27 replies to this topic

[Charles Chew]

Looking at the Codex's 4-Q Decision Tree approach for determining CCP, the methodology appears more applicable to determining CCPs for "Process Points" rather than non-process issues such as "Ingredients and Packaging Materials".

I have seen CCP determination applied with the 4-Q approach but it just does not make any sense to me. (How do you go about it on issues concerning non-process critical risk categories or do we simply ingore these areas of ingredients and packaging materials!)

However, if your rogram is audited against the Codex Guidelines, using any other form of CCP determination is a big NO to the food auditors.

Is the Codex Decision Tree a God-Sent Formula? Would anyone like to share their opinions on this?

Charles Chew

[Koko LMQ]

Sharing!!!!!!!!

When I develop the HACCP plan and process flow diagram, I use the wording of "receiving xxxx(raw material)" then I think you can apply the decision tree as the processing step not the material ones.

Q1 ; Do control measure(s) exist for receiving of this raw material ? (at this step and including the subsequent steps)

Q2 ; Is the receiving step specifically designed to eliminate or reduce the likely occurence of a hazard to an acceptable level? If there is the inspection point e.g. Ask supplier to send the Certificate Of Analysis when sending the raw material, I think it should be CCP of this identified significant hazard.

NY

[Charles Chew]

I use the wording of "receiving xxxx(raw material)" then I think you can apply the decision tree as the processing step not the material ones.

Dear NY,
Sorry for having totally forgotten about this topic I posted earlier.

From your suggested application, "receiving of raw materials" is indeed a step (part of the flow diagram) and the decision tree does apply.

Look at Q2 of the Codex Decision Tree - which reads as "Is this step specifically designed to eliminate or reduce the likely occurence of a hazard to an accpetable level"

The issue of potential hazards or CCPs in "ingredients and material packaging (I & M)" can not be determined using the above Q2 of the decision tree which refers clearly to "this step" while (I & M) is also clearly NOT a step but rather tangible items that are "used or applied" in the production of the final end product.

Can Q3 & Q4 be sensibly applied as well. I dont believe so as the wordings appear to be towards process steps as well.

Or - do we just ignore the determination of CCP for (I & M) but what if the primary packaging material used is not what was intended and is apparently suspected to be non-food grade. (If a CCP analysis is not done in the first place - this may be not be known at all.

What do you think?

Charles Chew

[Charles Chew]

Seriously, before one even goes to the extend of defining your Hazard Analysis Table, what if one of your main ingredients has a significant high risk of microbial contaminations (milk is a good example) - at what level of the HACCP program do you recognise the threat to safe quality food issues and what methodology to be used to determine the CCP of receiving an ingredient of this nature.

And, when that is recognised, what else is done to address this threat from such an INGREDIENT (reminder - this is not a PROCESS.....)

IMO - this is an impt. support element to the HACCP Principles. How can we have a HACCP System if this is not addressed correctly? Viewpoints - anyone!

Charles Chew

[petegilmartin]

e.g. Milk
Step Incoming Goods, Hazard (Temperatures out of range), Control Measure (Temperature controlled vehicle), Critical Limit <8C

Do control measures exist = Yes
Is the step specifically designed = No
Could contamination occur in excess....etc = Yes
Will subsequent step eliminate...etc = No

Therefore CCP

Is this on the right track???

[Charles Chew]

Do control measures exist? - I guess you may have a Suppliers Assurance Program (since this is an ingredient)

Is the step specifically designed...... - the problem is this is not a step but rather an ingredient. So is this question under the decision tree viable?

Would contamination occur in excess .... - it could contaminate process equipment though if other products of non-milk nature is produced in the same factory.

Will subsequent step eliminate ... - as an ingredient input - No!..but it is an allergy issue in thr form of lactose intolerant.

I am still of the opinion that the Codex Decision Tree was never designed to test the determiantion of CCPs in ingredients and packaging BUT merely the PROCESS aspect. Looking forward to other views.

Charles Chew

[petegilmartin]

I see where you are coming from. It does appear illogical to answer Q2 in this way when you want to address the ingredient (not the step). However, if you follow the decision tree through, it does give you the answer that you would have guessed. i.e. microbial hazard with a temperature control will always be a CCP. Therefore, although appearing slightly illogical, the decision tree prevails.

[Charles Chew]

Peter,
We probably agree that the Codex Decision Tree (albeit - has been around for too long) was specifically designed to address process controls and the approach has been used by ALL since.

As it appears obvious that this D-tree is meant for Process Only - do you think using a "decision tree" designed by the HACCP Team members for "ingredients and Packaging (I & M)" would be an acceptable approach. Afterall, there really is no decision tree under Codex that has been designed for I & M or is there

Again isn't it up to the management and the team to justify the approach if no other guidelines are available. Personally, I know Consultant Members are well aware of this issue and it would be good to hear their views too.

Charles Chew

[Simon]

Peter,
Again isn't it up to the management and the team to justify the approach if no other guidelines are available. Personally, I know Consultant Members are well aware of this issue and it would be good to hear their views too.

Yes where are all the Food Safety Consultants? I know you're out there. Anyone care to contribute to the discussion or are you only here for the

Regards,
Simon

[petegilmartin]

Charles,
I am not aware of a Codex decision tree for ingredients and materials. Have you yourself (or HACCP team) tried developing one or is it just an idea? Could get complicated I would imagine.

[Charles Chew]

Peter,

I am raising this issue here because I truly feel that Codex has flawed in this area.

To your question of whether a model has been established for a specific CCP assessment of ingredients and packaging materials - the answer is yes and the model has always been accepted by all and if not by a highly qualified IRCA Food Registered Auditor..............surely you cannot dispute an IRCA Food Auditor.

Really like to hear from others what their models or views are on this issue. Saferpak is not just for taking information but really a place for sharing. I think we have about 1000+ members here but only a handful truly get involved and Peter, thanks for having made your mark here.

Will justly post at the appropriate time.

Charles Chew

[Charles Chew]

Peter,

I have just sent Simon a version who as Adm. will pass them on duly to members who participated on this subject.. He will be sending you a copy soon.

Regards
Charles Chew

[okido]

Kia ora Saferpakkers,

Ten years ago when we started with HACCP we ran into the same problem with the CCP tree.
Platsic packaging production we are talking about.
We made some changes in it, made a decision table, but were still not pleased with the outcome. Some products or product components did not fit in. We decided not to use it.
The outcome of the hazard analysis is our foundation, we estimate the likelihood of occurrence and severity of the outcome on a scale from low to high. If one of them is high it is a CCP.
Some products or product components are by definition categorised low, middle or high.
Plain and simple.

No worries, Okido

[Charles Chew]

Kia ora okido,

You know if you have your HACCP program certified against the Codex Guidelines - the auditor would have instantly given you a Major NC. But the point is as what you have pointed out - relevancy.

Just like Peter, he had analysed out correctly the hazards on MILK - he knew where the CCPs were and what the Monitoring Control Procedures are. In short, he knew what to do with just a Hazard Analysis Table (w/o the damn CCP determination decision tree structure anyway)

Okido - you are right. Its just a simple risk assessment and as a matter of fact, I only use "High or Low" and had long discarded the "Medium" since God knows when.

BTW - did you have any problems with the auditors on the amended decision tree. Very interesting indeed! Tell us more.

Hei kona ra
Charles Chew

Edited by charleschew, 27 October 2005 - 04:33 PM.

[petegilmartin]

Richard A. Sprenger has developed the Highfield Simplified Decision Tree but I am told that Auditors do not like it (even though you get the same result!!)

[Charles Chew]

but I am told that Auditors do not like it (even though you get the same result!!)

Why

Was it a matter of factual error - can't be when you get the same result or auditors were just purely stubborn headed or would it be not in line with Codex again

Charles Chew

Edited by charleschew, 28 October 2005 - 09:48 AM.

[Koko LMQ]

Hey, I try to find some CCP D-Tree for I&M but I cannot. I think I used to see in the USFDA HACCP Blue book or SQF training manual. Anyone can help!

NY

[Charles Chew]

In the risk assessment, I would probably consider whether there are any hazards associated with the ingredients (static and non-static i.e. microbial). This way, we can highlight the specifics and deal with the source through appropriate means. Again, whether there are hazards associated with the primary packaging materials in particular.

With respect to any, will there be issues concerning cross contaminations during process including the presence of allergens (normally static in nature) etc. Furthermore, do we see the possibility of a reduction in or an elimination of during the subsequent process of these ingredients.

Question: Are all these necessary when Codex D-Tree on Process is already in place which means if required, we need to cover CCP Determination and Hazard Analysis for I & M on top of Process..............can be hell of a lot of work.

Charles Chew

[APPAJI]

Dear Charles
I am shareing one of the decsision tree for Raw material and Ingradients given by DNV , Norway for primery production
Q1- Is there a named Hazard associated with this RM?
If No then go to Next RM
If Yes then go to Q2
Q2- Will processing , including correct consumer use , gaurantee removal of hazard or reduce to safe levels?
If NO then RM quality is likely to be a CCP for this hazard
If YES then go to Q3
Q3- oes the hazard pose a cross contamination risk, to the plant or other foods which is difficult to control
If No then go to next RM
If Yes then RM quality is likely to be a CCP for this hazard

[Charles Chew]

Hello Appaji,

I am glad to notice your response from my invitation to participate in this forum as your wide expereinces are needed to hlep out others along in this area.

As a matter of fact, we are using the same decision tree format for the identification of hazards in ingredients and materials but am not sure if this approach was introduced by DNV, Norway as it is widely used in Malaysia.

Nevertheless, I am not quite happy with this approach as there are limitations in this format and should be further explored however short of any other approaches, this methodology is highly essential especially when dealing with ingredients at control source where toxins or other hazards that can no longer be reduced/removed by other subsequent process steps and invariably are categorized as significant risks with severe impact.

[APPAJI]

Hello Appaji,

I am glad to notice your response from my invitation to participate in this forum as your wide expereinces are needed to hlep out others along in this area.

As a matter of fact, we are using the same decision tree format for the identification of hazards in ingredients and materials but am not sure if this approach was introduced by DNV, Norway as it is widely used in Malaysia.

Nevertheless, I am not quite happy with this approach as there are limitations in this format and should be further explored however short of any other approaches, this methodology is highly essential especially when dealing with ingredients at control source where toxins or other hazards that can no longer be reduced/removed by other subsequent process steps and invariably are categorized as significant risks with severe impact.

Daer Charles ,
Thank you very much for intoducing this forum it also helpful to me as to every one, well on topic, I do agree on the confution but in of my innitial trainings on HACCP my tuter explained that the potential hazards you questioning and if no control exisists to eliminate or reduce to acceptable limit then the best control need to excert is avoiding such food ingradient from allowing into the Food Chain.
Hope we can discuss more on this
APPAJI

[Charles Chew]

voiding such food ingradient from allowing into the Food Chain.

Hi Appaji,
Yes, we are on the same page and I guess this is where ISO 22K is so effective in insisting communication does exist on both uptream and downstream with regards to identification, determination and control of suppliers' specifications.

Avoidance of hazards derived from ingredients into the internal food chain is essenital when no process control exists to remove or reduce to an acceptable level is possible. And, I guess this is where the D. Tree does come into play as well. Not entirely perfect but hey! there is currently no other way

[Charles.C]

Hi. I haven't seen the ISO doc so this is sorta off-topic (you did ask for contributions) however I offer these methodological comments on HACCP/Dtrees/risk analysis methods if any interest. It is of importance to note that regardless which way you start, both procedures require a risk analysis evaluation even if it is maybe a bit less explicit in the Dtree.
When I set up my HACCP plans (food) I had choice of using Dtree or occurrence/severity/probability matrix. The risk analysis routes had previously been extensively developed in other industries, eg pharmaceutical, importation of exotic species. Nonetheless I failed to find any which I could make directly satisfactory to myself, main problem being a working balance between number of categories/actions (as an example I found the item which can dwl this site a bit limited although beautifully described - it depends on what you/auditor want to get out of it).
I also met the option that some approaches recommended splitting out items like the raw materials ( plastics/ingredients etc) into a 'pre-HACCP' section on the assumption that these should have been automatically analysed in detail prior to the main process (via epidemiological history / certification routes etc). This manouevre enables xreferences in filling out yr D4 plans referred in other posts.
If you study the IT available Dtrees closely (eg USFDA, Canada) they (honestly) comment here and there regarding some 'assumptions' which are necessary to make use of their Dtrees (eg the USFDA scheme in the latest version I have seen (ca. 2003) allows you to classify raw seafoods (micro. aspect) as a Non-CCP if will be fully cooked by the final consumer. I found this logic not acceptable personally. As a result of such issues I didn't use Dtrees but I did have a lot of arguments with auditors who expected a nice standardised approach.
Regardless my critcisms I would still suggest anyone interested in food area to look up the US / Canada / NZ Dtree presentations for the best items I know of being currently available on IT.
If these comments mirror some other thread apologies in advance.
Regards / Charles.C

edit added - I just saw the file re iso22000, seems that pre-haccp is now respectabilised as PRP and still causing interpretative difficulties, c'est la vie.

Edited by Charles.C, 20 March 2006 - 09:58 AM.

[Simon]

Regardless my critcisms I would still suggest anyone interested in food area to look up the US / Canada / NZ Dtree presentations for the best items I know of being currently available on IT.

Charles can you provide web links for reference.

Thanks,

Simon

[Witch]

I would like to remind you, that a CCP is a switch and no dimmer. That in mind, I do not have problems with the decision tree. Every process that lovers the risk, is a prerequisite. It is not a problem to think about raw materials, qualify their risk potential and decide, whether a process dims it. If so, you might be able to dim to a level where via FMEA a CCP is not necessary.
What exactly is your problem?

Andrea