How I should process previous non-conformities on documents or records?

[newblues]

Hi,

Could you advise me how I should process previous non-conformities on documents or records?

There are cases that you find that you have errors in previous records/documents.
Corrective action may be quite simple, such as to set a clear requirements on SOP not to repeat such error. However, correction is sometimes impossible when you need to search for thousands of previous records in archive and correct them one by one.
Is there requirements or guidelines for such a case?

Thank you.

newblues

 

[John Broomfield]

Hi,

Could you advise me how I should process previous non-conformities on documents or records?

There are cases that you find that you have errors in previous records/documents.
Corrective action may be quite simple, such as to set a clear requirements on SOP not to repeat such error. However, correction is sometimes impossible when you need to search for thousands of previous records in archive and correct them one by one.
Is there requirements or guidelines for such a case?

Thank you.

newblues

newblues,

Kindly describe the nature of the nonconformity that concerns you most.

If the historical record is inaccurate then are the inaccuracies:

1. Significant, if so in what way?
2. Isolated or chronic?

If these inaccuracies are worthy of corrective action, instead of just correction, how was your management system changed to:

A. Stop inaccurate recording (possibly with automation)?
B. Help users to get their records right?

If the inaccuracy is systemic but analog then a global note may suffice provide the record cannot be accessed without making the user aware. If the inaccuracy is systemic but digital then a script may be developed and run to correct all the records with this inaccuracy.

Whatever you decide to do it should be proportionate to the risk and the old record should not be replaced with the corrected record; both should remain visible.

But all this depends on the nature of the nonconformity which is usually expressed as the opposite of the requirement.

John

 

[somashekar]

Records:
You must not resort to records correction in secrecy, but set your records generating process more robust.
The errors in records can be trivial to serious to legal issues and must be carefully assessed based upon the risks. There are methods for records correction that one must follow including records correction approvals. Older the records, less are the chances that you need to look into.

Documents:
When the activities are correct, but the document is not, errors have to assessed and planned revisions and approvals can be made. Again depending upon the risks one can prioritize the correction activity.
When activities are being done per the document, and there are errors, these are high risk areas where a document change activity must be immediately taken up. Again the priority depends upon the risk levels.

 

[newblues]

Hi John,

Thank you.

It's our rework instruction, related to my another thread I posted today.
"Reason for determining no adverse effect on reworked product"

On our rework instructions previously issued, only Yes or No (as a determination for adverse effect) has been stated, and we don't have any reason for determining No.

I can't say it's significant or not because each rework activity varies. In some case it might be significant. And of course it's chronic. Also they are all analog.

What kind of rework was to be done is stated on the rework instruction, so probably reason for each determination can be documented by reviewing previous records. Yes, it depends on how I consider the risk. I'll discuss with my colleagues.

Rgds,

Yoichi

 

[John Broomfield]

Hi John,

Thank you.

It's our rework instruction, related to my another thread I posted today.
"Reason for determining no adverse effect on reworked product"

On our rework instructions previously issued, only Yes or No (as a determination for adverse effect) has been stated, and we don't have any reason for determining No.

I can't say it's significant or not because each rework activity varies. In some case it might be significant. And of course it's chronic. Also they are all analog.

What kind of rework was to be done is stated on the rework instruction, so probably reason for each determination can be documented by reviewing previous records. Yes, it depends on how I consider the risk. I'll discuss with my colleagues.

Rgds,

Yoichi

Yoichi,

Repair is usually more significant than rework.

After all, repair is a design change.

Capturing this decision and its significance usually requires a cross-functional team including the product's designer or design representative.

And yes, it is important to record the reasons for the decisions. You may find that the decision-makers used criteria and that these criteria should form part of the record if not part of the procedure itself.

John