Chromatography Forum

Is system suitability for UV Spectrometric method necessary? Is there any guidance that I can read?
All UV Spectrometric methods in my current company only scan 1 time on standard 1 time on sample.

Thanks.

That is really more of a regulatory than a scientific questions, and the answer is "That depends". If your facility operates under cGMP or GLP then I would say system suitability should be part of of any method (otherwise, how can you be sure that your system is, in fact, suitable for the analysis?). As a minimum I would look to verify the wavelength accuracy and check repeatability.

Possible exceptions that I can imagine might be compendial methods (if they do not include system suit), very old validated methods which might be "grandfathered in", or if you have an SOP in place which allows you to skip system suit (and that SOP has been reviewed and found acceptable).

I'll throw in a disclaimer that I am not a regulatory expert (nor a spectroscopist, for that matter), so I may be way off base on this one.

Many of the USP raw material methods just state to run a blank, the standard, then the sample. If you are doing an assay of this nature, then perhaps that is all you need to do.

Many of the derivatization assays include the need to run multiple standards to create a curve to plot the sample against. This would be considered a system suitability.

I have never seen a UV method that required you to run a standard five times and calculate the %RSD though, so I wouldn't do system suitability in that regard for each assay.

Hi

Yes it kind of depends. For like limit test in pharmacopiea monographs I would say no, ie run blank and then your test solutions.

For like assays with a calibration curve (could be a dissolution test or something) I would check the correlations coefficient regardless if it would be formally required in a file or just as an in-house check.

Seen but not used myself a few cases where recovery is checked against calibration curve, or that there is a permitted intervall for the "e" coeffient (in lambert beers law)