Operational Qualification and Performance Verification OV PV

[Jumpshooter]

Do you'all do this on a regular basis or have it contracted to be done?
The reason I'm asking is because this is what I am being tasked to facilitate now. And, from what I can gather by reading up on OQ PV, these processes are NOT cheap nor easily done--usually requiring an external party to perform them.

To my knowledge our GC and HPLC has not had this OV PQ; it was just set up by me and I started to run STDs to assess suitability and linearity--then after being satisfied with its performance on the STDS I began to run test samples and have been obtaining good results (and I must acknowledge that the kind and knowledgeable assistance of fellow Chrom Forum writers has helped me to successfully troubleshoot problems when I encountered them).

If your instrument is ''working fine" (i.e., giving good/expected results) then conventional wisdom says that "OQ and PV" are subsumed. But is this assumption kosher or no? Do you'all do this (or have it done) and how frequently and at what co$t?

Thanks!

[JGK]

For the operational qualification OQ, you need to prove that your GC is doing what you commands tell it to do, for example. If you tell it to inject vial No. 3 five times, does it do it ?

You could start the OQ from the ground up ie:
Does the inlet/oven/detector go to the correct temperature?
do the gas flows set correctly( if they are set electronically) ?
Does the autosampler function as set?

For performance qualificaltion (PQ) you could test:

injection reproducibility at different volumes
detector linearity

Its probably best to check the manual to see if there's any other recommended tests.

[Jumpshooter]

JGK: Thanks for your thoughtful reply. So, I could put together a reasonable OQ and PV plan myself then? From other readings I've seen, it appears that a "third dis-interested party" need to conduct the OV and PV in order to lend credence/validity to the process. Do I have the proper understanding here---or---is that just a money making ploy by outside vendors of OQ PV services? The price quotes I have seem to be between $3K to $4K USD. Furthermore, the items they intend to check for OQ and PV are impossible for me to check. Such things as: detector flow accuracy test, inlet pressure decay, hydrogen cut off safety, detector signal noise and drift, etc...

[GaryR]

If you work in a regulated environment you definitely need a documented procedure for all aspects of instrument validation.
The IQ/OQ should be performed by a qualified vendor technician and fully documented.
Ongoing performance verification can be performed in-house using SOPs written around the tests performed during OQ. You shouldn't need to go to extremes like detector gas flows (IMHO) as your basic tests such as injector reproducibility and detector noise and drift will give adequate assurance that those parameters are functioning adequately.
For our GCs we do injector reproducibility, detector sensitivity (minimum area of reproducibility soln), oven temperature set-point from 50C up to 300C, detector noise and drift.
For HPLCs we also do injector linearity, column and sample compartment temperatures (where appropriate), wavelength accuracy, gradient pump proportioning valves, and some vendor specified diagnostics.
It is not necessary to use the vendor tests exactly, as long as you document your reasons sufficiently, and you cover all critical control parameters.
Even doing these tests in-house you are looking at a considerable cost as you need to factor in reagents and consumables and analyst time/costs.

[Ron]

In many cases if there is a possibility that results of the testing could become evidence in a legal action or regulatory actions it is a good idea to have an outside company perform the initial performance verification. It gets kind of shaky when the answer to the question of how you knew whether the instrument was performing properly is "We tested it ourselves". That can open up a whole can of worms.

[Consumer Products Guy]

We have an outside contractor do this each year. Yes, it is costly.

[Jumpshooter]

Well it appears that the concensus is: "get an outside agency/vendor to do the operational performance assessment of our instruments". I have no real reason to believe that my instruments (and HPLC and GC) are NOT performing as they should be given that all of my runs tend to "pass" and that there is satisfactory system suitability and such. But I am not naive and recognize that an outside inspector (e.g. FDA audit) could presume that I am "making them pass" (shudder the thought!). To date I have gathered quotes and they are $3,400 to $4,050, and six to eight hours are required by the contract engineer to conduct the testing. We will be issued a CD Rom that specifies the "passing of certification". I guess I could frame it and attach it to the ALS tower.

[Consumer Products Guy]

We have an outside contractor do this each year. Yes, it is costly.

[Jumpshooter]

Thanks all, apparently this is a necessary thing (to have an outside "dis-interested party" come in and put my instrument thru OV PV). I will take this quotes for service to our Purchasing Manager--maybe bring fresh donuts to assuage the process as she will likely fret when I submit my p.o.'s. We are a small operation and I am the "lab guy" so this is why I judiciously consider these things; you'alls commentary has been instructive--if not humbling.

[bhahn]

All of the above is correct. We also do this for people. Shoot me an e-mail at bhahn@aoti.net let's chat!

[tcay584]

Well it appears that the concensus is: "get an outside agency/vendor to do the operational performance assessment of our instruments". I have no real reason to believe that my instruments (and HPLC and GC) are NOT performing as they should be given that all of my runs tend to "pass" and that there is satisfactory system suitability and such. But I am not naive and recognize that an outside inspector (e.g. FDA audit) could presume that I am "making them pass" (shudder the thought!). To date I have gathered quotes and they are $3,400 to $4,050, and six to eight hours are required by the contract engineer to conduct the testing. We will be issued a CD Rom that specifies the "passing of certification". I guess I could frame it and attach it to the ALS tower.

I guess I disagree with everyone. If you have a solid procedure, and document your qualification activities thoroughly, there's no reason for anyone to suspect you are "making something pass". I fall on the side of the fence where I find it more valuable to be able to demonstrate to an auditor that the people running the equipment have the education, experience, and understanding to produce valid results. If you know your equipment, you should be able to answer any question an auditor may have. The final responsibility for generated results lies with the analyst....not some technician who has no idea of what your lab does, what you analyse, etc. Also, the testing is not complex nor costly at all. All the parameters and criteria are in the manuals. Go to a training course for the cost of one PM/PQ, then you bring the techniques in house.

[Consumer Products Guy]

Adding on to this thread: does your management allow you to repair your own equipment (e.g. diagnose/replace an Agilent active inlet valve cartridge, or change piston seal) and just document such in the maintenance log and be OK for cGMP, or require you to use a "legitimate" outside engineer? We would do our own check of the repair (e.g. overlay multiple injections, run system suitability each Sequence).

I even think a "not-qualified" module could be substituted in for a few days while repair is made, as long as one documents this and runs system suitability with each sequence. What do others think? Not everyone has a zillion qualified systems around...

[GaryR]

I concur.

The bottom line is: document, document, document.

As long as you document what you do, and there is sound scientific reasoning behind it, you should be OK.
Who's to say that the vendor qualified technician didn't get something wrong? We routinely run our internal PQ procedure following any service or vendor supplied PM.
Ultimately the buck stops with you, or your management, so if your quality system is up to scratch you will be OK.

[Jumpshooter]

Clearly there are two polar (no pun intended) views on this matter.
Both have their merits and concerns. I will float these options to upper management and see how we may proceed. Thanks.