Regulatory Open Forum

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

You posted correctly.

First, there is no such thing as CAPA. There are corrective actions and preventive actions, and they never occur together. This is important because the effectiveness check conditions are often different.

I recommend that you have multiple states to track the progress and report to management. I prefer to use "closed" to mean all the planned actions are done and "completed" to mean completion of the effectiveness check (regardless of the outcome). Otherwise there is a long period of inactivity in the metrics that looks like people are not working on the problem when they should be.

For a corrective action, determine the changes needed to eliminate the cause of the detected nonconformity and implement them. The effectiveness check determines if the changes actually eliminated the cause. As such, the effectiveness check is condition based, not time based. As you work on the corrective action, determine the effectiveness check. For example, in a production process nonconformance you may decide that you need three successive lots without detecting the nonconformance to declare the actions as effective.

For a preventive action, determine the changes needed to eliminate the cause of the potential nonconformity and implement them. The effectiveness check determines if the changes actually eliminated the cause. Here the effectiveness is looking to ensure that a problem that hasn't happened before hasn't happened yet. Check that the changes are still place. This is often time based. While working on a project involving a room full of injection molding machines. I noticed that the people were not wearing ear protection, so I raised it with the company owner, pointing out that he probably had a large number of worker's comp claims in his future. We set the effectiveness check to observe worker behavior two-months after the policy went into place.

If the effectiveness check shows the action are not effective, then either open a new action or reopen the prior one. In either case, for the metrics start all over as you work the steps. Write the decision in your SOP, so it is consistent.

------------------------------
Dan O'Leary CQA, CQE
Swanzey NH
United States
------------------------------

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

To expand on Dan's response, one technique that I have used for effectiveness checks for corrective actions is simply how often the defect or error occurred in the history of that product or production method (whichever is appropriate).  If it occurred once every thirty lots, then an effectiveness would be established if it did not occur in the next forty-five lots.  Depending on the risk of the error, this could easily be extended to sixty lots or even more.  In some cases for a very rare error, a time period can be used in order to prevent the effectiveness period from being excessive.

Regards,
James

------------------------------
James Bonds J.D.
Director Regulatory Affairs
Atlanta GA
United States
------------------------------

Original Message:
Sent: 02-Jan-2020 10:40
From: Dan O'Leary
Subject: Determining CAPA Effectivity Check Period

You posted correctly.

First, there is no such thing as CAPA. There are corrective actions and preventive actions, and they never occur together. This is important because the effectiveness check conditions are often different.

I recommend that you have multiple states to track the progress and report to management. I prefer to use "closed" to mean all the planned actions are done and "completed" to mean completion of the effectiveness check (regardless of the outcome). Otherwise there is a long period of inactivity in the metrics that looks like people are not working on the problem when they should be.

For a corrective action, determine the changes needed to eliminate the cause of the detected nonconformity and implement them. The effectiveness check determines if the changes actually eliminated the cause. As such, the effectiveness check is condition based, not time based. As you work on the corrective action, determine the effectiveness check. For example, in a production process nonconformance you may decide that you need three successive lots without detecting the nonconformance to declare the actions as effective.

For a preventive action, determine the changes needed to eliminate the cause of the potential nonconformity and implement them. The effectiveness check determines if the changes actually eliminated the cause. Here the effectiveness is looking to ensure that a problem that hasn't happened before hasn't happened yet. Check that the changes are still place. This is often time based. While working on a project involving a room full of injection molding machines. I noticed that the people were not wearing ear protection, so I raised it with the company owner, pointing out that he probably had a large number of worker's comp claims in his future. We set the effectiveness check to observe worker behavior two-months after the policy went into place.

If the effectiveness check shows the action are not effective, then either open a new action or reopen the prior one. In either case, for the metrics start all over as you work the steps. Write the decision in your SOP, so it is consistent.

------------------------------
Dan O'Leary CQA, CQE
Swanzey NH
United States

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

Hi Lauren,

Every corrective action to a "problematic" situation must lead to a preventive action, in order to avoid this problem to the future. That's why so many people call it CAPA, because they are linked together. Many times, as you can consider, there can be a preventive action without a corrective action. You can see or imagine a problem coming on your way and you try to avoid it.
In my experience, a universal CAPA effective period cannot be determined, because as Dan explained, each corrective action can be extremely different.
A corrective action that can be closed with a simple personnel training can last a few weeks, but an infrastructure corrective action can last months.
I believe you should breakdown the period of corrective actions as Dan advised.

Hope that helps.

------------------------------
Spyros Drivelos
Medical Devices Expert, RAC
Agia Paraskevi, Athens
Greece
------------------------------

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

Hi Lauren,

Yep, welcome and certainly posting questions like that helps to get some different perspectives.  The period of effectiveness checks (and the process for corrections, corrective action(s), and systemic corrective action(s)) may be dependent on the situation, issue observed, and how it can be examined for proper implementation.  This may be a couple lots, may be 2 months, may be 12 months ... this should be clearly stated in the corrective action plan or corrective action summary.  What I would recommend to you is if you set an effectiveness check at 12 months, you do not want 12 months and then check.  Remember corrective actions and preventive actions are a process in the quality system.  So instead of checking once at 12 months, there should be periodic checks over the 12 month period.  And make sure these are documented !  Most of the time I see companies doing activities associated with corrective action and preventive action, yet they do not write anything down.  During an audit, they talk for 15 minutes about everything they have done in the last 3 months, yet on the corrective action or preventive action record there is a big blank space with nothing from 3 months ago.

------------------------------
Richard Vincins RAC
Vice President Global Regulatory Affairs
------------------------------

Original Message:
Sent: 03-Jan-2020 02:06
From: Spyros Drivelos
Subject: Determining CAPA Effectivity Check Period

Hi Lauren,

Every corrective action to a "problematic" situation must lead to a preventive action, in order to avoid this problem to the future. That's why so many people call it CAPA, because they are linked together. Many times, as you can consider, there can be a preventive action without a corrective action. You can see or imagine a problem coming on your way and you try to avoid it.
In my experience, a universal CAPA effective period cannot be determined, because as Dan explained, each corrective action can be extremely different.
A corrective action that can be closed with a simple personnel training can last a few weeks, but an infrastructure corrective action can last months.
I believe you should breakdown the period of corrective actions as Dan advised.

Hope that helps.

------------------------------
Spyros Drivelos
Medical Devices Expert, RAC
Agia Paraskevi, Athens
Greece

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

You say that every corrective action must lead to a preventive action. This is the wrong approach. In English I call this the language trap because colloquial language does not align with the technical terminology.

The first concept is the "detected nonconformity". Eliminating the detected nonconformity is correction and is one of repair, rework, or regrade. Eliminating the cause of the detected nonconformity is corrective action.

The second concept is the "potential nonconformity". It hasn't happened, so it cannot be eliminated. However, eliminating the cause of a potential nonconformity is preventive action.

I recommend reading the definitions in ISO 9000:2015.

Also, GHTF/SG3/N18:2010 Quality Management System –Medical Devices – Guidance on Corrective Action and Preventive Action and Related QMS Processes includes the statement, "The acronym 'CAPA' will not be used in this document because the concept of corrective action and preventive action has been incorrectly interpreted to assume that a preventive action is required for every corrective action".

The term CAPA arose for 820.100 which is named Corrective and Preventive Action. People started to use CAPA as shorthand for the section's name. Notice that 820.100 covers both corrective action and preventive action in the same section. However, ISO 13485:2016 separates them because they are different processes designed to address different types of problems.

------------------------------
Dan O'Leary CQA, CQE
Swanzey NH
United States
------------------------------

Original Message:
Sent: 03-Jan-2020 02:06
From: Spyros Drivelos
Subject: Determining CAPA Effectivity Check Period

Hi Lauren,

Every corrective action to a "problematic" situation must lead to a preventive action, in order to avoid this problem to the future. That's why so many people call it CAPA, because they are linked together. Many times, as you can consider, there can be a preventive action without a corrective action. You can see or imagine a problem coming on your way and you try to avoid it.
In my experience, a universal CAPA effective period cannot be determined, because as Dan explained, each corrective action can be extremely different.
A corrective action that can be closed with a simple personnel training can last a few weeks, but an infrastructure corrective action can last months.
I believe you should breakdown the period of corrective actions as Dan advised.

Hope that helps.

------------------------------
Spyros Drivelos
Medical Devices Expert, RAC
Agia Paraskevi, Athens
Greece

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

The challenge here is that there is no single "right" answer to this. The answer for any given situation should generally be tied to probability and statistics - for instance, if something happened 1 time in 100, how many lots/units/cases do you need to determine with appropriate confidence that the problem was fixed? For something that happens at a high rate, or for which you make/do a lot of, this can be very fast. However, at other times it can be much longer. I once had an event that happened roughly 1 in 1000 so it took a very long time based on the speed that product was used to determine the design fix actually prevented the problem. But there are some devices that get used 1000 times a day, so they would take much less time to be able to verify such a change.

Other items, particularly preventive ones, are harder to base in statistical methodology - how do you determine statistics of something that didn't happen yet? Sometimes you can actually measure "mid-process" to show appropriate reductions, but not always - for instance if you are improving an SOP for, say, calibration timing to prevent opportunities to have an OOC, you may have to "pick" a time or a number of uses of the procedure to show that it is being followed and indeed addressing the original opportunity.

Ginger
​

------------------------------
Ginger Glaser RAC
Chief Technology Officer
MN
------------------------------

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

Thank you everyone for the helpful responses!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

IMO, once the actions are taken, the "CAPA" can be closed. 

I would have a separate procedure to follow up on preventive actions to determine whether they were, in fact, preventive.  I would not try to come up with a specific time period for each preventive action.  As Ginger notes, you can try to work out a statistical argument for how soon/often to follow up each specific PA, but I think that will just drive you crazy.  I would just check them all on the same schedule.  If that is a lot of checking, I might set a different frequency for groups of devices based on the level of harm that would be expected repeat of the event that the preventive action was intended to prevent, e.g., low harm annually, moderate harm quarterly, severe harm monthly.

------------------------------
Julie Omohundro, ex-RAC (US, GS), still an MBA
Principal Consultant
Class Three, LLC
Mebane, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com
------------------------------

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

I'm of the opinion that RA/QA folks spend way too much time debating about when a CA/PA is or can be "closed." Honestly, who cares? Track two metrics - when the work is completed and when VoE is completed. Depending on the system, even when VoE "passes." What matter is "how fast is the work getting done" and "did the action resolve the issue" and the amount of time which is appropriate probably varies depending on the complexity of the issue and investigation needed. The term "closed" just confuses the issue.

This isn't a shot at Julie (though I am tagging on her post) but seriously is a debate I've heard lots of times in meetings and training on CA/PA. I swear, the insistence among too many RA/QA pros (though refreshingly rare on this group) is why so many executives don't view our functions as "critical business leaders." I can just imagine those folks listening to a "when is a CAPA closed" debate and going "honestly this is a waste of time" even when they do care about whether the product works well and that their processes are compliant.

A far more important point was one someone made in this thread - to the extent you ARE doing work on the issue, document it somewhere and take credit for it - far more productive use of time.

G-

------------------------------
Ginger Glaser RAC
Chief Technology Officer
MN
------------------------------

Original Message:
Sent: 06-Jan-2020 17:32
From: Julie Omohundro
Subject: Determining CAPA Effectivity Check Period

IMO, once the actions are taken, the "CAPA" can be closed.

I would have a separate procedure to follow up on preventive actions to determine whether they were, in fact, preventive.  I would not try to come up with a specific time period for each preventive action.  As Ginger notes, you can try to work out a statistical argument for how soon/often to follow up each specific PA, but I think that will just drive you crazy.  I would just check them all on the same schedule.  If that is a lot of checking, I might set a different frequency for groups of devices based on the level of harm that would be expected repeat of the event that the preventive action was intended to prevent, e.g., low harm annually, moderate harm quarterly, severe harm monthly.

------------------------------
Julie Omohundro, ex-RAC (US, GS), still an MBA
Principal Consultant
Class Three, LLC
Mebane, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------

Ginger, no worries, I do not feel shot. :)

I agree, except that I think you may be overestimating executives, perched atop of that proverbial hill down which things flow.  Many of whom treat KPIs essentially the same way, and I think KPIs are where this (and similar notions and behaviors) typically come from.



------------------------------
Julie Omohundro, ex-RAC (US, GS), still an MBA
Principal Consultant
Class Three, LLC
Mebane, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com
------------------------------

Original Message:
Sent: 07-Jan-2020 12:01
From: Ginger Glaser
Subject: Determining CAPA Effectivity Check Period

I'm of the opinion that RA/QA folks spend way too much time debating about when a CA/PA is or can be "closed." Honestly, who cares? Track two metrics - when the work is completed and when VoE is completed. Depending on the system, even when VoE "passes." What matter is "how fast is the work getting done" and "did the action resolve the issue" and the amount of time which is appropriate probably varies depending on the complexity of the issue and investigation needed. The term "closed" just confuses the issue.

This isn't a shot at Julie (though I am tagging on her post) but seriously is a debate I've heard lots of times in meetings and training on CA/PA. I swear, the insistence among too many RA/QA pros (though refreshingly rare on this group) is why so many executives don't view our functions as "critical business leaders." I can just imagine those folks listening to a "when is a CAPA closed" debate and going "honestly this is a waste of time" even when they do care about whether the product works well and that their processes are compliant.

A far more important point was one someone made in this thread - to the extent you ARE doing work on the issue, document it somewhere and take credit for it - far more productive use of time.

G-

------------------------------
Ginger Glaser RAC
Chief Technology Officer
MN

Original Message:
Sent: 06-Jan-2020 17:32
From: Julie Omohundro
Subject: Determining CAPA Effectivity Check Period

IMO, once the actions are taken, the "CAPA" can be closed.

I would have a separate procedure to follow up on preventive actions to determine whether they were, in fact, preventive.  I would not try to come up with a specific time period for each preventive action.  As Ginger notes, you can try to work out a statistical argument for how soon/often to follow up each specific PA, but I think that will just drive you crazy.  I would just check them all on the same schedule.  If that is a lot of checking, I might set a different frequency for groups of devices based on the level of harm that would be expected repeat of the event that the preventive action was intended to prevent, e.g., low harm annually, moderate harm quarterly, severe harm monthly.

------------------------------
Julie Omohundro, ex-RAC (US, GS), still an MBA
Principal Consultant
Class Three, LLC
Mebane, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com

Original Message:
Sent: 02-Jan-2020 10:02
From: Lauren Stover
Subject: Determining CAPA Effectivity Check Period

Hello everyone, I've been a member for a few years but this is my first post, so hopefully I'm doing it correctly.

My company recently opened up several CAPAs and we are having trouble determining how long each CAPA should be left open to verify effectiveness. How would you recommend calculating how long of a period a CAPA's effectivity check should be?

Thanks!

------------------------------
Lauren Stover
Chesterfield MO
United States
------------------------------