By sdc on Wednesday, July 21, 2004 - 11:25 am:

I work in the pharm business. My new boss is a recent PhD grad in mech. engineering. With every set of data I produce he wants to know the accuracy of the result (and the tighter the better). I have "qualified" the HPLC methods I am currently running to a +/- 2% accuracy level, but he wants to see better. Specifically I am looking for HPLC courses that use a combination of statistics and method controls which will allow me to determine actual accuracy levels of a single result. For example, if he really wants an accuracy level of 0.5%, what do I have to do in the method, and what type of statistical tests do I use to prove/disprove that level?

Thanks.
s

-------------------------------------------------------------------------------------------------------
By dlorenzo on Thursday, July 22, 2004 - 09:12 am:

Was your boss born on earth? Did he studied statistics?
100+ - 1%!!! Please! :-)

-------------------------------------------------------------------------------------------------------
By scarredand wary on Thursday, July 22, 2004 - 09:36 am:

Not withstanding the knowledge of a mechanical engineer, what does He know about the errors inherent in the science of chromatography?

Does He know that glassware measurement of liquids has an error of 1% or WORSE?

Does He know the injection variability of a gas sample loop is about the same?

Does He know the reproducibility of integrators integrating the same generated analog peak signal?

Why do companies put people in charge of areas in which they have no technical competence?

I wish you well.

-------------------------------------------------------------------------------------------------------
By judd on Thursday, July 22, 2004 - 11:11 am:

I think what you need to do is walk the poor guy through the errors inherent in each step of the process from weighing samples and standards, dilution, injection, and so forth.

It certainly will do you no harm to identify and quantify these things if you have not done so already.

For him to arbitrarily impose accuracy and precision standards is silly, but you should have at least a reasonable idea of what kind of accuracy and precision is considered acceptable for your assays and where your sources of error lie.

And why would a mechanical engineer be running a chromatography lab anyway?

-------------------------------------------------------------------------------------------------------
By Michele on Thursday, July 22, 2004 - 12:37 pm:

Give your boss a copy of the industry guidance from the FDA, and point out that you are well within the guidance. And I don't think that you need a course in HPLC and how to handle the data; I think that he does.