MS-QA_06-2019 rev5

ANSI/AARST MS-QA 2019 25 Radon Measurement Systems Quality Assurance Laboratories—All Methods 7.6.1 MinimumWarning and Control Limits The limits presented in this standard of practice are the basic minimums, and it is expected that labs will develop their own limits based on at least 20 different types of QC checks and then develop, document and periodically evaluate their own limits specific to their operations using the methods referenced in Section 3.5 or other statistically-defensible control limit generation algorithms. 7.7 Establishment of In-Control Operations Informative advisory —Because recalibration can entail a large set of chamber exposures (at least five of each unique parameter combination or as consistent with the requirements in Section 7.5 ), it is important to keep all factors as consistent as possible to allow stable, in-control operations without having to generate new calibration factors. 7.7.1 Adequate Statistics Required This standard does not dictate specifics for the triggers that mandate recalibration but does supply the following principles that shall be incorporated into the laboratory’s QAP: a) At least approximately 20 valid QC checks of each type used by the laboratory to monitor stability (at minimum, 20 lab spikes , 20 lab blanks and 20 laboratory control samples , as relevant) are necessary to establish adequate statistics to determine in-control operations. (Informative advisory —An optional type of QC check, lab duplicates or recounts, is recommended for at least periodic assessments of imprecision components due to laboratory activities.); and b) Determination of stability. The determination of calibration function stability shall be based on at least the variability and mean for: 1. background as determined using matrix blanks , 2. the RPE for laboratory spikes , and 3. laboratory control samples (such as matrix spikes ) during every analysis session. 7.7.2 QC Interval and Continued Assessment of In-Control Operations The QC interval is defined as the shortest interval of time needed to gather at least 20 of each type of QC check necessary for the laboratory’s quality system . After calibrations in-control operations are established, an initial pool of QC results is to be collected and analyzed, which will produce control charts with QC limits. These QC limits are used as a baseline against which future QC results will be compared and demonstrate the validity of measurements made during the time periods when QC results are within limits. The data needed for continued assessment and documentation of stable calibration functions shall include: a) At least approximately 20 QC checks of each type used by the laboratory to monitor system performance that are charted, evaluated statistically and used to characterize the in-control distribution (variability and mean) of each QC check type. b) At least approximately 20 QC checks of each type from subsequent QC intervals , which are charted, evaluated statistically and compared with the mean and variability of the initial post- calibration , in- control set of QC results. c) An ongoing analysis of QC results demonstrating that there are no statistically significant (probability >5%) changes in the mean or the variability of the current set of QC check results when compared to both the: