
If this is not possible, it must be considered to adjust the control measure or consider to categorize it as OPRP (only in exceptional cases).
 
What do you think?


Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 31 October 2019 - 07:26 AM

Coz UHT is your control measure for biological hazard right??. and also as i said limits applied to your finished product, not the semi finished coz it is finished product that we consume not the semi finished one.

Purpose of UHT is to reduce microbial load, before UHT the microbial load will vary so you cannot fix critical limit, You can fix microbial count limits at receiving stage that bcomes your acceptance criteria. But after the application of UHT microbial load in the product will be almost uniform in count (may vary a bit) and of course you can set critical limit here and it can be made as your CCP

Hi Mahantesh, i dont think restrict the micro load of incoming bacterial is enough. Because during the process, there are times where se-mi products are standing still (due to machine break down, lack of packaging, etc.) and if not properly controlled can lead to failure of heat treatment like pasteurise or UHT. So that's why we set the specs for semi products, especially before heat treatmets like thermization, pasteurization and UHT in order to control the maximum micro load before burning them with heat. What do you think?



mahantesh.micro

    Grade - MIFSQN

  • IFSQN Member
  • 166 posts
  • 88 thanks
28
Excellent

  • India
    India
  • Gender:Male
  • Location:Bangalore- Karnataka (India)
  • Interests:Food Safety Standards, Specification developments, Procedures
    Other interests are- cricket and indoor games

Posted 31 October 2019 - 08:15 AM

Hi Mahantesh, i dont think restrict the micro load of incoming bacterial is enough. Because during the process, there are times where se-mi products are standing still (due to machine break down, lack of packaging, etc.) and if not properly controlled can lead to failure of heat treatment like pasteurise or UHT. So that's why we set the specs for semi products, especially before heat treatmets like thermization, pasteurization and UHT in order to control the maximum micro load before burning them with heat. What do you think?

Ofcourse you can set the limits for semi products but i dont think it should be CCP.



Dat Hoang

    Grade - AIFSQN

  • IFSQN Associate
  • 34 posts
  • 0 thanks
0
Neutral

  • Vietnam
    Vietnam

Posted 31 October 2019 - 08:55 AM

 

Hi Charles,

 

Thanks for your shared Coca Decision Tree, its a good reference, Still we often find its very confusing to distinguish between ISO oPRP and CCP definitions. for example the flexibility of the word "Immediately" for actions taken when there is deviation in question 5.

That's why we developed more specific definition (still align with ISO) about oPRP as following. 

 

oPRP: are PRPs that are typically linked to the production process and are identified by the hazard analysis as essential, in order to control the likelihood of the introduction, survival and/or proliferation of food safety hazards in the product(s) or in the processing environment. OPRPs do not reduce hazards, but safeguard conditions to avoid proliferation or (re)contamination.

OPRPs are points in the production process with a smaller food safety risk or where no measurable limits exist. This means that while OPRPs are essential, immediate corrective action towards the product is not always required.
 
CCP: are normally linked with process parameters where 100% measurement can usually be applied and timely reaction upon deviation is needed. For a CCP loss of control implies potential direct food safety risks.
CCPs often ensure reduction or even full elimination of the hazard.
Timely reaction upon deviation is defined as reaction towards the end product in such a time frame that the affected end product batch can be isolated before release.
 
Besides, using the decision tree is not the final step in categorizing specific control measures.
 
After having categorized the control measure via the decision tree as CCP, we will verify, that the CCP is or can be designed in such a way that the following can be applied:
o Measurable critical limits,
o Appropriate ways of monitoring and,
o Timely corrections in case of failure.
If this is not possible, it must be considered to adjust the control measure or consider to categorize it as OPRP (only in exceptional cases).
 
What do you think?

 

 

Hi Tom,

 

Please note I wrote > 4.

Yes, I saw the tree (thks). I suggest you compare it to the requirements of the Standard. Unfortunately it has at least one critical omission. Maybe more.

The lengthy explanative text attached (slightly inaccurately quoted) illustrates how difficult it is to present a usable, fully compliant decision tree, particularly for the 2018 version. Unfortunately the tree itself is supposed to provide a direct definitive answer.

(One of the team who developed iso22000 noted in 2005 that the Codex Tree is not an acceptable basis for implementing iso-haccp. Nonetheless, many "fudged" adaptions of it have been accepted by auditors).

 

PS - If you would like to see a fairly intelligible and more (but not 100%) compliant decision tree (for iso22000:2005) can try this post/attachment -

 

https://www.ifsqn.co...18/#entry138212

(a compilation of various proposed alternative trees/procedures for 2005 version is included in post 30 of above thread)

(2 hopefully near 100% compliant, non-tree, Procedures for 2018 version are included in the above thread, Posts 18,22)

 

Of course, the 2018 version has added some further demands/interpretations

 

PPS - the haccp portion of the iso22000 standard has evolved/changed several  times since its original concept, particularly regarding the OPRP factor. ISO-haccp  is IMO probably the most complex/confusing version of haccp currently in use. Even from the original 2005 version  I suspect many auditors found it necessary to accept any reasonably logical haccp presentation which included those elements of the Standard regarded as most "important". The 2018 version of the Standard has, unfortunately, even further increased the complexity IMO.

Hi Charles, What is the critical omission you are mentioning?



Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 20,542 posts
  • 5662 thanks
1,544
Excellent

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

Posted 31 October 2019 - 08:56 AM

 

Hi Charles,

 

Thanks for your shared Coca Decision Tree, its a good reference, Still we often find its very confusing to distinguish between ISO oPRP and CCP definitions. for example the flexibility of the word "Immediately" for actions taken when there is deviation in question 5.

That's why we developed more specific definition (still align with ISO) about oPRP as following. 

 

oPRP: are PRPs that are typically linked to the production process and are identified by the hazard analysis as essential, in order to control the likelihood of the introduction, survival and/or proliferation of food safety hazards in the product(s) or in the processing environment. OPRPs do not reduce hazards, but safeguard conditions to avoid proliferation or (re)contamination.

OPRPs are points in the production process with a smaller food safety risk or where no measurable limits exist. This means that while OPRPs are essential, immediate corrective action towards the product is not always required.
 
CCP: are normally linked with process parameters where 100% measurement can usually be applied and timely reaction upon deviation is needed. For a CCP loss of control implies potential direct food safety risks.
CCPs often ensure reduction or even full elimination of the hazard.
Timely reaction upon deviation is defined as reaction towards the end product in such a time frame that the affected end product batch can be isolated before release.
 
Besides, using the decision tree is not the final step in categorizing specific control measures.
 
After having categorized the control measure via the decision tree as CCP, we will verify, that the CCP is or can be designed in such a way that the following can be applied:
o Measurable critical limits,
o Appropriate ways of monitoring and,
o Timely corrections in case of failure.
If this is not possible, it must be considered to adjust the control measure or consider to categorize it as OPRP (only in exceptional cases).
 
What do you think?

 

 

Hi Tom,

 

It seems that iso22000 (2018) has now abandoned the OPRP definition/interpretation which you quote and has moved to a more operational viewpoint  via introducing the terminology “significant hazard” as borrowed from long-established traditional haccp.

I guess this is an attempt to clarify/simplify the, seemingly, permanently confused CCP/OPRP situation and IMO is a massive improvement. Unfortunately, together with this change, the section presenting methods for distinguishing CCPs/OPRPs has been further complicated. Perhaps 2 Steps forward, 1 Step back !

 

JFI here are a few  extracts from  iso22004 (2014 and 2005) –
 

 

Control measures applied at CCPs and as OPRPs are intended to control all the hazards assessed as significant during hazard analysis. Control measures or combinations of control measures applied at CCPs and as OPRPs should be validated to ensure that they are effective. Validation is not required for PRPs.

 

—  control  measures  applied  at  CCPs:  these  control  measures  are  managed  by  a  HACCP  plan.  Such control measures have defined critical limits that can separate acceptable product from potentially unacceptable (unsafe) product. In addition, their implementation can be monitored in a manner that enables detection of any loss of control within a timeframe sufficient to effectively control affected product. Failure to meet critical limits will result in a potentially unsafe product.

—  control measures applied as OPRPs: these control measures cannot be managed by an HACCP plan but can be managed by similar means such as an OPRP plan. In the case of OPRPs, such control measures do not have a critical limit, but should have an action limit or an action criterion, which demonstrates that the OPRP is in control. Failure to meet these established criteria will lead to corrective actions.

 

 

Categorization of control measures: The organization may focus on having as many of the control measures as possible managed by operational PRPs and only a few managed by the HACCP plan, or the opposite.

 

 

Regarding yr “immediately” comment above, a possible, simple, solution is given in the “semi-official” ISO22000 (2005) tree shown on sheet 3 of my previously referred excel. I used similar logic in an expanded self-compiled tree on sheet 4.

 

My overall conclusion is that for the HACCP/CCP/OPRP section of the iso22000 Standard (2005 or 2008),  any approach, decision tree or otherwise, should preferably -

  1. Cover the requirements of the Standard
  2. Be readily comprehendable/implementable
  3. Be “Logical”
  4. Give a direct result.

IMO your decision tree does not fully provide (1, 4) and, possibly due my own limitations, (2).

 

However if your tree works OK for you and audits satisfactorily then you should have no problem. :smile:

 

PS - Re - critical omission, i was particularly referring to the monitoring aspect as discussed above.. Of course if one compares to 2018 there are, subjectively, others.


Kind Regards,

 

Charles.C




Share this


0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users