14 replies to this topic

[anna898]

Hi All

I have a question as Im a bit confused   and I hope you can explain or advice 

SO for my HACCP I normally do a hazard analysis where I assess the likelihood and severity for each hazard( B/C/P) indicating which risks and steps are high risk , (normally M/M, M/H, H/H) then only these steps I put through the CCP decision tree ( codex) to see which of them will be a CCP others will be controlled by PRP/OPR 

 

Recently an auditor told me that I should put all process steps through the decision tree to see which ones I should control via PRP and which ones will be a CCP.

 

In my understanding the hazard analysis would be pointless if all the steps have to go through decision tree. if a step is managed by PRP and is not significant (L/M, L/H) the decision tree will indicate the same.

Moreover I have seen some haccps for a meat industry where a deliver/intake step  is controlled by PRP (with B= M/H), and in other HACCP it is a CCP.

 

It is a bit confusing and Im not sure If I'm doing it wrong ?

 

 

 

regards 

Anna

[VThiruselvi]

Hi Anna,

 

what you has being doing is correct. some of the risk assessment may combined together with other process flow, hence when to be deciding the PRP or CCP, you have look into the process flow and each individual product. 

[Charles.C]

Hi All

I have a question as Im a bit confused   and I hope you can explain or advice 

SO for my HACCP I normally do a hazard analysis where I assess the likelihood and severity for each hazard( B/C/P) indicating which risks and steps are high risk , (normally M/M, M/H, H/H) then only these steps I put through the CCP decision tree ( codex) to see which of them will be a CCP others will be controlled by PRP/OPR 

 

Recently an auditor told me that I should put all process steps through the decision tree to see which ones I should control via PRP and which ones will be a CCP.

 

In my understanding the hazard analysis would be pointless if all the steps have to go through decision tree. if a step is managed by PRP and is not significant (L/M, L/H) the decision tree will indicate the same.

Moreover I have seen some haccps for a meat industry where a deliver/intake step  is controlled by PRP (with B= M/H), and in other HACCP it is a CCP.

 

It is a bit confusing and Im not sure If I'm doing it wrong ?

 

 

 

regards 

Anna

Hi Anna,

 

Yr difficulties arise from variations in opinions within the HACCP  "community".

 

Yr OP contains queries on 2 topics -

 

(1) How to determine CCPs,

(2) How to determine PRPs.

 

The queries also overlap difference between  "systems" of HACCP, eg, Codex, NACMCF, ISO, FSMA.

 

Regarding (1) Codex, NACMCF,,ISO support yr "basic" procedure. However various standard textbooks/Literature publications use alternative methods such as you refer. Presumably auditors may similarly  also differ depending on their own training histories. afaik, GFSI-recognised FS Standards typically claim to "base" their haccp Principles on Codex (probably following GFSI itself ?).

 

Regarding (2) Codex, NACMCF, ISO support defining PRPs in advance of the hazard analysis which thereby predetermines related likelihoods in the subsequent  hazard analysis. Again, some recognised texts determine PRPs from risk determination within the hazard analysis.

 

One well-known CCP decision tree (Campden) modifies Codex by adding an initial "Stop" question, eg - is the appropriate control measure handled by a PRP ?

 

FSMA introduced it's own risk assessment variant which I believe was borrowed from other RA areas. Personally I have so far failed to comprehend it.

[anna898]

Thank you Charles and VThiruselvi for your answer

 

I do understand that there can be a different approach and outcomes of the applied HACCP plan depending on the standard or even decision tree used. Sometimes even using codex decision tree and campden one can give different outcomes. 

 

 

But can auditor request to change my Haccp because he thinks I should put all process steps in a CCP decision tree? 

[GMO]

To my mind it is the right approach you are taking, also no auditor has the right to say an "interpretive" element like this is right or wrong.  For one, how you're doing it is absolutely how Campden BRI trained me on Level 4.  If they can point to a standard and explain where you're not complying with it then fair enough but I bet they can't... 

[El Molino]

Our HACCP plan is based on the CFIA FSEP framework with the forms outlining all the requirements of hazard analysis and CCP determination. Since SFCR, CFIA has watered down the programs but most federally inspected plants have maintained the FSEP framework - check out Form 8 combined which provides the PRP, oPRP and CCP determination in a full spreadsheet summary where you add each step of the process. I don't think an auditor should tell you to " change" your plan as it is your program - not theirs - they are just there to determine if your FS program meets the standard.

[Charles.C]

Thank you Charles and VThiruselvi for your answer

 

I do understand that there can be a different approach and outcomes of the applied HACCP plan depending on the standard or even decision tree used. Sometimes even using codex decision tree and campden one can give different outcomes. 

 

 

But can auditor request to change my Haccp because he thinks I should put all process steps in a CCP decision tree? 

So which audited Standard are you talking about ?

[Hoosiersmoker]

Our HACCP plan takes all identified hazards into account. If they are controlled by PRPs it is stated in the plan as such with no need for CCPs. If you exclude certain hazards just by identifying them as already being controlled by your PRPs, where is the assessment for those written down? If the assessment / analysis isn't recorded somewhere it technically doesn't exist.

[anna898]

Hi All 

Thanks for all replies, 

I've done some research, even asked my uni professor.

And As I though I'm doing it right. If my hazards analysis states that identified hazard is not significant and its controlled by PRP, there is no need to put it through the CCP decision tree. 

We are not certified by BRC/GFSI just FSA and SALSA, it was auditor from FSA who said that my HACCP is wrong. Maybe he need a HACCP refresher

Thanks again for your help

[jbana]

Hello. I find this confusing too, Has anyone used the decision tree for the categorisation or prerequisites and operation prerequisites from Campden BRI fifth edition ?

[GMO]

Hello. I find this confusing too, Has anyone used the decision tree for the categorisation or prerequisites and operation prerequisites from Campden BRI fifth edition ?

 

There is a 6th edition now just to throw that into the mix, but honestly I wouldn't recommend it, 5th is simpler, 6th tries to incorporate FSMA which isn't relevant unless you export or are based in the US.

 

oPRPs are not required unless you work to an ISO 22000 based standard.  I think something between CCPs and PRPs is a good thing but I'm not sure oPRPs are it.  There is a flow in v5 which doesn't include oPRPs.

 

Are you working to FSSC 22000 or other standards?

[marco89634]

 

The auditor's recommendation to submit all steps of the process to the CCP decision tree is in fact a profound deviation from the logic of the Codex Alimentarius.

 

Codex defines a significant hazard as: "A hazard identified by a hazard analysis as reasonably likely to occur at an unacceptable level in the absence of control, and for which control is essential given the intended use of the food." (Source: GENERAL PRINCIPLES OF FOOD HYGIENE (CXC 1-1969), Section 6 "Definitions", Page 13). Its primary objective is to separate negligible hazards, managed by basic good hygiene practices (PRP/GHP), from those whose management is essential for food safety.

 

"The HACCP team should next evaluate the hazards to identify which of these hazards are such that their prevention, elimination, or reduction to acceptable levels is essential to the production of safe food (i.e. determine the significant hazards that have to be addressed in the HACCP plan)." Source: GENERAL PRINCIPLES OF FOOD HYGIENE (CXC 1-1969), Section 19.6 "List all potential hazards...", Page 43, Paragraph 2

If a hazard in stage X is assessed as not significant (e.g. Low Probability / Low Severity), the analysis for that hazard at that stage is concluded. It will be managed by the PRPs. T

The HACCP system, by its very nature, is no longer actively addressing it. To proceed further would be an inefficient and illogical allocation of resources. The CCP decision tree (or similar tools) is not an investigation tool to discover which hazards are significant; it is a classification tool to be applied only to hazards that have already passed the threshold of significance.

"CCPs are to be determined only for hazards identified as significant as the result of a hazard analysis." (Source: GENERAL PRINCIPLES OF FOOD HYGIENE (CXC 1-1969), Section 19.7 "Determine the critical control points (CCPs)", Page 44, Paragraph 2.)

This single sentence is the key. It states unambiguously that the universe of application for determining CCPs is limited to significant hazards only.

 

As further confirmation, the instructions for use of the decision tree itself, reported in the annex, specify: "Figure 1 Example of a CCP decision tree - apply to each step where a specified significant hazard is identified" (Source: GENERAL PRINCIPLES OF FOOD HYGIENE (CXC 1-1969), Annex IV "Tools to determine the critical control points (CCPs)", Page 53.) Applying the decision tree to a non-significant hazard is a methodologically incorrect action. The tree, by its structure, would almost always result in "Not a CCP", making the exercise, as you rightly pointed out, redundant and useless.

 

 

Example: Meat Industry.

Product: Raw, refrigerated meat.

Hazard at Reception: Proliferation of pathogens (e.g. Salmonella, Listeria monocytogenes, E. coli) due to abuse of temperature during transport. Risk Analysis: The probability of a failure of the truck's refrigeration system is not negligible, and the severity of poisoning by these pathogens is very high. The hazard is highly significant. CCP Determination:

The decision tree, applied to this significant hazard, will inevitably lead to identifying the temperature control at reception as a CCP. It is the last possibility to control that hazard before the raw material enters the process.

 

Example Biscuit Factory.

Product: Flour, sugar (dry, stable ingredients).

Hazard on Receipt: Contamination by foreign bodies (e.g. wood fragments from pallet), presence of pests, chemical contamination from improper transportation. The risk of bacterial growth related to temperature is zero. Hazard Analysis:

These hazards are important, but their probability and/or severity is effectively managed by robust Prerequisite Programs (PRPs): supplier qualification and audit, visual inspection of trucks and bags, pest control program, certificates of analysis. There is no measurable "critical limit" for the absence of a foreign body that makes the step a CCP. The hazard, although present, is not considered "significant" in the sense of requiring a CCP, because the PRPs provide sufficient control. CCP Determination: Since there is no significant hazard that requires specific control at this step (other than the PRPs), the decision tree is not even applied.

[Scampi]

 

 

As further confirmation, the instructions for use of the decision tree itself, reported in the annex, specify: "Figure 1 Example of a CCP decision tree - apply to each step where a specified significant hazard is identified" (Source: GENERAL PRINCIPLES OF FOOD HYGIENE (CXC 1-1969), Annex IV "Tools to determine the critical control points (CCPs)", Page 53.) Applying the decision tree to a non-significant hazard is a methodologically incorrect action. The tree, by its structure, would almost always result in "Not a CCP", making the exercise, as you rightly pointed out, redundant and useless.

 

 

 

 

And I could not disagree more and you're assuming a hazard analysis has NOT been completed properly .  Just because you've identified a significant hazard does not automatically make it a CCP

 

 

A 4-Step process

The Codex CCP decision tree involves answering four (4) questions. These questions must be answered in order. The answer to each question will either be ‘YES’ or ‘NO’.

Keep in mind that as part of completing a hazard analysis (HACCP Principle 1) you are required to identify ‘significant hazards’. You would therefore use the decision tree to answer the questions about the significant hazards you have identified during that process.

Let’s take a look a the four questions now.

Question 1: Can the significant hazard be controlled to an acceptable level at this step by prerequisite programs (e.g., GHPs)?*

Question 2: Do specific control measures for the identified significant hazard exist at this step?

Question 3: Will a subsequent step prevent or eliminate the identified significant hazard or reduce it to an acceptable level?

Question 4: Can this step specifically prevent or eliminate the identified significant hazard or reduce it to an acceptable level?

[Killian]

And I could not disagree more and you're assuming a hazard analysis has NOT been completed properly .  Just because you've identified a significant hazard does not automatically make it a CCP

 

 

A 4-Step process

The Codex CCP decision tree involves answering four (4) questions. These questions must be answered in order. The answer to each question will either be ‘YES’ or ‘NO’.

Keep in mind that as part of completing a hazard analysis (HACCP Principle 1) you are required to identify ‘significant hazards’. You would therefore use the decision tree to answer the questions about the significant hazards you have identified during that process.

Let’s take a look a the four questions now.

Question 1: Can the significant hazard be controlled to an acceptable level at this step by prerequisite programs (e.g., GHPs)?*

Question 2: Do specific control measures for the identified significant hazard exist at this step?

Question 3: Will a subsequent step prevent or eliminate the identified significant hazard or reduce it to an acceptable level?

Question 4: Can this step specifically prevent or eliminate the identified significant hazard or reduce it to an acceptable level?

 

Maybe I'm misreading something, but I think his point was you shouldn't apply the decision tree to hazards which are not deemed to be significant, not that all significant hazards are CCPs.

[GMO]

Maybe I'm misreading something, but I think his point was you shouldn't apply the decision tree to hazards which are not deemed to be significant, not that all significant hazards are CCPs.

 

That's how I read it too.  

 

There is a "get out" in both Campden and Codex decision trees where it asks up front nowadays if it's already controlled by prerequisites so if you do accidentally decide to apply the decision tree to non significant hazards, they're likely to get screened out.  I think that was the point.