Preventive Action (PA) and Corrective Action (CA) - One or Two Procedures?

[potdar]

Nice job Patricia.

the system I use is almost similar, though the continuous improvement aspect is missing from it. I am definitely planning to add that. And mine is a plain excel sheet - definitely not as colorful!

One additional aspect I include though, is whether any change in documentation is required? If the answer is positive, the CA / PA does not close till the document has been officially revised and the revision is put on record.

 

[Patricia Ravanello]

Hi Potdar,

Thanks for your feedback.

Documentation changes precipitated by Corrective Actions are captured in a linked database called the "Action Item Database", which, as you stated, requires that the necessary document revisions be made before the Corrective Action can be closed out. It is part of the verification and validation steps in the Corrective Action.

Sounds like you have an excellent methodology which is very pro-active, precluding the possibility of errors or omissions. We all have so much to learn from each other!

Thanks again,
Patricia

 

[Helmut Jilling]

How wonderfully simple. Just a / between the CA & PA. I have had trouble getting preventive action documented unless a corrective action occurred. I have been trying to improve my form and this is a good answer.

Thank you.

If a corrective action happened, it does not lead to a preventive action. The ISO standard clearly defines a preventive action as a potential failure.

Don't want to hijack this thread, we have beaten this on many other threads already. If you don't agree, there are several other threads if you do a search.

 

[Patricia Ravanello]

If a corrective action happened, it does not lead to a preventive action. The ISO standard clearly defines a preventive action as a potential failure.

Sorry, I'm too tired to look up the other threads...thought I'd reply here.

You're right...when a corrective action happens, it does not always lead to a preventive action per say, however, potential failures may be identified through a broad range of activities, including a previously validated corrective action.

This is how a C/A might generate an opportunity for a P/A:
Once a Corrective Action is implemented and validated, Management should ask themselves if an opportunity for a Prev. Action exists through the application of the corrective action to a similar process or scenario, thereby preventing a failure mode that could precipitate a failure like the one that led to the original corrective action.

Corrective actions are not always the catalyst for a Preventive Action (see attachment), but they can be.

Patricia Ravanello

P.S. Please note, the attachment is only 1 page of a 3 page document.
Following the validation of a Corrective Action (page 1 of SOP-0005) is a "Decision Diamond": "Could the application of this corrective action to other processes eliminate potential nonconformities?" If the answer is "yes", you link to Page 2, the Preventive Action Flow Chart.

 

[Jim Wynne]

If a corrective action happened, it does not lead to a preventive action. The ISO standard clearly defines a preventive action as a potential failure.

Don't want to hijack this thread, we have beaten this on many other threads already. If you don't agree, there are several other threads if you do a search.

Sorry, Helmut, but you made the statement here, and it would make no sense to answer it elsewhere. The fact is that corrective action can result in PA that satisfies the (nonsensical) ISO definitions. If we do CA in the process where the NC occurred, subsequent actions in that process should not be considered "preventive" in the ISO sense. But, if we take the opportunity to extend the actions to other processes where the same NC (or one that's substantially similar) has the potential to occur but hasn't, then ISO preventive action has been done.

 

[Helmut Jilling]

Sorry, Helmut, but you made the statement here, and it would make no sense to answer it elsewhere. The fact is that corrective action can result in PA that satisfies the (nonsensical) ISO definitions. If we do CA in the process where the NC occurred, subsequent actions in that process should not be considered "preventive" in the ISO sense. But, if we take the opportunity to extend the actions to other processes where the same NC (or one that's substantially similar) has the potential to occur but hasn't, then ISO preventive action has been done.

I know, this has gone round and round in other threads. In TS, your approach would not be appropriate, because that step is addressed in TS cl 8.5.2.3 "Corrective Action Impact."

In ISO, there is nothing that says we can't, but why open 2 different forms for the same event. Not efficient. On the 8D form that many people use, the intent of step 7 was to apply this to other processes. In a 5 step type of CAR form, I would do the immediate and extended corrective action in the action section. I wouldn't open a separate form all over again.

But, as we have agreed in the past, whatever gets users to the best final answer is the important part. The semantics are not all that important.

 

[Stijloor]

But, as we have agreed in the past, whatever gets users to the best final answer is the important part. The semantics are not all that important.

Helmut,

Hallelujah!

Stijloor.

 

[TamTom]

Hello,

when I read this discussion I got a question:

We have very similar products, when I make a corrective action on one part X and use this for the other parts Y and Z as well, for the parts Y and Z this is not preventive??
We use part family FMEA a corrective action can goes in there and is used for all further developments is this preventive or lesson learned.
So after all this very long theard I didn't find the answer.

Regards,

Tamara

 

[Ahsas]

If the answer to "Is corrective action necessary?" is NO, (which means their metric is on plan and/or meeting its stated objective), then they have to answer the question, "Is there an opportunity for continual improvement?" (you can't initiate a Continual Improvement unless your metric is meeting objective, and no actions are necessary - any action taken before you're meeting your objective is, by definition, "corrective", and not an improvement). If the answer is "YES", you hyperlink to the Continual Improvement. If the answer is "No", you answer the question "Why not?" (possible answers: "no resources currently available, no improvements identified, timing constraints, etc.), and that's the end of the sequence.

However, this demostrates to the auditors that Sr. Management is always assessing the opportunity for continual improvements, and reviewing if they are feasible, cost-efficient, and timely. It further hyperlinks them to all the Continual Improvement Actions that have been initiated...neat, tidy, readily accessible....Sr. Management knows right where to go to answer the auditor's question..."What Continual Improvement projects have you undertaken since the last audit?" and "Can you show me some of them?"

It's almost idiot-proof!

It is an interesting observation that the continuous improvement can't be initiated unless the metric has met its objective.

In my experience, often times the departments have the processes in place but no specific metrics and their targets or objectives, except that the output must satisfy the customer's need. The measurable processes are considered 'very complex'.

The process owners (a.k.a. department managers) when confronted with the deficiencies in the process or the failures (found during verifications by the quality dept.) would often cite the prevailing continuous improvement (CI) culture (in the organization) to circumvent the need to take immediate preventive actions to fix the process. The need to prove the effectiveness of the 'as is' process and take credit for operating it successfully is predominant.

Sadly enough, the much needed kick (for making change) has to come from outside (the external customers) which, of course, is part of the continuous improvement culture.

 

[harry]

Hello,

when I read this discussion I got a question:

We have very similar products, when I make a corrective action on one part X and use this for the other parts Y and Z as well, for the parts Y and Z this is not preventive??
We use part family FMEA a corrective action can goes in there and is used for all further developments is this preventive or lesson learned.
So after all this very long theard I didn't find the answer.

Regards,

Tamara

Sorry, Helmut, but you made the statement here, and it would make no sense to answer it elsewhere. The fact is that corrective action can result in PA that satisfies the (nonsensical) ISO definitions. If we do CA in the process where the NC occurred, subsequent actions in that process should not be considered "preventive" in the ISO sense. But, if we take the opportunity to extend the actions to other processes where the same NC (or one that's substantially similar) has the potential to occur but hasn't, then to ISO, preventive action has been done.

Hi Tamara,

I am sorry that you felt confused but I think Jim had given a good answer to your query.