Do we have to document all corrections and corrective actions in ISO 9001 10.2.2 ?

[Jen Kirley]

The standard clearly allows for errors/nonconforming outputs NOT to be treated via a corrective action. ISO 9001:2015 10.2.1 b). evaluate the need for action....

There will be non-critical, non-repetitive issues where correction only is adequate.

I would agree except for the extensive guidance given in ISO/TS 9002:2016, excerpted as:

10.2.1 The intent of this subclause is to ensure that the organization manages nonconformities, and implements corrective action, appropriately.

When a nonconformity occurs (including those arising from complaints; from identified nonconforming outputs [see ISO 9001:2015, 8.7]; problems arising from external providers or other relevant interested parties; audit results; or the effects of unplanned changes), the organization should take action to investigate what has gone wrong, to correct it if possible, and to avoid similar issues from recurring in the future. The organization should seek to eliminate permanently the causes and consequent effects of problems that could have a negative impact on its:...

The organization should take action to control or correct any nonconformity. This can be achieved by containing the problem while investigations continue...

When evaluating the action needed for a nonconformity the might be instances where the cause of a nonconformity cannot be eliminated, therefore, the organization should consider taking actions to be able to detect and minimize the effects of the nonconformity if it were to occur again.

The standard plainly enough states that corrective action should be appropriate, just not selective.

That said, I understand small things should get small actions as long as we stay aware of repeats and/or occurrence elsewhere. That is why I put the Corrective Action Log into the Resources, which (based on three criteria) includes a low-risk level "3W" option (Who will do What, and When); the entries in there can be tracked for recurrence.

 

[Sidney Vianna]

So, in your assessment, ANY discrepancy/error/nonconformity, irrespective of how trivial/inconsequential/isolated, MUST have a corrective action. Is that your position?

 

[Jen Kirley]

So, in your assessment, ANY discrepancy/error/nonconformity, irrespective of how trivial/inconsequential/isolated, MUST have a corrective action. Is that your position?

Please see my edit, which I regret I did not get in before your post.

 

[Kronos147]

The standard clearly allows for errors/nonconforming outputs NOT to be treated via a corrective action. ISO 9001:2015 10.2.1 b). evaluate the need for action....

There will be non-critical, non-repetitive issues where correction only is adequate.

So, in your assessment, ANY discrepancy/error/nonconformity, irrespective of how trivial/inconsequential/isolated, MUST have a corrective action. Is that your position?

FWIW, I believe in logging, and then determining thresholds where logical (like Jen says, based on risks, opportunities, context, needs and expectations of interested parties, etc... and of course resources!) where to take formal corrective action.

Without the log, your system may not be able to recognize what they 'think' is non-repetitive, is in fact just that.

The point of the standard is consistent results. Systems should determine when it makes sense to commit to the corrective action process to prevent it from becoming a paperwork excercise.

 

[Sidney Vianna]

Without the log, your system may not be able to recognize what they 'think' is non-repetitive, is in fact just that.

As long as the organization has been able to drive fear out, sure.

But imagine what the workforce might think when they find out that the manager is "keeping tabs" of screw-ups. If they fear being punished, they will hide instances of undetected problems.

 

[Jen Kirley]

As long as the organization has been able to drive fear out, sure.

But imagine what the workforce might think when they find out that the manager is "keeping tabs" of screw-ups. If they fear being punished, they will hide instances of undetected problems.

That's the challenge management must rise to. Problems should be addressed for what they are: process ineffectiveness, and not punish people when improvements could be helping them to perform better. Letting even small nonconformances go unrecognized introduces the risk that we will subject ourselves to the "death by a thousand cuts" syndrome.

 

[Kronos147]

But imagine what the workforce might think when they find out that the manager is "keeping tabs" of screw-ups. If they fear being punished, they will hide instances of undetected problems.

That is a risk, for sure!

That boils down to culture. There was some of that at E.S.M. when I got there.

I explained that it was the process that was wrong and that as a management, we needed to plan for the process success better and that is why we keep stats. We can't work on every perceived problem, but if there are some stats, it tells the organization where to put their efforts.

I was so glad to see our organization 'transition' from blame-casting to root cause analysis and effective corrective actions! The staff engages me with continual improvements\risk mitigations that us managers would just not have seen from above the trenches, as it were.

 

[John Broomfield]

But do they record every mistake they make?

No one is disputing the need for correction and corrective action (and records thereof) where preventive action has failed.

Mistakes may not result in nonconforming services or products but would be recorded if they do.

 

[Jen Kirley]

But do they record every mistake they make?

No one is disputing the need for correction and corrective action (and records thereof) where preventive action has failed.

Mistakes may not result in nonconforming services or products but would be recorded if they do.

The standard says "nonconforming outputs." If a mistake does not result in a nonconforming output, how would we even know about it? It is not practical to expect perfection. 10.2 refers to outcomes.

 

[John Broomfield]

Thanks for clarifying that.

My reading of this thread had implied otherwise.