Noise in LOQ

[mojo]

Which noise result is used when calculating the LOQ? We have a noise report from ChemStation which lists 6xSD, ASTM and PtoP.

[Jumpshooter]

You have those three options to choose from--the three you listed in your post--it's your choise.

[Dan]

Mojo,

The subject of signal-to-noise has been discussed before. Here are just a couple of links:

http://www.sepsci.com/chromforum/viewtopic.php?t=6442

http://www.sepsci.com/chromforum/viewtopic.php?t=6245

If you seacrh the forum, you may find more.

I don't think that there is any "standard" option. The peak-to-peak measurement technique may be the most popular. A chromatography data system usually uses RMS noise, but that has been changing and more options are available now.

I agree with Jumpshooter, you must choose the option that works for you. You should consider documenting the noise technique that you use in an SOP. At the very least, be consistent.

Regards,
Dan

[shaun78]

When I can get away with a calculated LOQ by QA here, I will generally take it from linearity because it is generally always required in the work I do.

Calculating LOQ from linearity you get (10*sigma)/S where,

sigma = residual standard deviation of the regression line;
S = slope of the regression line.

Hope this helps.

Shaun

[Shanextan]

The important thing is you need to test the LOQ to see if it meet your criteria. The QA will see your tested LOQ, not the calculated LOQ. --Shane