RE: Applicability of SPC to Intra and Inter Assay controls

["Prevette, Steven S" <Steven_S_Prevette@rl.gov>]

>The best example is the inter assay control.  First, a standard curve is
generated using "known" sample concentrations (3-5 tested from each sample).
Then a given sample (control) is tested 3-5 times giving a range for that
sample.  The next time that an assay is run, a sample (control) from the
same lot is tested along with any other samples to be tested in the same
manner.

>It is my contention that the variablity and repeatability of the assay
should be determined via SPC on the control sample.

>Does this application truely fall under SPC or traditional statistical
methods?

I assume that "traditional" methods are currently used (most likely ANOVA or
t-test or chi-square).  This, I am guessing, is due to each test is being
treated as an individual test, and only being analyzed within that single
result.  Descriptive statistics are being used.  But you have seen that
perhaps there should be an analysis (inferential) done across the results.
This would then become a time-series analysis, very applicable for SPC.

Steve Prevette
Occupational Safety and Health
Fluor Hanford, A Fluor Global Services Company
ASQ Certified Quality Engineer
steven_s_prevette@rl.gov
509-373-9371