Help
Since it appears to me that some support is available for all 3 possible categories of the 2nd control measure (CM) referred, I have suggested a generic answer to yr basic question, ie what to do if this CM fails. It appears to me that ISO 22000 attempts to present the requirements in some kind of chronological safety order.
If the CM is a PRP, >>7.8 (inc. 7.2.3, 7.10.3 et al)
If the CM is (applied at) a CCP, >> 7.6.5, 7.10
If the CM is an OPRP, same as CCP, (presumably based on some sort of “OPRP limits” ).
(Note that a rather intriguing conceptual distinction is made [10.1] regarding the follow-up to a failed OPRP as compared to a CCP, although this is not relevant in present example).
All 3 will presumably have potential implication to 8.4.2
Last 2 may have relevance to ISO 22004, 8.2 (esp. last para.) .Frankly, I would probably invent a pseudo-validation so as to include PRP within this comment also, if it were otherwise excluded.
(Of course, ISO 22004 implies that all the OPRP/CCP distinguishing efforts are basically a waste of everybody's time. Probably the next revision will suggest all CMs should be labelled PRPP or PRPR [ primary or retrospective]
).In the specific example, I would hv thought all 3 options would have the same 1st consequence assuming that the initial results were confirmable, ie batch rejection. Many of the subsequent supplier consequences will IMO, as Tony indicated, depend upon / fall under the category of “root cause”, usually a section of the standard CA form. Plus some kind of version of the American "3 knocks(?)" rule.
I hope the above partially embraces the original thread heading (strictly should hv included PAS220) and yr example.
Rgds / Charles.C
