Page 1 of 1

setting specification for inter laboratory precision

Posted: Fri Feb 08, 2008 11:45 am
by michael101
Hello
I have some questions regarding the setting of specifications between two sites for the transfer of a particle size method, specifically inter laboratory precision

Validating a particle size method as per USP allows the following variation for repeatability: (before any specification is set for inter laboratory precision)
Dv 10 <30%
Dv 50 <10%
Dv 90 <15%

My questions are as follows:
Is there any paper or document on the setting of such specifications between sites.
How do I take external factors(two different instruments, analysts, etc) into account when setting these specifications.
Should the results be reported within a range (+/- 5um etc) or as a %RSD.
Any comment would be greatly appreciated
Thank you

Posted: Mon Feb 11, 2008 10:22 am
by RichardS
There are a few particle size method validation papers about, PASG published a good starter a few years ago, but none cover method transfer specifically.
The USP / EP RSD values are so large that they can comfortably be used for method transfer. Most methods can be set with much tighter specifications and no problems.

Normally I'd suggest
1) Check RSDs on the same sample as it recirculates. Should be very low. This is checking the repeatability, unless we have no dissolution or aggregation we cant validate the method
2) Size different preparations of the same sample, reproducibility, here we're checking on sampling.
3) 2 analysts on the same instrument to as part of an intermediate precision test. This is really checking the sample preparation is free of operator bias

If all that is OK we can start to look at method transfer. Laboratory temperature and humidity between sites can play a part here. It helps if you're using the same model of instrument and sample tank units as size range and tank volume will vary between models.