Wednesday, July 8, 2009

Verifying or Validating CAPA

The topic of verifying or validating CAPA is a very important step in the investigation process and one that is often overlooked. Many investigations will adequately investigate the nonconformance, determine the root cause, may identify actions needed to correct and prevent the recurrence of the nonconformance, but will often fall short of the mark when it comes to the verifying or validating the corrective and preventive actions to ensure that the actions taken are effective.

Verification or validation of a corrective and preventive action is mandated by the quality system regulations as defined in 21CFR820.100 (a) (1) (4):

Verifying or validating the corrective and preventive action to ensure that
such action is effective and does not adversely affect the finished device;

What this means is that once you have clearly identified the actions that you need to take in order to correct and prevent this problem, you must take it one step further and verify or validate that the action(s) will in fact work to fix the issue at hand. Why is this an important step in the investigation process?

Well, first of all, investigations are to have what I like to call a “defined lifecycle”. By that I mean:
(1) Identify the nonconformance.
(2) Contain the nonconformance. That is, immediately identify how you are going to contain this nonconformance to ensure that nonconforming product does not enter the marketplace.
(3) Next, you will investigate the nonconformance to identify the root cause
(4) Identify and implement certain actions to correct the nonconformance
(5) Finally, you must validate or verify that those actions will work and will not adversely affect the product

This “defined lifecycle” is supposed to be part and parcel of a clearly defined process that will not only investigate the issue at hand, but also to ensure that when investigated correctly, you should not have a recurrence of this same issue. Hence, the importance of verification or validation of the actions identified in the investigation. Where there is no verification or validation that the actions will work, then the likelihood of a recurrence of this same issue is not only probable, but inevitable.

An investigation is supposed to be a finite event that leads you not only to identify the root cause, but whatever actions are taken should prevent a recurrence of the nonconformance. If it does not, then you can rest assured that there has been a break in the “defined lifecycle” and the nonconformance is likely to repeat itself.

No comments:

Post a Comment