Suggestions on glassware

[tcay584]

Hello,
Please pardon the lengthy post, but I need some general industry feedback on an issue I am trying to handle with my QA department. We are a GMP compliant laboratory doing both QC and method development/validation work. We use worksheets to provide traceability to all equipment used (HPLCs, columns, ovens, balances, etc) for purposes of identifying cal due dates and qualification/calibration status. A question arose during a review of a validation in which we use graduated cylinders for mobile phase preparation (500mL of each component) and also a volumetric pipet to add 2mL H3PO4. A reviewer questioned us on our glassware. Specifically, the question was "How do I know you used a volumetric pipet to add the H3PO4"? Also, "How do we know which cylinders you used for mobile phase preparation"? Now, I have used certified graduated cylinders for pump flow rates during PQs which had a serial number and certificate of traceability...but not during standard or mobile phase preparation for routine use. Our SOP says to use only class A glassware for standard and sample prep, and so I thought we were covered by that caveat. Our QA department is asking us to "Prove it" and I don't know how to reasonably incorporate that into our lab...or if the regulations actually require that. In our SOPs, we say "Use a volumetric pipet to transfer...." and also "Transfer to a volumetric flask" when the precision of the method warrants it (and all standard preps warrant it). However, I don't even know if I can find calibrated, serialized, 1000mL graduated cylinders. It has been recommended that I consider some form of uniquely identifying each piece of glassware in my lab and reference that identifier in all analyses in the same manner as you would identify an HPLC or a balance. Please let me know how you handle glassware traceability in your labs. I will be using responses to this post in a discussion with our QA group, so please be as explicit as possible as to 1) How you have handled this issue 2) If it really is an issue and 3) any suggestions if it really is an issue that I need to handle but have overlooked. I have worked in many labs, and never faced this sort of question, so I'm a bit lost and am really hoping you all can help me out here.

[grzesiek]

That was never an issue for me

But i must say that tis is really f.cked up man, this kind of questions make me sick

I would handle it this way:
1. because we only have class A certified glass or volumetric pipets etc.
2. because only qualified personel works here

[Peter Apps]

Congratulations ! QA has to find something to question, that they are picking at this means that they cannot find anything else to complain about, so the rest of your QC must be up to standard.

Years ago, when I still worked in regulated labs I was told; specify and document whatever makes a difference, and only what makes a difference.

Apart from robots with bar code readers or continuous video surveillance there is no way to prove that a procedure was carried out as specified. At some stage we have to accept that lab staff can and do follow the instructions as given. You could offer to have each step ticked off on a check sheet, which generates the documentation so beloved of QA but still does not prove that it was actually done.

While not necessarily agreeing that volumetric cylinders are the most repeatable way of mixing mobile phases, having had at least one case where the same volume in different cylinders gave different retention times, individually specifying them is a slippery slope. What happens when the one specified for the method gets broken ?, do you have a validated substitute waiting in the wings ?

Pose the question; what happens if the mobile phase composition is not as specified ? Answer; retention, resolution, and/or peak shape or size will change. So if you have control limits for those parameters you have QC on your mobile phase even if it is mixed in a bucket. And if those parameters do not change despite variation in mobile phase composition (i.e. robustness and ruggedness) then you do not need to specify narrow limits (i.e. individual cylinders) in mixing procedures.

On a purely practical note, if you are mixing two 500 ml volumes (as opposed to making up to a total of 1000 ml) you can just as easily use two 500 ml volumetric flasks. If you want traceability, weigh the liquids and record exact weights, or hit the print button on the balance.

Peter

[krickos]

Hi

I also work under QC/QA with additional development/validation in the GMP area.


We basicly use the same approach as you do, you did not state if you had any other quality of glassware present in your lab.
What we did some +10 years ago was that we tossed all less quality glassware to avoid mixups but unsure if that was related to an audit comment or not.

1) How you have handled this issue

Only Class A metric cylinder etc present in the lab.

2) If it really is an issue.

No I do not think so, especially if you you stick to one grade of quality.

3)

any suggestions if it really is an issue that I need to handle but have overlooked. I have worked in many labs, and never faced this sort of question, so I'm a bit lost and am really hoping you all can help me out here.

Again, can not see this as an issue if you stick to the grade you have and do not mix grades.
To put some numbers on things (joggling memory here a bit so may not be 100% accurate):
Newer grade A 1000ml cylinder to lesser grades: +/- 5-7,5-20ml
5ml Voll pipette A grade to less grade: +/- 0,015-0,022ml

So teorectically a mix/different of grades can have impact on variation.

GMP requirements: Did not find any specific in USP when breifly checking, but the European pharmacopiea states that grade A is to be used for volumetric glass ware.

[DR]

In my notebook, I note exactly the type of glassware used for everything -

eg: 200mL ACN via 500mL grad. cyl. combined with 800mL buffer via 1L grad. cyl. and 2mL TFA via vol. pipet...

Also, if your SOPs state that standards and samples require vol. glassware, explain to QA that mobile phase is neither and that 1) method validation reports should address mobile phase variations that far exceed those introduced by competent chemists using class A (or B or whatever you have) cylinders and 2) NOBODY uses serialized cylinders - and nobody I know (in Pharma) has to write down serial numbers of volumetric glassware as they go through it either (though I know it's out there).

[LC_labrat]

It seems this is more about documentation rather than good science.

"How do I know you used a volumetric pipet to add the H3PO4"?
Should be documented in notebook.

How do we know which cylinders you used for mobile phase preparation"?

If the question was posed due to concern over using a 4000 ml cylinder to measure 500 ml, resulting in an error of measurement due to volume differences. Then either state in the method glassware to be used, or make it a training topic, document the training attach a summary of the training and that is your proof. Further proof is document cylinder size if deemed necessary.

I've never been asked to use calibrated graduated cylinders for mobile phase prep. I would look into the error differences in volume from a calibrated vs non calibrated cylinder.

[tcay584]

Thanks so much to all of you for your replies. I have a meeting set for next week to discuss this issue with QA. LC-labrat, the required volumes for our measuring devices are in the associated SOP for the reason of associated error like you mentioned. We do say, use a 500mL grad cyl to measure 500mL of ACN..... The issue QA has is that unless we can correlate a unique identifier to each piece of glassware, how do they have documented proof that we used a 500mL grad cyl. Basically, they are asking if we might etch an "equipment ID" number to each piece of glassware we use in the lab and document "I used CYL1234 to measure 500mL ACN...." so they could go back and check that, indeed CYL1234 is a 500mL grad. cyl. I have never heard of this being done, and really can't get my head around the fact that at some point you have to trust that your people are in fact educated, trained, and experienced enough to recognize good lab practices. I've taken it so far as to say "How can you prove that just because I wrote that I used MeOH, I didn't actually use ACN"? Other than, as DR pointed out, short of video surveillance, you have to trust your people at some point. The glassware etching is going too far IMHO, and I'm trying to gather enough evidence to show QA that this is 1) Not industry standard and 2) Will raise more questions than it answers insofar as if someone sees this....will they not think "Jeez, how dumb are these people....let's audit harder". Please keep the suggestions coming....you guys are the best resource I have to show what industry standard practices are and I really appreciate your help.

[Rootfa]

A good way of differentiating is to use e.g. 500ml for a measuring cylinder and 500.0ml for a grad flask. This also applies in methods, ie mix 500ml Acetonitrile with 500ml water means measuring cylider accuracy is ok (method development should cover the +/- ranges during validation anyway), and dilute 5.0ml of stock solution to 100.0ml in methanol means use a 5ml volumetric pipette to transfer into a 100ml volumetric flask.

Like the previous poster we also removed all class B glassware from the lab. It makes everything som much simpler.

[lmh]

They're somewhat missing the point.

You can make your staff record some etched number of the cylinder as much as you like, but it's far more relevant to know that your staff really filled the cylinder exactly to the mark (and for that matter didn't hold the cylinder in sticky hot fingers for 10 minutes to mess up its calibration). It's also surprisingly easy to record what you think you ought to have done rather than what you did (quite honestly, but absent-mindedly). In a very record-rich environment, some people begin to regard making records as an almost independant task, unrelated to the thing they're actually recording. Too many unnecessary records can weaken tracking rather than strengthen it, analogous to the way too many safety notices can lead to all notices being ignored.

If the retention times and peak shapes of appropriate standards and controls are within specification, then the method worked, and the mobile phase cannot have been too wrong. That is a large part of the point of including all those controls in the sequence.

[Bruce Hamilton]

Where I worked, we engraved an obvious unique number on all of our A-grade volumetric glassware ( pipettes, volumetric flasks, and measuring cylinders ).

At the time of engraving, we checked the item "matched" A-grade certification. This was in response to an infamous "certified 25ml A-Grade volumetric pipette " that delivered 50ml and existed along with it's manufacturer's certificate, in our lab store. We also had a certified 50g weight that was manufacturer calibrated and certified as 500g.

The acceptance test was not a full calibration, but just several replicates using water at 10, 50, 100% for measuring cylinders, and 100% for volumetrics.

If I recall correctly, there is typically a retest of 5 or 10 years on A-grade volumetric glassware - details in relevant ASTM specifications.

The issue of using glassware correctly ( fill to mark etc. ) is a matter of training - hence auditors want to see relevant staff training records.

I was always surprised at how few analysts knew how to identify pipette calibrations ( Blow Out versus Drain ) or measuring cylinder calibrations ( To Contain versus To Deliver ). Or even recognise that the calibrations are only valid for aqueous solutions.

Some chemists even tried to calibrate To Deliver measuring cylinders as To Contain, and wondered why they failed... That's why there is no point in using volumetric flasks calibrated as To Contain to dispense solutions in place of measuring cylinders.

I would suggest that you will also have to appease your client's quality people, and engraving unique numbers is not too onerous, and once present are easy to enter in workbooks. We used a standard notation of "100 ml [ x ] of water", where x = unique number for calibrated items.

That way, your quality people will be more confident defending you, should you need it.

Please keep having fun,

Bruce Hamilton

[tom jupille]

As far as mobile phase prep is concerned, a validated method should have system suitability criteria attached. If you meet system suitability, your method is suitable for the intended purpose.

By way of support, look at the attachment to this document, which tells FDA people how much they are allowed to tweak the parameters of the method.
http://www.fda.gov/downloads/ScienceRes ... 092147.pdf. You have to be practically a lawyer to parse the mobile phase limits, but it amounts to about 2% -- so long as you meet system suitability.

[Consumer Products Guy]

Agree with Tom, above. Our test procedures state to use a volumetric pipet class A when one is required, and allow use of graduated pipet or grad cylinder when that's permitted. So one just follows our test procedures, and ours are pretty well-written and confusion-free.

[MikeD]

Quite similar questions turn up from time to time on the FDA subforum of http://elsmar.com/Forums/. The OP coud also post there.

The advice often given is to ask the auditor to show the exact clause in xxxx regulation/standard where they are claiming non-compliance.

[smartglasspro]

Ok thanks guys, but are these suggestions are valuable for each country?