Dear Zeeshan,
You didn’t mention if you are already certified or not.?
You asked why people don’t precisely follow the 22000 standard. Some possible reasons are –
1. Parts of 7.4.4 are ambiguous to some (many?) people.
2. The consequences of an exact adherence to 22000 format for some practical situations are rather unpleasant, eg dozens of OPRPs. I noticed this problem in some published water systems.
3. ISO 22004 offers an excellent justification for alternative options. (have you read it yet?)
4. Auditors are simply not requiring a close match. In fact the expectations seem to vary a lot.
No doubt there are more.
Regarding tree presentations, I hv myself so far sort of interpreted the items in 7.4.4 (a-g) as related to 3 basic factors (somewhat disregarding, but not forgetting, the equally important conceptual features of OPRPs etc). Other people may well disagree, it’s a very flexible / interpretative topic as illustrated by all the posted scoring tables. –
A. The fundamental need for a properly implemented CM. Max need > +CCP
B. The ability / likelihood (AL) of a chosen / analysed / implemented control measure (CM) being optimally "sucessful". So max (AL) > +CCP (this tends to create OPRPs by "default" but the allocation should also make sense within their conceptual definition)
C. Handling a situation where a CCP might hv been considered appropriate but for which no practical critical limits are readily stated.
Obviously many of the items discussed below are not black and white [hence all these discussions 
], an occurrence which has probably dismayed the 22000 designers. (note to PAS for their next prescriptive category??)
As related to (A-C), I (very approximately) interpret 7.4.4 (a-g) as –
(a) > B. High effectiveness tends to CCP (I suppose "effectiveness" is a theoretical/practical composite as described in 7.4.4 a,b etc of ISO 22004 however achievance of the CCP criteria of def.3.10 is presumably a crucial part of any performance objective target [eg see note3 in above 22004 ref.])
(b) > B. Continuous monitoring (but variously interpreted) tends to CCP
© > B. (variously interpreted) but typically considered as CM required to be in the “best” location, usually nearest the end of the process chain with no subsequent equivalent competitors.
(d) > A. (as I understand it, this is a sort of circular logic). If there was no initial significant chance of a CM failing (inc. due to varying process conditions), no reason to invoke HACCP. So a sig. likelihood tends to CCP. This parameter is coupled to (e) and the pair are a sort of re-iteration of the starting situation.
(e) > A. Linked with (d). (eg Risk = Likelihood x Severity) (in the current application anyway). So max. severity tends to CCP (but not necessarily independently of likelihood of course).
(f) > B. A designed / appropriately deployed CM should be optimum, therefore tends to CCP.
(g) > B. (variously interpreted). One viewpoint is that a single CM which establishes full control is optimum. IMO, take yr pick either way and validate.
From the published / posted items in my previous link, you will see there is general agreement on some of above opinions, disagreement on others. IMO, for several of the contentious items, as long as you can justify yr point of view, it probably doesn’t matter to the auditor (particularly in view of the "clarifications" in ISO 22004 which take precedence over ISO 22000).
I may hv got some of the above wrong, anyone feel free to comment!.
As far as the minimum number of criteria are concerned, this is typically a result of an intuitive prioritisation / weighting within the elements of 7.4.4 (a-g). Some people’s posts here hv implied that “1” was auditorily acceptable for them. The Procert example is the shortest one I think I hv seen published. The justification may relate to the process etc. And the auditor 
So, referring to yr latest tree, on the basis of above logics, I hv some reservations about the interpretation / positioning of the first, second (b), sixth (f) boxes on LHS.
It is also equally important to analyse the steps passed through prior to category 7.4.4 (a) box as illustrated in the Procert flow chart. The details will depend on yr process again.
You may also find that yr specific process is more compatible with the scoring system as in yr previous post. The latter has the advantage (especially using 3-step score) of more easily working with non-black/white situations. The cost may be more work and thinking. As usual, it all depends. (FSSM’s vegetable / excel sheet is an example of the use of scoring options as per his interpretation.)
Hope the above comments are intelligible, the 22000 standard is a classic of unclear statements IMO however I can appreciate that this whole risk topic is very dificult to present in an auditable format so (reluctant) kudos to ISO for at least making the attempt.
Rgds / Charles.C
Categorization of CM.pdf 82.96KB
319 downloads
