Chromatography Forum

I have a few questions concerning the method validation in HPLC and possible replies to external evaluation questions.

I guess that from the start we can not question the reactives provided by the companies who sale different kits for HPLC.
Considering this, whenever we run a batch we pass the standard sample at the begining and one more time after the last sample at the end of the batch. At the end we pass it as a sample, just to see if we get the same value as the one given by the company.
Sometimes we get a different value, and my question is what is the maximum allowed difference that we can accept?

The second question, if the preparation of the standard sample involves additional steps of manipulation how can we be sure that the prepared standard has the right concentration and how can we prove this?
How can you be sure that the prepared standard is realy good?

Thanks in advance to everybody

When the method is validated for quentification, there should have been measurment accuracy ranges established for standard solution (C ± x%).

So if you have a STD solution of concentration C and you analyse it as thought it was a sample it should return a sample within the established range if the assay system is working properly.

There are an number of vaiables at work in HPLC and injection of the same solution will rarely provide Identical results. there is always some variation.

I guess that from the start we cannot question the reactives provided by the companies who sale different kits for HPLC.

Why not? Numbers they give you may be rigth but it is for their HPLC, in their lab, using their reagents, run by their analysts etc. It may (should!) work pretty much the same in your lab but you can't be sure, can you?

Considering this, whenever we run a batch we pass the standard sample at the begining and one more time after the last sample at the end of the batch. At the end we pass it as a sample, just to see if we get the same value as the one given by the company.
Sometimes we get a different value, and my question is what is the maximum allowed difference that we can accept?

If the kit you have is validated by the company who sells it, it should have this information somewhere (repeatability, intra-assay precision). But again - it's their number, not yours. If they don't have it then obviously you need to validate this method for yourself.

The second question, if the preparation of the standard sample involves additional steps of manipulation how can we be sure that the prepared standard has the right concentration and how can we prove this?
How can you be sure that the prepared standard is realy good?

What kind of manipulations?

Thank you for your answers.

Kristof, by manipulation, to answer your question, I mean intermediate steps before injection (like derivatization of the standard in case of amminoacids) step that involves pipetting solutions. Can we just consider in this case that the technician works perfectly all the time or we should check pipetting errors to make sure that everything is OK?

Thank you

Do you have an SOP that covers pipetting?

Wouldn't meeting your precision and accuracy (and intermediate precision) requirements take any such "errors/variability" into account?

If you work in GLP/GMP environmet your pipettes should be calibrated and you should know their precision. In any case operator-to-operator variation should be included in intermediate precision assays i.e. they should be performed by at least two operators. (Same goes for equimpment, columns, mobile phase batches etc. - at least two of each. Some regulations - Japanese if I recall - actually tell you how it is supposed to be arranged.)

If you work in GLP/GMP environmet your pipettes should be calibrated and you should know their precision.

Kristof - do you mean Eppendorf-type pipets or also glass Class A volumetric pipets? If the latter, do you also calibrate your volumetric flasks?

Actually both are calibrated and have defined precision, but only automatic pipettes can be calibrated (to certain extent) in-house. No, we don't calibrate our glassware.