Requires that both the function and calibration status of equipment sent out for calibration be checked per a defined procedure and the results be recorded.

Considering we just sent sent out various types of instruments and equipment out to have that very thing done for us, by an A2LA accredited company, by the way, what does the group think this requirement means? I'm thinking I would rather not do a mini calibration on every pipette and timer we have calibrated.

Tom from Clayton

You don't 'recheck' them when they come back, and the calibration laboratory is supposed to have procedures applicable to the instrument/device they are calibrating. You have to tell the laboratory what you want. For example, some laboratories charge more if you want a copy of the pre- and post- data, for example.

Yes, the laboratory must record results and per the requirements of Clause 4.13.2 (and sub-Clauses) must be applicable to specific tests or calibrations.

However, if you need the results and you should get the uncertainty, you may well have to pay extra for it.

In A2LA's November newsletter the question presented was "Does the requirement in Section 5.5.9 of ISO 17025 apply to equipment that has been sent out for calibration?"

The response from A2LA was: "Yes. When equipment is returned from being calibrated, the lab is responsible for performing a check per a defined procedure and recording the results of that check...."

Section 5.5.9 requires the function and calibration status of the equpment be checked. My calibration person is going beserk and if i knew how to spell apolpletic I would also use that word to describe how she feels.

This is a case of checking the checker but the checker is an accredited calibration lab and I am not.

Tom from Clayton

Hello, Tom!

In essence, you are correct about the perceived approach and your frustration within. :D

Remember, the procedures are about establishing policies and processes; not about dictating specific steps to take.

I have spent many an hour with the technicians I work with on this, getting them to understand the value for this check.

IMHO, the clause is less about checking the checker. Pipettes are sent off for calibration, pistons changed, etc. The check you are being asked to do is less about checking the calibration performed, and more about verifying that nothing has changed with the device during the time it was calibrated and the pipette ending back on the bench.

First, I would suggest you develop procedures for establishing a process that is in control. Check several pipettes first, then back off on the number of verifications and the number of test points. E1154 and the ISO spec both cover this concept.

Next, I would think you would want to check pipettes every now and then anyway, especially with repeatibility/reliability checks between operators; a huge source of error.

My interpretation of that clause in 17025 is more about mitigating risks involved with standards; equipment sent out for calibration to be used for calibration. One does not want to get back a standard thermometer, check 200 instruments, then discover something strange happened when it was first sent back to you.:)

I would start with developing processes that even though you send them out for calibration, that the devices still produce reliable, consistent readings with the different environments/ operators that are present.

Thanks, Brad. Our processes include checks to ensure everything is in control and if out of control although they would not specifically point to pipettors, they would indicate a problem that would have to be corrected.

Our issue was.... well it really doesn't matter because you have cleared things up nicely. We currently examine returned items fairly thoroughly to ensure they have been calibrated correctly but our written procedure was a little weak. We consider this correction to be an opportunity for improvement.

Thanks again all.

Tom from Clayton