If the final yield does not meet with the acceptance criteria, then in that case we have to fill the deviation or not?
Yes
⊠You fill deviation & find root cause & take CAPA to prevent this type of mistake again
Yes you have to initiatief deviation
I am asking for process validation batches.
What would be your rationale for not issuing a deviation? Wouldnât be better to do a root cause analysis to identify why the yield did not result as expected in those PV batches?
These are commercial lots manufactured for process validation, which did not comply with an acceptance criteria. Those batches nust not be released until process validation is approved. PV cannot be approved without complying with all acceptance criteria; hence, a deviation is expected to be issued to evaluate impact on PV batches (commercial product) and root cause analysis.
1 Like
Sir, sampling was done at different stages for PV batches. So, in many products the yield does not meet the acceptance criteria. If we are regularly manufacturing similar products, no deviations were found and all the results were found with limits.
So what should we do in this case?
A common practice is if the yield is expected to be lower, then the acceptance criteria is modified accordingly; or yield is documented only as reference until enough data is available to define actual specification.
If you already got a yield out of specification because PV protocol was issued that way, then I would suggest to follow your SOP accordingly and issue a deviation where you evaluate impact on product quality. Investigation would conclude that there is no impact on product quality and can be released, since the root cause was due to additional sampling taken for validation purposes.
So in other words, even if you know what caused your out of spec, and there is no impact, you must follow your SOP and document deviations as for GMP requirements.
After completion of the batch manufacturing, yield reconciliation to be calculated to find out the exact yield.
i.e; in this case you have to consider the QC sampling quantity also to be included in obtained yield. Then there is no question of yield loss i.e, OOS .
This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.