I gave a talk recently in Bethesda, MD on glucose meters. On one of my slides, I showed large differences among glucose meter values for a proficiency survey sample but qualified things by saying that these differences might be due to matrix effects. After the talk, a commentator said, yes there are matrix effects which could likely explain these results. Now this person until recently developed glucose meter assays, so I defer to his expertise.
So one thought is how useful are proficiency survey samples if much of the variation is not relevant to the assay.
Another thought is how much variation is explained by matrix effects. Is it possible for matrix effects to appear across reagents lots for the same glucose meter?
Does anyone have an opinion about this?