Hello all

I am new to this forum but I have already read a few interesting views.

The company I work for uses the 8D methodology in order to record discrepancies and provide corrective and preventive actions. Now I need to find a way to measure the effectiveness of the process. Has anyone come across a similar situation? How do you measure your c.a.p.a. system?

I do apologise in advance if there is a similar thread already. It might be the case of this subject having been discussed in the past.

Thank you

Hello all

I am new to this forum but I have already read a few interesting views.

The company I work for uses the 8D methodology in order to record discrepancies and provide corrective and preventive actions. Now I need to find a way to measure the effectiveness of the process. Has anyone come across a similar situation? How do you measure your c.a.p.a. system?

I do apologise in advance if there is a similar thread already. It might be the case of this subject having been discussed in the past.

Thank you

Welcome to the Cove :D.

Scroll down to the bottom of this page and you'll see a list of related discussion threads. Have a look, and post back if you have questions.

I suggest measuring effectiveness of both actions on the basis of root causes.

The intent of Corrective Action is to find the root cause and apply a fix to it. The effectiveness of corrective action is the degree to which the root cause already fixed, reappears. Often the symptom (customer complaint, inspection failure) has several root causes. The analysis of the corrective action system should focus on which root causes reappearred and/or how the corrective actions were applied. Either the corrective action process identified the wrong root cause or the application and verification of the "fix" was not effective.

In preventive action the analysis should begin with the appearance of the symptom you did not want. Again the analysis should focus on what root cause was most likely and how mitigation was applied.

Note that this approach may be confused by the 8D. The lack of appearance of repeat issues may be the result of enhanced inspection and not accurate root cause analysis.

Hello all

I am new to this forum but I have already read a few interesting views.

The company I work for uses the 8D methodology in order to record discrepancies and provide corrective and preventive actions. Now I need to find a way to measure the effectiveness of the process. Has anyone come across a similar situation? How do you measure your c.a.p.a. system?

I do apologise in advance if there is a similar thread already. It might be the case of this subject having been discussed in the past.

Thank you

Measuring the effectiveness?

Simple!

Corrective action: The absence of recurrences.
Preventive action: The absence of occurrences.


Stijloor.

Not that simple

How many times have "late delivery" reoccurred?

A fault tree analysis will find many root causes for the same symptom i.e. late delivery. The key is which root cause did we try to fix and did the same root cause reappear?

How many times have "late delivery" reoccurred?

Which means that whatever has been implemented before is not effective.

I do not want to trivialize measuring the effectiveness of PA and CA processes, but sometimes there is a tendency to make simple concepts way too complex. Going back to basics (intent and what we're trying to accomplish) brings us back to earth.

Stijloor.

Which means that whatever has been implemented before is not effective.

I do not want to trivialize measuring the effectiveness of PA and CA processes, but sometimes there is a tendency to make simple concepts way too complex. Going back to basics (intent and what we're trying to accomplish) brings us back to earth.

Stijloor.This of course assumes the process has been repeated under the same relative conditions as before, and that the only thing different is the change made through c.a.p.a. If even one variable is different, such as a different shipping company in the delivery process because various shippers are contracted, then one could doubt if a difference is from c.a.p.a. unless the change affects both carriers.

You see? You need to look for performance that the change actually impacted.

This of course assumes the process has been repeated under the same relative conditions as before, and that the only thing different is the change made through c.a.p.a. If even one variable is different, such as a different shipping company in the delivery process because various shippers are contracted, then one could doubt if a difference is from c.a.p.a. unless the change affects both carriers.

You see? You need to look for performance that the change actually impacted.

A detailed in-depth analyis as part of CA would bring this to light.

Stijloor.

I think we are in fundamental agreement. I also opt for the simplist approach. In evaluating effectiveness of anything, we want answers that we can act on. We control only two actions when we perform corrective actions: our ability to find a root cause and our ability to perform a "fix". Reoccurance, as a measure of effectiveness, needs to be evaluated in the context of our actions.

Late delivery is a good example to use as there are hundreds of reasons why a company is never perfect in delivery for any length of time. Root causes are different for each event and one gets frustrated in measuring corrective action effectiveness in terms of "reoccurances". :bonk:

I think we are in fundamental agreement. I also opt for the simplist approach. In evaluating effectiveness of anything, we want answers that we can act on. We control only two actions when we perform corrective actions: our ability to find a root cause and our ability to perform a "fix". Reoccurance, as a measure of effectiveness, needs to be evaluated in the context of our actions.

Late delivery is a good example to use as there are hundreds of reasons why a company is never perfect in delivery for any length of time. Root causes are different for each event and one gets frustrated in measuring corrective action effectiveness in terms of "reoccurances". :bonk:

Chuck,

Excellent points!

As an auditor (I am one too ;)), I am sure that you have observed sloppy CA and PA practices. Most of the CA (rarely PA) attempts that I have observed were merely quick-fixes; a topic extensively discussed here at The Cove.

My approach to this topic may be somewhat "blunt", but it usually gets the message across.

Recurrences are mostly due to ineffective CA. And I do understand that there are exceptions.

Great discussion.:agree1:

Stijloor.

:applause:Right On

Great discussion and most helpful.

Now if we can get the rest of the manufacturing world to listen. :nopity:

:applause:Right On

Great discussion and most helpful.

Now if we can get the rest of the manufacturing world to listen. :nopity:

Chuck,

As long as the organizational "firefighters" get the rewards instead of the folks working dilligently behind the screens to prevent difficulties, things will remain the same. :frust: :frust:

Stijloor.

Which means that whatever has been implemented before is not effective.

I do not want to trivialize measuring the effectiveness of PA and CA processes, but sometimes there is a tendency to make simple concepts way too complex. Going back to basics (intent and what we're trying to accomplish) brings us back to earth.

Stijloor.

Einstein said "things shodul be made as simple as possible but not simpler. Some things are not as simple as we'd like. Let's take another look at what ChuckHughes is saying: there are multiple factors that cause late deliveries. weather, unstable yields during manufacturing, long and unpredictable cycle times at the manufacturer, one of their suppliers is late with a raw material delivery, the planning process did not generate the correct buy order for raw materials, the planning process did not a correct build signal...

Let's say that we issue a corrective action to a supplier for their late deliveries and we/they find the root cause to be that they have unpredictable yields due to a very specific physical cause in their manufacturing process. The supplier finds the root cause and eliminates it, puts SPC monitoring in place for the causal factor. (Corrective action for most QMS systems only require action on the cause not an all of ht ehundreds or thousands of potential causes that might possibly exist) There are no further late deliveries for the next 3 months. Then their supplier sends them some material that is defective and they don't discover it until final test in their final product. A series of late deliveries ensues until this new and different root cause is isolated and fixed.

Was the first corrective action effective: yes.
When the original Problem reoccurs due to the second very different root cause (defective raw material) does this mean that the first corrective action was not effective? no. it's still in place and the cause in the suppliers' manufacturing process has not reoccured...it is effective.

So effectiveness fo corrective actions can't be measured simply by the occurence rate of the PROBLEM. They can be measured by the occurence rate of the CAUSE - two very different things.

Einstein said "things shodul be made as simple as possible but not simpler. Some things are not as simple as we'd like. Let's take another look at what ChuckHughes is saying: there are multiple factors that cause late deliveries. weather, unstable yields during manufacturing, long and unpredictable cycle times at the manufacturer, one of their suppliers is late with a raw material delivery, the planning process did not generate the correct buy order for raw materials, the planning process did not a correct build signal...

Let's say that we issue a corrective action to a supplier for their late deliveries and we/they find the root cause to be that they have unpredictable yields due to a very specific physical cause in their manufacturing process. The supplier finds the root cause and eliminates it, puts SPC monitoring in place for the causal factor. (Corrective action for most QMS systems only require action on the cause not an all of ht ehundreds or thousands of potential causes that might possibly exist) There are no further late deliveries for the next 3 months. Then their supplier sends them some material that is defective and they don't discover it until final test in their final product. A series of late deliveries ensues until this new and different root cause is isolated and fixed.

Was the first corrective action effective: yes.
When the original Problem reoccurs due to the second very different root cause (defective raw material) does this mean that the first corrective action was not effective? no. it's still in place and the cause in the suppliers' manufacturing process has not reoccured...it is effective.

So effectiveness fo corrective actions can't be measured simply by the occurence rate of the PROBLEM. They can be measured by the occurence rate of the CAUSE - two very different things.

Ask what the customer of the process (internal and/or external) cares about....

Stijloor.

Ask what the customer of the process (internal and/or external) cares about....

Stijloor.

OF COURSE they care about reoccurrences of the Problem. But our job is to figure how to improve the occurence rate of the Problems. As an organization there is a huge difference in our actions between reoccurence of the Problem due to the same root cause (inneffective CA) and reoccurence due to different root causes. Is the corrective system broken or are our basic processes broken? The answer generates very different responses...

I've seen guys get beaten up for their inability to solve problems when the Problem reoccurs due to different causes. When in truth they are very good at solving problems and their CAs are very effective. It's just that the underlying systems are not as good as they can be. so new root causes occur. you get pretty demotivated pretty fast when mangement tells you YOU are broken.

On the other hand I've aslo seen inneffective CA systems where the solution is typically a bandaid (discipline the operator, retrain the operator, guess at cause and implement an inexpensive neat solution that doesn't fix anything) and managment encourages 'quick fixes' or ignores it all together - that is also demotivating.

It's all about root cause all the time...

OF COURSE they care about reoccurrences of the Problem. But our job is to figure how to improve the occurence rate of the Problems. As an organization there is a huge difference in our actions between reoccurence of the Problem due to the same root cause (inneffective CA) and reoccurence due to different root causes. Is the corrective system broken or are our basic processes broken? The answer generates very different responses...

I've seen guys get beaten up for their inability to solve problems when the Problem reoccurs due to different causes. When in truth they are very good at solving problems and their CAs are very effective. It's just that the underlying systems are not as good as they can be. so new root causes occur. you get pretty demotivated pretty fast when mangement tells you YOU are broken.

On the other hand I've aslo seen inneffective CA systems where the solution is typically a bandaid (discipline the operator, retrain the operator, guess at cause and implement an inexpensive neat solution that doesn't fix anything) and managment encourages 'quick fixes' or ignores it all together - that is also demotivating.

It's all about root cause all the time...

Excellent points! :applause:

Stijloor.

Thank you all for your replies. :applause:

It has been a real eye opener.

very simple and specific definition.
Measuring the effectiveness?

Simple!

Corrective action: The absence of recurrences.
Preventive action: The absence of occurrences.


Stijloor.

very simple and specific definition.

did you read the posts fololwing this one? as I and others pointed out this method may not be as straightforward as they seem at first reading. Has your perspective on ths been altered or not? if not how not?

Thanks Bev

If it were as easy as restating definitions, everyone would be a consultant.

Time will develop an appreciation for the difficulty in making successful corrective and preventive actions.

Chuck

Thanks Bev

If it were as easy as restating definitions, everyone would be a consultant.

Time will develop an appreciation for the difficulty in making successful corrective and preventive actions.

Chuck

My intent was not to trivialize the Corrective Action Process and the Preventive Action Process. Trust me, I've been around for a very long time to understand and appreciate the challenges. But the OP asked about how do you MEASURE the effectiveness of CA and PA (results of these processes). If these processes work effectively (as intended) then my definition is appropriate.

BTW, I am a consultant for 21 years....;)

Stijloor.

did you read the posts following this one? as I and others pointed out this method may not be as straightforward as they seem at first reading. Has your perspective on ths been altered or not? if not how not?

Apparently it has...See the OP's post (http://elsmar.com/Forums/showpost.php?p=266773&postcount=17).

Stijloor.

sorry Stijloor - I should have been more clear. I was directing my question to eternal_atlas as his response was to quote your first response without addressing any of the points made subsequent to that post...

If we are evaluating the effectiveness of the situation, we set a date (typically 90 days after implementation of the last action) by which the situation should not recur.

For evalauting the effectivenes of the system, my organization tracks recurrences based on variables such as:

Crew
Area
Situation Category
Root Cause Category
...and a few others...


Data Analysis allows to track for recurrences, but also allows us to do a pareto analysis of our nonconformances.

We have established triggers for our nonconformances which guide us on the level of "problem solving" to appy to an abnormal situation. The pareto analysis indicates whether we can tighten our triggers...implying that our system is effective at its current level (i.e., recurrences are diminishing for that particular abnormality)...and we can now intensify - or redirect - our focus.

For example, if we had problem with unscheduled downtime, we might set the trigger for corrective action at all unscheduled downtime > 90 minutes. After one year, we would review the trend with our unschedule dowtime occurrences. An effective CAPA system should - knock on wood - indicate that the number of these situations is decreasing. For the following year, we could set the trigger to all unschedule downtime events > 60 minutes. Theoretically, we would see a jump in CAPA reports for this in the beginning of the year, but the cycle of dimishing numbers would occur throughout the year.

Essentially, we have two means of tracking the efficiency of our CAPA system. One is the tracking of recurrences - the more direct means of demonstrating the efficiency. The other is the tightening of our triggers.

sorry Stijloor - I should have been more clear. I was directing my question to eternal_atlas as his response was to quote your first response without addressing any of the points made subsequent to that post...

Bev,

No problem, I understand!

Stijloor.

did you read the posts fololwing this one? as I and others pointed out this method may not be as straightforward as they seem at first reading. Has your perspective on ths been altered or not? if not how not?

I agree with your perspective that degree of absence of re-occurences matters a lot to ensure the effectiveness.. if it shows a good trend when compared with previous impact and if the root cause needs for a corrective action plan,let us set us an short term objective and monitor closely to see the improvement trend which is a good measure for effectiveness of the corrective action..

as u cited.. discipline of the operator.. CAs may be counselling,training,awareness etc..

effectivness measure : No. of Counselling Session / Training for the operators and No. of occurence of non-disciplinary actions..

I hope i answer your point.. if i misunderstood..let me know..

Hi Meta124
On our end, we measure the timeliness of the following: 1. Due date of return of 8D Report versus the actual date of returned. 2. Commitment date versus the actual implementation date of corrective and preventive action.
Hope this helps.
Best regards,
Raffy

Hello all

I am new to this forum but I have already read a few interesting views.

The company I work for uses the 8D methodology in order to record discrepancies and provide corrective and preventive actions. Now I need to find a way to measure the effectiveness of the process. Has anyone come across a similar situation? How do you measure your c.a.p.a. system?

I do apologise in advance if there is a similar thread already. It might be the case of this subject having been discussed in the past.

Thank you

as u cited.. discipline of the operator.. CAs may be counselling,training,awareness etc..

effectivness measure : No. of Counselling Session / Training for the operators and No. of occurence of non-disciplinary actions..

I hope i answer your point.. if i misunderstood..let me know..

Sorry - it may be a language problem, but I dont' believe you understood my post. what you state as "effectiveness measure: No. of counseling sessions..." is really a measure of how the corrective action was implemented. Effectiveness is a measure of how well the corrective action worked for the stated cause. In this example you cite of a problem caused by lack of operator discipline and your CA of counselling and training, we would measure effectiveness by whether or or not the same problem reocurred due to an 'operator error' or lack of discipline to follow the process'.

It seems you track a lot of data on your CARs. It seems much too complicated for the minimum number of CARs we go through annually. Would you mind posting how many CARs you average in a year?

Here is a simple measure:

Measure Repeat Concerns - so basically if a concern occurs that has been seen previously then the 8D was not effective - as in the preventative action was not effective. Stijloors' explanation above is simple and correct.

Regards,

This is my first time blogging....It's been my experience that nonconformance investigations are quickly closed out to meet a timeline prescibed in an SOP and to release product, which ultimately leads to poor investigations, incorrect root cause(s), and finally ineffective CAs. Take a look at the number of new nonconformance investigations per month and if it's increasing then it's usually a good indication that you're not getting to root cause (assuming you didn't introduce a new product)...if the root cause(s) is incorrect, then your CA will also be incorrect.

Anyone ever consider measuring the reduction of RISK as the measure of corrective or preventive action? If what you did lowers the occurrence or increases the likelihood of detection of a problem then you have reduced risk of the problem. FMEA concept of RPN - severity x occurrence x detection. Of course, you have to have an FMEA for the process or system in which the problem occurs.

Yes....However, you just have to be cautious that people don't select a lower risk ratings so that you have a less complicated Corrective Action, otherwise you won't have a true indication of the effectiveness. I've seen this happen when people are spread too thin.