Developing our custom test methods and test equipment

[ga2qa23]

Hi All, I have a prototype Class 2 medical device (handheld battery-powered instrument for surgery). My team and I are trying to develop our own test methods and test equipment. I would greatly appreciate any and all advice/resources you can point me towards.

Custom Test Method = custom-made "implants" to allow us to measure post-operative results better in a CT Scan. We don't follow any FDA-recognized consensus standard. At most, we base our test methods on several scientific articles. I only read the FDA Guidance on Bench Testing and it wasn't helpful: https://www.fda.gov/media/113230/download. All it says is to provide "justification for the methods used to perform the testing" and "clinical/scientific/engineering basis for the acceptance criteria".

• Does our test method need to be 'validated' first? Do we define our own test protocol and test report specifically for validating our test method?
• If so, what sample size should we use for 'validation' of the test method? (Do we need run a test with 3 'test samples' (known dimensions) through the CT Scanner, and confirm that our measurements are accurate?)

Custom Test Equipment = custom-made test jig, to test each "finished" product unit coming off the assembly line, to confirm that the sensors work properly. It's simple to make but it's very unclear how to handle the development, manufacturing, and 'validation' of the test jig itself.

• Should we have the CAD prints of this test jig be documented through our change control process?
• How should we document the production info? (Should it be part of our medical device files somehow?)
• Do we need to 'validate' this test jig first? Do we need to develop our own "IQ", "OQ", and "PQ" for this homemade test jig?
• If we need to 'validate' this test jig, then is a simple test with 3 'test samples' OK?

 

[ECHO]

Test Equipment,
Yes, you should document the design and how to build it. It doesn't have to be to the same rigor as your medical device but make sure you are able to recreate the jig at a later date. The "how" you document is a bit more flexible, I have seen some companies have a mini DDP for their test fixture, some just throw it all into their equipment spec.

The short answer to your "validate" question is "yes". that said that doesn't always mean "test it".
I would first write out the answers to the following questions, "What happens if my jig doesn't work right? what does it mean for my jig to not work right? do I have other fail safes (either in the design or in my process)?"
You should use the answers to the questions above to dictate how thoroughly you want to test it.

Lastly, if you decide to test with "3 test samples" you should write out why you decided to use 3.

All that said, you should have a SOP for this.

 

[ga2qa23]

Thanks for your advice. That makes a lot of sense.

Does DDP = Design Development Plan?

Regarding 'validation', I guess I specifically mean if it requires the same stringent level of "process validation" that FDA expects for medical device components and finished products. Our medical device has a sensor inside that must be calibrated before it can be "finished" in the manufacturing process. The test jig does that final calibration step on the medical device. The test jig is meant to be the final "100% verification check" of our manufacturing process, so that we don't need to validate the manufacturing process. So I guess the test jig doesn't have any fail safes.

Do you know of any online stores that sell a template SOP for this purpose? I unfortunately don't have an SOP for this specifically. I only have SOPs on process validation and statistical techniques.

 

[ECHO]

Does DDP = Design Development Plan?

Yes

Regarding 'validation', I guess I specifically mean if it requires the same stringent level of "process validation" that FDA expects for medical device components and finished products. Our medical device has a sensor inside that must be calibrated before it can be "finished" in the manufacturing process. The test jig does that final calibration step on the medical device. The test jig is meant to be the final "100% verification check" of our manufacturing process, so that we don't need to validate the manufacturing process. So I guess the test jig doesn't have any fail safes.

In the past, I would validate test equipment in a less stringent level and the test equipment would be added to the process validation. I don't know if you are allowed to do that but if you are trying to shift the onus of the "validation" of the manufacturing process to the test jig, I would image the test jig would need to be thoroughly tested.

Do you know of any online stores that sell a template SOP for this purpose? I unfortunately don't have an SOP for this specifically. I only have SOPs on process validation and statistical techniques.

None come to mind but I will look around for you.

 

[ga2qa23]

Thank you so much. Yes, please let me know if you found an SOP for this purpose. I'd happily fork over hundreds of dollars for the opportunity to get started on the right foot. I just can't find any actual guidance / white papers / procedures out there already.

My only other option is to contract a test laboratory vendor specifically for building new test equipment. But that is a lot of cost and I need to do my own research first before diving in.

 

[Tidge]

I don't have my copy of ASTM Volumes handy, but I would recommend review of the ASTM catalogue... it is possible that they have a validated method that will meet your needs, but if not there is very likely to be an area of concern where they lay out the (consensus) approach to developing an innovative test methodology. See the early slides -up to slide 12- of this presentation (I have no affiliation or knowledge of the eventual subject matter) to get an idea how a well-respected standards organization validates a test method. This is a common approach used by other organizations (e.g. ASHRAE).

Ah... and since "test equipment" was also mentioned, custom test equipment can also become part of a consensus standard. In my experience UL standards were notorious for getting "home-brewed" pieces of equipment into their (eventual consensus) standards, but this is not unique to that organization.