For long term performance analysis and control a well characterized sample with known values can be included in every method run. Do you think that performance trend data of this sample could also be used to abbreviate validation exercises, e.g. instead of running an intermediate precision experiment with 6 determinations use the data from the performance trending?