Page 1 of 1

Pooled Precision by looking at peak areas

Posted: Sat Mar 26, 2016 12:59 pm
by Rhodium
We have a pooled precision requirement that must be met in order to transfer an HPLC test method from lab to lab. The pooled precision uses peak areas to make this precision determination. Should there be a concern that since the two labs are using different makes and models of detector (Waters uv/vis and Agilent DAD), this could lead to different peak areas/sensitivities? From a detector v/s detector standpoint, what all should be taken into consideration if the peak areas are shown to be different - thus causing a pooled precision failure? Some immediate concerns:
- photomultiplier tube in the Waters uv/vis detector can result in higher sensitivity versus the Agilent DAD
-different flow cells (path length and volume)
-different lamp life
-optical differences

Also, is looking at peak areas to determine pooled precision across labs a common practice?

Thanks in advance for the input.

Re: Pooled Precision by looking at peak areas

Posted: Sat Mar 26, 2016 3:22 pm
by Peter Apps
If "pooled precision" takes the results from different labs and calculates the mean and standard deviation for all of them then you cannot do it using peak areas - for the reasons that you mention. You have to calculate the measurement results (concentration, mass or whatever it is you want to measure) for each determination and pool those across the labs.

Peter