Measuring Corrective Action Effectiveness

[TacitBlue]

Hi All,

We are updating our internal Corrective Action procedure; however, we do not have a solid means of measuring corrective action effectiveness. What is the simple best practice for measuring effectiveness that can be applied in production, engineering, design, or in other words across the board applicable to any CA?

For example, is anyone using a template of some kind that can quantitatively measure or graph or some type of log?

Some of the ideas is the measure reoccurance, but how do you do that and present it? or another idea was to look at lost productivity? but, again, how do you do that in concert with the CA effectiveness?

Please note, I would appreciate responses to not include references to other past posts or questioning why I am asking this. I am looking for a fresh perspective.

Thanks

 

[Mike S.]

I primarily consider does the cause(s) addressed by the corrective action(s) reoccur?

 

[jmech]

I don't think there is one "simple best practice ... applicable to any CA". The measurement method/tool/template/practice needs to be suitable for what is being measured.

When the corrective action is for a process that occurs frequently, check if the process/outputs now conform.

When the corrective action is for a process/event that occurs less frequently (or you don't know if it will ever occur again, such as producing a specific custom product or enacting a specific contingency plan), you might need to be more creative in measuring effectiveness.

 

[TacitBlue]

I primarily consider does the cause(s) addressed by the corrective action(s) reoccur?

Mike, we use TipQA Corrective Action module to record and monitor CAs. However, beyond going back or monitoring within a lengthy database, how would we monitor for reoccurrence? Should each CA recipient keep their own log?

 

[Mike S.]

TipQA? I pity you. That's awful software.

I usually don't do this review formally on every CA. The assumption is, and the norm is, that the cause(s) addressed by the corrective action(s) do not reoccur. But when a problem crops up we look to see if it is a repeat and if so, is the cause the same.

 

[Bev D]

There is no ‘fresh’ approach. Just old fashioned knowledge and real math.
Just because you want something, doesn’t mean you’re gonna get it.

 

[Sidney Vianna]

There is no ‘fresh’ approach. Just old fashioned knowledge and real math.
Just because you want something, doesn’t mean you’re gonna get it.

Spot on.

I don’t understand the question “how we would monitor recurrence”. The same way you identified the original problem that triggered a CAR.

 

[ChrisM]

"All corrective actions are different". You can bet that if there was one way of monitoring their effectiveness, some software developer would have already come up with a program to do the monitoring, from which they would be making a lot of money.

Part of the CA process would be to identify and state what was required from the arising action, e.g. to eliminate the issue or to reduce it (ideally with a qualifying number or percentage). Checking the effectiveness would then determine if this criteria had been met.... For example: no recurrence, then it's been eliminated; recurrence reduced from once a week to once a month, and the target was to reduce to once per month or less - corrective action has been effective

 

[Steve Prevette]

It really comes down to the word "effectiveness". Have you had any recurrences of this issue? Now that can be a tough issue - such as that was not a recurrence of a knee injury since the first one happened to a left knee, and the second to a right knee.

And waiting for a recurrence can be very lagging, and you likely want some assurance that you won't have a recurrence.

So next is to look at (and I did this as a Contractor Assurance person) - an independent review of the causal analysis and "grading" it for completeness. Grading did the corrective actions planned reasonably address the identified cause(s). And finally, did the corrective actions as documented in the closure meet the specifications of the plan.

Simple "Management by Wandering Around" tours can also spot if the corrective actions took hold, and analyzing "close calls".

Finally, I'd suggest comparing it to taking your car into a repair shop for a repair. How do you "know" if the repair was does properly? Use the same logic as how you answer that question.

 

[Randy]

I am looking for a fresh perspective.

Fresh perspective? How about.........Does the corrective action work the way we wanted it to work? YES or NO

History is full of examples of effective corrective action, just snag one