How to deal with too many CARs (Corrective Action Requests), PARs (Preventive Action)

[LesPiles]

The same old problem, we, quality guys, are facing every day in our systems.

For Upper Management, or the rest of team we're part of, their simple solution is "categorize them" by priority, and or by the gain we should have from their resolution.

In this friday afternoon we're all tired from this extra busy week, do some of you would like to share their methodology of evaluation or, better, their evaluation template ?

Thank you in advance !

[dwend]

My solution is to use this the CAR process as judiciously as possible so as to not have high numbers of CARS open at one time. For internal audits, we have a "corrections" process that will allow minor tweaks without invoking the full process. From there it is anything that directly affects product conformity, customer related, system failure, down to documentation errors as a lower priority with time open a consideration as well. I manage our CAR process but sending out documentation to individual owners who have the responsibility to request team help if needed. We put up a Rolling Action Item List every 2 weeks for review in front of upper management who then can turn up the voltage on their prods if needed. Ultimately its a resource decision which usually falls in the domain of upper management. Bang for the buck is hard to calculate. Be interested in what others have to say.....

[QCAce]

Don't write so many. Combine related nonconformances into one CAR.

[Jennifer Kirley]

Not knowing the issues these CARs were written for, it is hard to advise. But if there are many, many CARs I wonder if they are being issued for trivia or if the CAs are effective?

I do agree that nonconforming issues can be combined into one CA. But if multiple incidents of the same thing need to be respond to, I wonder if there isn't a systemic issue that the typical CAR would handle. Just a thought...

[ssz102]

the urgent CAR will be first done
and then you should be find out the root cause to results these too many CAR

[p1stonbroke]

Raise opportunities for improvement which do not invoke the whole process
If they have not been actioned next time you audit, then they get the full Non-Conformance.

In terms of PARs, it depends how you view this...
one view:
preventive as in preventing something which has never happened yet but which might happen...
or
Root cause - preventing a recurrence of something that went wrong

I tend to condier that witth a CAR, comes a PAR (second variant) and that genuine opportunities for improvement can be documented as PARs, as well as those which might come out of (for example) FMEA

Yet another option is to permit persons to resolve audit findings before the end of the audit, prove they have solved the problem, write the finding up in the audit, but recording the fact that it has been resolved, without formally raising them as CARs - this approach is offten well recived and means that they take the ones you feel obliged to raise formally much mroe seriously.

If you raise too many CARs/PARs, or too many of them are open/outstanding/late, then they won't get the focus they should and the net result is less than if they are raised for issues that really need resolving, with a less formal approach for those with reduced implications

[silentrunning]

If you write too many CARs they will become meaningless and be ignored by everyone except the poor soul that has to close them out.

[Steve Prevette]

What is "too many"? What would be "too few"? Why?

I do a lot of work for my employer with their Corrective Action Management system. Questions include:

1. Are these self discovered, or a result of external sources? Note that if you stop looking, CAR's will go down. If you change for some reason to look harder, CAR's will go up.

2. What does the control chart show - is the rate of self assessment increasing or decreasing or staying the same? Is the rate of CARs increasing or decreasing or staying the same?

3. How do the data split out by subject area, by severity? A combination of Pareto Charts and Control Charts may help here.

4. What sort of causal analysis is done? Only on a case by case individual basis? Or by logical groupings of CAR's?