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

[Bill Ryan - 2007]

I'm trying to determine if we even *have* any system(s) that constitute as PA (which I now know we do).

Just a quickie response (one of our "embers" is about to explode into a full-fledged fire )


Do you have a Preventive Maintenance Program? I, at least, don't see any quarrels with this being a Preventive Action.

Bill

 

[RoxaneB]

... and I know the standard calls for separate documented procedures...

It does? Where does it state that they need to be separate? As long as the process is documented, that's all that matters, right?

, but when the rubber meets the road, does it really matter whether the action is one or the other? Both terms are sub-clauses of the overall clause of IMPROVEMENT (8.5). At the end of the day, if I 've helped correct or improve or prevent anything, I could care less which one it should be applied to.

Do I care about the application? Yes. Why? Because my organization has a Business Management System and there are reasons why we use the tools that we do.

If you do not have the proper application, how do you propose to get buy in from the users of the BMS? If you do not have buy in, how do you propose to have an effective system? If you do not have an effective system, how do you propose to truly move from a reactive to a proactive approach?

A process for the application of corrective/preventive/improvement based activities needs to be established and formalized. There can be no improvement, however, without standardization.

 

[Bill Ryan - 2007]

I stand corrected. There is no shall that they be separated (but why did they separate them as sub-clauses?).

If you do not have the proper application, how do you propose to get buy in from the users of the BMS? If you do not have buy in, how do you propose to have an effective system? If you do not have an effective system, how do you propose to truly move from a reactive to a proactive approach?

I don't go to my boss and ask him whether I should approach an issue from the corrective or preventive point of view. I go take care of the issue with the "Quality Tool(s)" I have been trained in, and send the report on to him. He is the administrator of our QMS. I don't care which column he places the action, with its documentation, under. It's not worth the discussion if I don't happen to agree with where he puts it.

A process for the application of corrective/preventive/improvement based activities needs to be established and formalized. There can be no improvement, however, without standardization.

You lost me there (but I'm normally pretty set against "absolutes"). Are you saying that if I tackle an issue using a tool that is not "standardized" within our QMS, I have no chance of improving/correcting/preventing something?

Bill

 

[RoxaneB]

I stand corrected. There is no shall that they be separated (but why did they separate them as sub-clauses?).

Perhaps to distinguish between their functions and applications; to show that they have different reasons for occurring?

I don't go to my boss and ask him whether I should approach an issue from the corrective or preventive point of view. I go take care of the issue with the "Quality Tool(s)" I have been trained in, and send the report on to him. He is the administrator of our QMS. I don't care which column he places the action, with its documentation, under. It's not worth the discussion if I don't happen to agree with where he puts it.

If that is the system that you have, great! But note that in the end, the issue is classified as either Preventive or Corrective or whatever terminology your organization uses.

So how does your approach differ when you take either the corrective or preventive point of view?

You lost me there (but I'm normally pretty set against "absolutes"). Are you saying that if I tackle an issue using a tool that is not "standardized" within our QMS, I have no chance of improving/correcting/preventing something?

How can you improve upon something that is not standardized? One could argue that the first attempt to standardize is, in itself, a form of improvement. Chicken and egg argument.

Try this...

You have an accepted way of doing things with an accepted set of requirements for the output. That is a process. This is your standard. This is how you do it and this is what you get when you follow the process.

Suddenly, you start doing things differently from the norm or, for whatever reason, your output is not meeting the requirements it is supposed to. This calls for Corrective Action.

Or, Option 2, things are still okay, but the numbers are showing a distinctive trend to the outer limits. Preventive Action.

The results of either scenario could result in:

1. The abnormality addressed and the standard returned to; or,
2. The abnormality addressed and a new, better way standard discovered (also called improvement).

In the end, Bill, what we seem to have here is a difference within our systems. What works for you is great! What works for me is great, too! Again showing that even though ISO 9001:2000 provides us with the requirements to meet, we still find ways to make our systems unique.

 

[db]

No such animal

I contend that there is no such thing as a preventive action. All preventive action is based on what we know has happened before. We eat right and get plenty of exercise because experience shows that the opposite is bad for your health. Sure, it might prevent a heart attack, and I've never had a heart attack, but it is based on others having a heart attack. What it does is to prevent another heart from failing. The same thing is true with preventive maintenance. We perform it because we know that without it things fail. We are still reacting to something failing, it is still corrective in nature.

Okay, with that said, what do we do? How can we meet 8.5.3, if there is no preventive actions out there? Well, we cheat a bit. If a failure has not happened to a particular process, product, activity, or equipment, then anything we do things to prevent failure. We can call that preventive action, should we choose to. Remember, the standard allows you a lot of flexibility. Use it! SPC can be preventive, or corrective. It all depends on how you define it.

It is your QMS/EMS, and you make the rules.

 

[David Hartman]

I contend that there is no such thing as a preventive action. All preventive action is based on what we know has happened before. We eat right and get plenty of exercise because experience shows that the opposite is bad for your health. Sure, it might prevent a heart attack, and I've never had a heart attack, but it is based on others having a heart attack. What it does is to prevent another heart from failing. The same thing is true with preventive maintenance. We perform it because we know that without it things fail. We are still reacting to something failing, it is still corrective in nature.

Okay, with that said, what do we do? How can we meet 8.5.3, if there is no preventive actions out there? Well, we cheat a bit. If a failure has not happened to a particular process, product, activity, or equipment, then anything we do things to prevent failure. We can call that preventive action, should we choose to. Remember, the standard allows you a lot of flexibility. Use it! SPC can be preventive, or corrective. It all depends on how you define it.

It is your QMS/EMS, and you make the rules.

Dave,

Would it be "preventive" if I have a stable controlled process, but I implement a change that will make the process more robust (i.e. reducing variation around the norm)? Or, have I "corrected" the process by eliminating variation (even though the previously established limits were still conforming to spec)?

 

[db]

Dave,

Would it be "preventive" if I have a stable controlled process, but I implement a change that will make the process more robust (i.e. reducing variation around the norm)? Or, have I "corrected" the process by eliminating variation (even though the previously established limits were still conforming to spec)?

Why would you want to make the process more robust? Because data suggests that a more robust process will reduce variation. Variation is the cause of nonconforming product. Your action would be corrective in nature. Now, I sure you realize that I am engaging in purely philosophical banter. In the real world, But if you really analyzed it, you would probably agree that most of the preventive things we do are nothing more than corrective.

BTW, I vote for two separate procedures because I think the scope would be different. And this coming from a guy that doesn’t believe in preventive action.

 

[David Hartman]

BTW, I vote for two separate procedures because I think the scope would be different. And this coming from a guy that doesn’t believe in preventive action.

It's a good thing then that I voted for one, just to "prevent" yours from counting.

BTW: I do understand where you're coming from and can see that virtually any action could be interpreted as corrective. So in effect my vote has actually "corrected" your voting error.

 

[db]

It's a good thing then that I voted for one, just to "prevent" yours from counting.

BTW: I do understand where you're coming from and can see that virtually any action could be interpreted as corrective. So in effect my vote has actually "corrected" your voting error.

Or did I correct yours!

 

[RoxaneB]

Dave,

Would it be "preventive" if I have a stable controlled process, but I implement a change that will make the process more robust (i.e. reducing variation around the norm)? Or, have I "corrected" the process by eliminating variation (even though the previously established limits were still conforming to spec)?

Now that the Dave's are done correcting or preventing themselves...????...., within my organization's system, well, we'd disagree with db's comment that this is corrective. Why? Because you are still conforming to spec, even though you have variation around the norm. You are conforming, that means you are meeting requirements.

The activity of making the system more robust would be Improvement at my organization because we a standard process that is apparently stable and now we wish to reduce the amount of variation.

A preventive action would be when the difference from the norm was taking a downward trend to one of the limits. Not yet out of spec, but apparently heading in that direction. PA would entail maybe changing some equipment that has worn out, or getting a cup of coffee for the guy taking measurements for the samples...you get the idea.