10 replies to this topic

[QAD_Rebisco]

Hi!

 

Does anyone have an updated risk assessment matrix for HACCP hazard analysis? We just had our SGS Audit Recertification last October 2016 and the Auditor asked for an updated reference in which we based our risk assessment. Attached is the old one we used from SQF (i forgot what edition). Does anyone have an updated copy recognized by SGS? And a simpler one? Thanks in advance.

Attached Files

•  HM.011 appendix 2.pdf   106.01KB   701 downloads
•  HM.011 appendix 3-4.pdf   59.88KB   602 downloads

[Charles.C]

Hi QADRebisco,

 

One wonders what the auditor's approval criteria were ? All risk matrices are subjective/debatable in one way or another. Sounds like nitpicking unless yr subsequent usage was also at issue.

 

Do you mean that the auditor was -

 

(a) not happy with some particular item within yr current risk matrix or

(b) simply that yr reference supporting the matrix was date-wise too old for the auditor's liking or

(c) not one of the auditor's Company approved matrices (!!!?) ?

 

Yr example attached would not be my own choice but can probably be found in current use (somewhere).

 

The raw material risk assessment  tree you have attached is, IIRC, essentially identical to that in Mortimore's 2013 Book on HACCP. Too old ??

 

PS  - I don't recall ever seeing an official SQF-approved risk matrix ? Seen in a SQF webinar perhaps ?

[QAD_Rebisco]

Hi Charles!

 

Thank you for your comment. We were at fault because we simply relied on the hand-outs our Consultant gave us way back then (who have already resigned by the way) and then we just simply modified it to be a part of our document. We were so complacent that this matrix is applicable forever since it was accepted during the initial audit wherein the Consultant (who was also part of our Company from Corporate) participated in answering during that time. But lately, we were caught off-guard with the question because it never crossed our mind that the matrix could be phased-out and told that we were using an old one (we encounter different auditor each year from SGS due to their resignation). It's not that it was really a big deal to the auditor. She was just looking for references on where we based our HACCP documentation and (I think) measuring our in-depth knowledge about this. Can you please help/enlighten us about the most accepted matrix and it's source? Thank you.

[Charles.C]

Hi QADRebisco,

 

IMEX different users have their own personal preferences as to "style".

 

Theoretically, all risk matrices have their own pros/cons, eg with respect to size, but afaik any matrix will be auditorially acceptable if it can be justified, ie acceptably referenced. The most commonly used is probably 5x5 but many people also use 3x3. Other styles/sizes, eg asymmetical, 4x4,6x6 are also in use but I think less common.

 

I suggest you (a) nominate the matrix size you prefer (otherwise the options are legion) (b) how old an associated reference can you accept ? eg max 5 years ?

[QAD_Rebisco]

Hi Charles,

 

I think we'll try the simpler one, the 3X3. 3 - 5 years old reference would be acceptable. And as to the decision tree, we also got problem with the one we're using for raw materials especially for allergens. Can you help us with this? There are 3 questions and we answered it this way:

                Q1 - Is there a hazard with this raw material? (Yes, chemical hazard or allergen...proceed to Q2)

                Q2 - Are you or the customer going to process this hazard out of the product? (We cant answer No because it's not

                        supposed to be a CCP since there is an Allergen Control Program, so we answered Yes. But our justification is,

                        processing out of the product means proper labelling or making the customer aware of it's presence. We can't literally

                        process this out of the product, right? Since it is an allergen and inherent to RM? No problem with Q3 though, we're

                        just stucked with Q2 since the auditor was asking for the process and we haven't really justified it that well. So how

                        are we supposed to answer Q2? Or do you have a more relevant/sensible Decision Tree for RM?

 

Thank you very much.

[QAD_Rebisco]

By the way, my name is Mel M. 

[Charles.C]

Hi Mel M,

 

OK, actually the options in a 3x3 are relatively limited so can probably request a few 5x5s or whatever other posters can reference/put in. 

 

As far as the RM tree goes, IIRC the English anguage is flexible. I think there are some manouevres for this in Mortimore's book mentioned above. Are you familiar with it already ?

 

Some of the tree options may also depend on the BRC standard,

 

Other people are wlcome to comment of course.

[Charles.C]

Hi Mel M,

 

One way to provide recent risk matrices is to quote books however I’ve tried to find accessible responses which have some degree of status. This rapidly becomes difficult for > 3x3 matrices due the number of permutations in use (see below).

Anyway here is an example for 3x3 using one of the 2 common formats (the other format is in the slightly older last file below)

 mat1 - salsa,2015 - risk assessment.pdf   177.91KB   514 downloads

 

For 5x5, recent food examples seem scarce. 2 recent, slightly different, generic 5x5 examples follow. There are many others viewable.

 mat2 - ASQ,2013 how-to-apply-risk-to-your-quality-management-system-.pdf   3.65MB   454 downloads

 mat3 - Adelaide Uni,2012 - Risk Matrix.pdf   87.31KB   374 downloads

 

The choice for format within a given matrix size may depend on individual preferences (see last file below).

 

I have slightly enlarged the scope of the post below so as to provide a little more background context.

 

To illustrate the (bewildering) range of official (mainly US) matrices in use, here is a quite detailed comparative review from a rather surprising source  –

 mat4 - Risk Matrix,2015-Possibilities for Global harmonisation.pdf   6.14MB   403 downloads

 

Finally, this 2010 file contains the most readable explanation I have seen of how you might wish to select a particular matrix permutation within a given size (eg 5x5). It looks at/compares examples of a variety of matrices/permutations from 3x3 up.

 mat5 - Assessing your risk matrix,2010.pdf   656.2KB   421 downloads

 

Will revert on the raw material tree (unless anyone else cares to analyse it first).

[QAD_Rebisco]

Thanks Charles,

 

This is a bit complex but it seems helpful. Need to reassess our hazard analysis though. I think i'll go with the matrix 5. Currently, what we put in our hazard analysis is the resulting number of Severity and Likelihood. I think we need to just indicate the Low, Medium and High risk to make it simpler.

 

So if the auditor ask where we based this matrix that we used, what would be the appropriate answer? 

 

Mel

[Charles.C]

Hi Mel,

 

The answer to yr query likely depends on yr specific choice of matrix. My opinions are summarised below. Yr local experiences may differ(?).

 

IMEX, auditors have not been particularly concerned over the age of a reference for a risk matrix or its exact risk pattern as long as –

 

(a) the reference is not truly “antique” (cf NACMCF is 1997!) or attributed to some “obscure-looking” source. These are subjective terms of course.

(b) the general design/implementation looks internally consistent based on the auditor’s  own routine experience/guidance. Basically the matrix should not be visibly, significantly, atypical. "Typical" has some further theoretical implications but such are rarely considered in routine food haccp texts.

(c) the haccp conclusions/actions derived from the matrix are (i) compatible with  previous  auditor experiences/guidance for hazards commonly encountered elsewhere and (ii) seem “not unreasonable” for other hazards.

 

Recently I have used matrices whose "action" portion matches the format shown on pages 69, 72 in the (then) "official"  2004 file attached. (and also one of the options [Pg23] in the mat5 file/my previous Post).  Matrix theorists do suggest certain features regarding desirable formats but it seems those used in practice are mainly based on experience. The preferred choice may also relate to the specific risk assessment methodology used in the hazard analysis which can vary between users.

The haccp literature demonstrates that a substantial variety of matrix sizes and formats are in common use.

Two refs as per above are probably better than one.

 

So you may need to do a little research around yr chosen matrix format if you specifically seek an "official", "recent", food reference. If "recent" is not mandatory, life may be easier. Or you may get lucky.

 

 Risk Matrix, Guidance for Risk Managers,2004.pdf   683.37KB   387 downloads

[Charles.C]

addendum

 

If you prefer to retain yr current 5x5 (ie as in OP) and can accept a (2011) non-official source as a ref., I have located an example of  a near- identical matrix/CCP rule as per yr OP. Probably taken from the same original source as you mentioned. IMO this would represent an acceptable reference.

(Personally i find yr OP matrix a little "over" risk averse but I guess that should make the auditor happy).

 

 risk matrix,5x5.pdf   886.23KB   426 downloads

 

Regarding yr decision tree query in Post 5, the tree/procedure you describe is similar to that in the HACCP text (2013) by Mortimore et al as referred earlier. So this text book should be an acceptable reference.