What is the recommended validation approach for an existing manufacturing equipment with the following details:
Has a controller with basic HMI
Old equipment which is not CFR part 11 compliant
The equipment has multiple access control and levels
No E-records or E-signatures available
No audit trail
We perform the traditional mechanical validation (IQ/OQ/PQ) with no special focus on the computerized validation or the CFR compliance.
Is this approach acceptable by the FDA?
[quote=eliasmorany]What is the recommended validation approach for an existing manufacturing equipment with the following details:
Has a controller with basic HMI
Old equipment which is not CFR part 11 compliant
The equipment has multiple access control and levels
No E-records or E-signatures available
No audit trail
We perform the traditional mechanical validation (IQ/OQ/PQ) with no special focus on the computerized validation or the CFR compliance.
Is this approach acceptable by the FDA?[/quote]
First a little background. CFR is “Code of Federal Regulations.” There are many ‘titles’ under the CFR, 21 is for Food and Drugs. Under Title 21 is Part 11 for electronic records and electronic signatures (and Part 820 is for Quality Systems, another commonly-cited Part under Title 21).
Presuming your manufacturing operations are for medical devices, the primary regulation would be Part 820, Subpart G, section 70, Production and Process Controls. That drives validation / qualification of manufacturing processes / equipment.
Part 11, in and of itself, does not impose any ‘new’ regulations. If there is a “predicate rule” that requires a record and/or a signature and if you choose to create/manage that record electronically and/or apply the signature through electronic means, then Part 11 applies.
Given all that, you indicate no records required by predicate rule are collected and no electronic signatures are applied, so that would indicate that Part 11 is out of scope.
Finally back to your original question: how to validate. The IQ/OQ/PQ approach is more of a de facto approach rather than a prescribed one. But it’s effective and quite acceptable so the approach would be fine. Note that “traditionally” 3 batches / lots were run to pass PQ; however, this is no longer generally acceptable. You need to have a statistically sound justification for the number of runs / samples to pass PQ.