Devices from IQ, OQ or PQ process to be used for verification, validation and summative?

Given2Fly

Registered
Hi there,

I've been trying to find some clear FDA guidance on requirements around devices to be used for verification and validation testing but I still seem to be at somewhat of a loss, so I hope you helpful folks can shine some light on it for me...

We are currently starting formal verification of a ventilator device and we will be using devices that have been manufactured in a "Verified" manufacturing process (IQ, OQ) but not a validated process (PQ). I have read in 21 CFR Part 820: 820.30(g) that design validation should be run on "initial production units, lots or batches or their equivalents". Does this mean that devices used for validation need to come off a validated (PQ) manufacturing process?

  • So, would it be deemed acceptable to run validation with devices coming from an non-validated process?
  • And how about summative Human Factors studies? should that also be run with devices from a PQ'ed process?

Secondly, I am also curious about making changes to production equivalent devices during verification testing:
  • If we find the need for, say, a redesign of a mechanical component to meet a design requirement and we need to upgrade some (or all) devices under test with that change, can we do that in our test lab or do we need to send devices back to the CMO to have them rebuild the devices in a verified mfg environment?
Any help on the above would be much appreciated!
 
So, would it be deemed acceptable to run validation with devices coming from an non-validated process?
It depends on the type of design validation required and the associated risk. As an example, let's say your design validation does not put any patient at risk or it does not require a clinical trial, in that case, there is no need to do process validation before design validation. You may redesign the device based on the validation activities, so it does not make sense to put time and effort into process validation until you are sure the device meets its intended use in the current design.

On the other hand, if you have a high risk device that requires a clinical trial for design validation, it makes sense to validate certain processes prior to design validation. For example, your sealing process needs to be validated for a sterile device prior to starting a clinical trial.

I know people speak of V&V (verification & validation) lumping everything into one. In the typical progression, you do design verification first, then design validation and process validation only after all design verification requirements are met. The reason for this is that design verification results may indicate a redesign is necessary.

And how about summative Human Factors studies? should that also be run with devices from a PQ'ed process?
Again it depends on the risk and the type of summative study. Keep in mind that not every manufacturing process requires validation--only those processes where the output cannot be verified. Many times, manufacturers will perform process validation to reduce the amount of quality control activities or reduce the scrap rate even where the process output can be verified. These process validations are not required to be done prior to design validation.

If we find the need for, say, a redesign of a mechanical component to meet a design requirement and we need to upgrade some (or all) devices under test with that change, can we do that in our test lab or do we need to send devices back to the CMO to have them rebuild the devices in a verified mfg environment?
It sounds like you are talking about rework of existing devices. It is not clear how you would update your device history record to show the rework activities if they were to take place at the test lab.
 
When changing out a component change does that call out for a project
When you make any change to your device, you need to evaluated whether the change affects performance, safety, or use of the device and determine if any testing needs to be repeated. Example: If I must change a medical device material, certain biocompatibility tests may need to be repeated.
 

Given2Fly

Registered
Hi indubioush- thanks a lot for the explanation- This helps. Could I ask for some clarification on the following:
there is no need to do process validation before design validation
Do you know if this is specifically called out in any FDA guidance documents or standards?

So, it sounds like your advice it to finish design verification before starting mfg process validation (PQ). We are planning to do parallel path due to project timeline pressures (for a change!) and manage change with regression if needed but, again, is there a formal document from FDA that lays down any constraints around the order in which these activities are done?

It sounds like you are talking about rework of existing devices. It is not clear how you would update your device history record to show the rework activities if they were to take place at the test lab.
To answer your question, my thoughts are: if we need to redesign based on an anomaly found in verification, we would replace/modify a component in all or some of the devices in the lab (we will build a lot of samples for verification). Then we would update the device config file (a tracker of deltas with respect to the device history record), so the DHR stays the same, but we would have amendments. Do you think that would work? What is the alternative?

Thanks again for the help.
 
Do you know if this is specifically called out in any FDA guidance documents or standards?
The FDA guidance document does not really discuss the relationship of process validation and design validation. Regarding this subject, here is the wording from GHTF/SG3/N99-10:

"The interrelationship of design control and process development may, for some technologies, be very closely related. For others the relationship may be remote. The product should be designed robustly enough to withstand variations in the manufacturing process and the manufacturing process should be capable and stable to assure continued safe products that perform adequately. Often this results in a very interactive product development and process development activity."

In addition, ISO 13485 requires that production software be validated prior to use and that sterile barrier systems be validated prior to implementation. Other than that, it is up to you as the manufacturer to determine whether your any of your production processes require validation prior to design validation.

Regarding your other question about replacing components in the lab, I can't provide guidance because I don't know anything about your device, which test you are doing, etc. Also, I am not familiar with having amendments to a DHR. Sorry I can't help here.
 
Top Bottom