Daily automatic pipet calibration check

[Consumer Products Guy]

QA wants us to initiate a day-of-use automatic electronic pipet calibration check to comply with cGMP even though we send these out twice a year for "official" calibration check, and they've never been out of calibration. Does ANYONE already do this, and could comment on how many weighings of water they use, if they have to measure the temperature of such water using a calibrated thermometer, if they do multiple volumes from the same electronic pipet or just the volume they will be using, and how close the results must be, e.g. less than 2% RSD. Of course a qualified analytical balance must be used.

This type of information might help us to write an SOP (I know no one will want to "donate" their own SOP for this). Thanks.

[Peter Apps]

Can't you use the history of stability to argue that daily calibration checks are not required ? This kind of QA can become an infinite regress - how do you know that the balance is stable ?, how do you know that the certified mass is stable ?, how do you know that the fundamental constants really are constant ?!

This smacks of QA people not having enough to do to keep themselves busy.

Peter

[Axar]

Can't you use the history of stability to argue that daily calibration checks are not required ? This kind of QA can become an infinite regress - how do you know that the balance is stable ?, how do you know that the certified mass is stable ?, how do you know that the fundamental constants really are constant ?!

This smacks of QA people not having enough to do to keep themselves busy.

Peter

The procedure of calibration must be easy and fast to perform - that's for shure!
Do they need it? I guess yes. I can imagine how laboratory recalls all previously issued certificates just becouse one of pipets failed a years calibration... Such fail prejudices all analysises performed from the moment of last calibration.

[H.Thomas]

This type of information might help us to write an SOP (I know no one will want to "donate" their own SOP for this).

Eppendorf "donates" a SOP: http://www.eppendorf.com/script/binres. ... a57efb48f1

A SOP for testing can also be downloaded from Brand: http://www.brandtech.com/manuals.asp

Maybe there are similar offerings from the manufacturer of your brand.

[Peter Apps]

I think that it is pretty well established that the least reliable component of an autopipette is the nut on top of the plunger - so proving that on Monday morning the pipette was within spec does not validate the results for Monday (or Friday) anyway. What validates a set of results is the QC samples / calibrants / system suitability or whatver that is run with the samples, and if these QC samples are prepared using the same procedure - including the same pipette - as the real samples, then surely it follows that the pipette, and everything else including the operator, was functioning properly for that batch. Calibrating a pipette every day then becomes a pointless reduncancy.

Peter

[krickos]

Hi

Well to my knowledge there is no such thing as a cGMP requirement to control you automatic pipetts on a daily basis. Calibration every 6months or so is not uncommon in my experiance.
If there is, I would very much like a reference. However if they are eletronic motor driven I can somehow see a difference compared to manual adjusteble. But at the same time many drug product QCs uses motor driven Hamilton syringe dilutors and I seroiusly doubt they check those on a daily basis.

To extend the discussion further, is your motordriven (for manual titration or automatic) devices checked on a daily basis for accuracy/precision?! Do you daily check you volumetric glass pipett that it complies to the USP class A volumetric glassware requirement? Do not think so.

So from an more overall perspective it seems like your QA department has gone "bananas", as discussed in another thread, at some point, analytical chemists, SOPs, the quality of the equipment, calibration schedules, SSTs/QC samples at some point has to be "trusted" or we will spend more time checking stuff rather than do what we are supposed to do.

[lmh]

OK, I'll admit from the outset I don't work in a regulated environment, but extending the discussion to manual pipettes, I have been wondering what other labs do.

The rules we (locally) must abide by are that pipettes (of all types) should be calibrated once a year by a professional, and we should check them "regularly" to make sure they remain calibrated. The big question is what does "regularly" mean.

My feeling has always been (a) that the most unreliable item in the lab is probably me; (b) "regularly" can be defined by looking at how much work I'm prepared to repeat if something goes wrong.

The yearly calibrations by a professional are useless for QA purposes. They get the mechanics of the pipette overhauled, which is necessary and good, but the fact that a skilled technician has confirmed the pipette is working correctly in his lab, after servicing, before returning it to me, is unfortunately no proof that (a) it was working the last time I used it before sending it to be calibrated, or (b) that I'll get the right value in my lab.

I am much happier to rely on my own measurements to confirm I get the expected results. If a pipette fails, potentially it casts doubt on all results since it last passed. When I first set up our lab's organisation, I chose to do a very cursory check on a monthly basis: 4 weighings each at high and low volume. This meant the most work I'd ever have to repeat was 1 month. It's not a huge burden, but I suspect it's a lot more cautious than many non-regulated labs. 10 weighings with associated checks on temperature and atmospheric pressure seem excessive to me; I'm not bothered about a systematic 2% error, or assessing the s.d. of measurements accurately (given random biological variation it's irrelevant). Ensuring the pipette is correct at a particular temperature and pressure is also of limited use if the lab environment isn't similarly controlled.

Of course standards have slipped rather as time has gone by, and I struggle to persuade colleagues that such checks are necessary. I'd be grateful for any feed-back.

[JGK]

If you are a heavy user of pipettes, frequent testing photometrially (rather than gravimetrically) may be less time consuming.

The last company I worked for used a unit from this company http://www.artel-usa.com/about/default.aspx

[Consumer Products Guy]

If you are a heavy user of pipettes, frequent testing photometrially (rather than gravimetrically) may be less time consuming.

That's been discussed, but our UV-visible spectrophotometer is not cGMP-qualified, that would cost thousands of dollars a year. Our cGMP assays are by GC, HPLC, and a few titrations.

[unmgvar]

i agree with Peter's first comment
where do you stop?
and then after that those same QA people come asking you why nothing get's released and everything is investigated

first of all as suggested use history of validation to show that no complex procedure is necessary
request your QA to do their work and give you a risk assessment of what would they feel is feasible for them as a procedure to perform taking into account all other parameters of deviation of balance and things around the procedure
after they look at you with "eye balls of fish", because they have no clue what you are asking for, since they only look to cover themselves and barely get what you do in the lab anyway

propose the simplest possible procedure
when used for the first time or at the beginning of the day for any of them check 3 points
close to the lower end, upper end and middle, once for each, taking into account for temp of course while weighing.
show R^2 and confidence of the points at a certain limit
do not do anything more then this
tell them it is like the pH check basically.
remind them, historically annual check performance is always good.
if they ask for more, then show them what is done in the automated check for an analytic balance daily or pH

always add to this the amount of time required per check, transform this into an annual result and also show them how much work this will mean is lost and the man power cost to do it
then they can really see cost effectiveness and risk assess,
what they should have done to begin with, if they knew their job

[HW Mueller]

This is an amusing discussion. When I was heading the neurochemical lab of he neurosurgery in Giessen I noticed that the differences among technicians in measuring parameters like glucose (on a special apparatus, not chromatography) greatly exceeded the variations I obtained by testing the pipets. I noticed this only, after introducing a voluntary regular check with commercial testing substances, and through voluntary participation in Ringversuche (interlab tests). The differences diminished when I had everybody remove the pipet from the taken up solution by moving the tip allong the vessels walls above the solution, and by having them pipet onto a dry wall of the analyzing apparatus´cell.
All the validation can not cover for non-professionality.

[lmh]

HW Mueller, Absolutely!

As a young research assistant I struggled with how to deal with new post-doctoral scientists with dubious pipetting skills. A lowly technician can't criticise a PhD! My solution was to check a set of pipettes before post-doc arrived, in absentia, and hide them in a dirty drawer. On the post-doc's first day I would then take them out while the post-doc watched, and say "I'm sorry, I didn't get round to having these serviced before you arrived. They're probably OK, but do check them by weighing some water, and let me know if they're inaccurate. I can always see if we have another set somewhere". This way I could check each new post-doc...

[JGK]

If you are a heavy user of pipettes, frequent testing photometrially (rather than gravimetrically) may be less time consuming.

That's been discussed, but our UV-visible spectrophotometer is not cGMP-qualified, that would cost thousands of dollars a year. Our cGMP assays are by GC, HPLC, and a few titrations.

That's why I included the link to the Atrtel site, their equiment and reagents are used by companies which do require cGMP assays

[Consumer Products Guy]

Update: QA is now relaxing to allow risk-based management. This would be a weekly check, so worst case scenario upon a recheck would be that one week's data would be suspect. They say we can specify in the SOP to use 25C water because there is a 25C calibrated water bath, without having to delve into the details.

I haven't asked them if a week is Monday through Friday, Monday though the next Monday, or exactly 7.00 days from the time of last check. I think we can do 4 weighings at each of low, middle, and high ranges of the electronic pipet, and everyone will just use the same unit during that week for stuff that affects the data (like a sample aliquot, not addition of a reagent).

[Juan2011]

[quote="Consumer Products Guy"]QA wants us to initiate a day-of-use automatic electronic pipet calibration check to comply with cGMP even though we send these out twice a year for "official" calibration check, and they've never been out of calibration. Does ANYONE already do this, and could comment on how many weighings of water they use, if they have to measure the temperature of such water using a calibrated thermometer, if they do multiple volumes from the same electronic pipet or just the volume they will be using, and how close the results must be, e.g. less than 2% RSD. Of course a qualified analytical balance must be used.

This type of information might help us to write an SOP (I know no one will want to "donate" their own SOP for this). Thanks.
I do agree with some of the guys opinions. Many QA do not know what do we do and whats the impact of a particular task. They just tell us to do it as they are in QA. It is sad sometimes, they better learn some scientific data handling knowlege also. For them , keeping everything neat and clean and documented is called QA. I bet, so many QA guys do not know how the rsults are calculated. they expect clean calculation sheet and good result, rest do not maater ha ha ha
with my experience in bioanalytical labs where regulations are strick, we check the pipettle at three points, lower medium and high , each point 10 replicates using water. convert the mass into volune using formula , density=mass/volume. then check the manual for the pipette for acceptable rsds. Usually we follow manufactures given RSD. Doing once a month is enough. I never had problem with MHRA &FDA . Dont just do what QA says, justify your action with explanation ortherwise life will be hell when you need to calibrate or check for everytime you use ur pipette ha ha ha all the best