Explanation on definition CA/PA directly from ISO or FDA

[01mercy]

@Bev D thank you for your comprehensive input.

I do agree in principle that it is not about the words but about that the right actions are being done to fix and prevent a problem.

However we live with the reality that we have to write down processes (and lower lvl docs) for people to be able to execute work in a reproduceable way in order to have reproduceable results.
In this context, correct and understandable definitions are important. Because if it is not understandable people can interpret it in different ways and that will create unwanted deviation.

Many of the CAPAs I have supported,one of the causes is along these lines that the content of the documentation needs to be revised in order to be more clear about what is exactly meant, including definitions.

 

[Bev D]

@01mercy - I was actually responding to Tidge’s last post. We are in 100% agreement that words do matter in getting people to do the right thing. As I stated, I have seen many people stop their investigation into teh causal mechanism because they made corrections (rework, repair or scrap) and they believed that that was sufficient. That corrections = corrective actions.

P.s. there is NO such thing as a “CAPA”. This is an old reference to the two separate requirements for Corrective Action (after the defect occurs) and Preventive Actions (before defects occur) and only adds to teh mis-understanding of Corrective Actions to prevent reoccurrence. You are either performing a Corrective Action or you are performing a Preventive Action

 

[Sidney Vianna]

The meaning of the term CAPA - Elsmar Cove Quality and Business Standards Discussions

Can somebody clarify with examples the following: Correction Corrective action Prevention Preventive action Or can somebody provide me a CAPA template? Thank you in advance.

elsmar.com

 

[Ed Panek]

If it helps this is a snippet of a recent FDA audit of our QMS regarding CAPA - Edited for anonymity

According to the Employee, there have been XXX CAPA’s generated since the previous inspection. I randomly
selected XX for review with no deviations noted. The quality data sources reviewed were complaints and
nonconforming material reports.

I reviewed the following procedure relating to complaints.
XXX“Complaint Handling and MDR Reporting” revision M

I selected for review XX complaints covering the period since the previous inspection. No deviations were
noted. According to Employee, there have been no MDR’s since the previous inspection.

I reviewed the following procedures for nonconforming material. XXXX

The firm generated a total of XX nonconformance reports covering the previous inspection. I reviewed all
nonconformance reports with no deviations noted.

I reviewed SOP- XXXX “Correction and Removal, Recall” revision D. According to Employee, the firm has not been involved in any recalls or corrections and removals since the previous FDA inspection.

As you can see the FDA is very clinical and precise in how they report. They usually wont dig unless they see problems.

 

[Tidge]

FWIW, my use of CAPA (or CA/PA, or CA, or PA) is 100% informed by the FDA's QSIT. In the QSIT "CAPA" is a system, and the closest that doc comes to the non-FDA standards for corrective actions is IIRC changes of case between "Corrective Actions" and "corrective actions." I'm not trying to redefine any terms, this is/was just another example of the FDA not being 100% aligned with certain other standards, even if they are commensurate with them.

My medical device companies have literally all had a policy on Corrective Actions (an abstract usage for a process motivate by 21 CFR 820), and corrective actions (tangible usage for instances of actions that were taken, a la ISO 9000) are only one subset of activities within that policy. The way the FDA uses these words are not entirely dissimilar from the way they have used the word "validation" in multiple ways.

 

[pseudoazurin]

Hi, I am looking at a 13485 QMS with the CA follows NC …etc like we normally see, but PA does not follow CA, so my first question would be does/must NC lead to PA procedure (may justify to be non-necessary)?

In this system, PA is kind of independent and detached process, and is taken place and generated from Management Review meeting. It seems to be OK, but would like to see what other thinks, whether there is any potential issue or non-compliance to ISO 13485 or 21 CFR 820. Both CA and PA are under 8.5 Improvement, and PA section follow CA section, but I don’t think there is a sequential relationship that PA follows CA, like design output follows design input. Any thoughts?

 

[Bev D]

<Big Sigh>. You are correct to ask about this. First CAPA is a misnomer for the whole corrective action preventive action thing. No one should ever use it.

Corrective Action occurs once a defect or Non Conformance has actually occurred and been detected. Corrective Action is action against the cause to prevent recurrence. This prevention of recurrence is not “Preventive Action” per ISO’s own definitions.

So-called Preventive Action is taken BEFORE a defect or nonconformance has actually occurred and been detected. Preventive Actions are things like mistake proofing, FMEA, DOE, 5S, Single piece flow, sensors, etc. to PREVENT a defect from actually occurring.

“CA” and “PA” are two separate and distinct things. They do not and cannot belong together - like matter and anti-matter. There is no sequential relationship as there is no relationship.

ISO removed Preventive Action from the standard years ago in part because of this confusion.

 

[pseudoazurin]

<Big Sigh>. You are correct to ask about this. First CAPA is a misnomer for the whole corrective action preventive action thing. No one should ever use it.

Corrective Action occurs once a defect or Non Conformance has actually occurred and been detected. Corrective Action is action against the cause to prevent recurrence. This prevention of recurrence is not “Preventive Action” per ISO’s own definitions.

So-called Preventive Action is taken BEFORE a defect or nonconformance has actually occurred and been detected. Preventive Actions are things like mistake proofing, FMEA, DOE, 5S, Single piece flow, sensors, etc. to PREVENT a defect from actually occurring.

“CA” and “PA” are two separate and distinct things. They do not and cannot belong together - like matter and anti-matter. There is no sequential relationship as there is no relationship.

ISO removed Preventive Action from the standard years ago in part because of this confusion.

Thanks, Is there any official record/reference or good discussion about this on record? What is FDA's thinking about CAPA? In general CAPA is consider a single subsystem, rather than 2 subsystem, and for QSIT, it is kind of audited one after the other, of course not that it can't be a 2 independent system.

 

[Bev D]

I have no idea how the FDA views this; perhaps another member can enlighten us.

What is QSIT?

The 'official' definitions are actually part of the last (?) ISO9000 standard. It's been so long that I don't remember. I am sure that one of our esteemed auditors such as @Sidney Vianna or @Randy can tell us.

We of course have answered and discussed this very question many many times on this forum. A search might help you find the more recent postings on this topic. we seem to answer it 2-3 times a year even now...

 

[Sidney Vianna]

There are so many discussions on the lack of compatibility between the term CAPA and the (normative) definitions in ISO 9000. Example. The meaning of the term CAPA - Page 2 - Elsmar Cove Quality and Business Standards Discussions