MS-QA_06-2019 rev5

ANSI/AARST MS-QA 2019 26 Radon Measurement Systems Quality Assurance Laboratories—All Methods 1. results from the most recent QC interval , and 2. the initial post- calibration QC results. Laboratories are permitted to use a rolling assessment “look-back” period after the initial in-control limits are set or other systems designed to ensure that their operations are in control and that analysis results are accurate. Such QC records and analyses are to be available to auditors and credentialing authorities, subject to confidentiality agreements. 7.7.3 Minimum Intervals between Calibrations As batches of sensitive material, equipment and all the details involved in device analyses change over time, warning and control limits are to be recalculated based on changes in QC results. The same calibration functions can be used by CAD and ATD analysis laboratories for up to 5 years if: 1. No new materials are used in the device or new instrumentation introduced, including counters, shielding, electronics, charcoal, plastic, containers, etc.; and 2. There is documentation, including QC data, demonstrating that there has been no statistically significant (5%) change in the population (mean and variability) for each type of QC check. This is done by comparing the current QC results to the most recent QC interval and the initial post- calibration population of QC checks. 7.7.4 Use of Calibration Factors Beyond Any 12-Month Period Tests for both changes in the mean and the variability shall be conducted and documented to validate the continued use of the calibration factors beyond 12 months. A two-tailed F-test (Snedecor and Cochran, 1983), as well as other methods, can be used to test if the variances of two populations are equal, and these results will indicate whether the equal-variability t- test is appropriate. An unpaired two-sample t-test (Snedecor and Cochran, 1989), among others, can be used to determine if two population means are equal. If these analyses demonstrate that there is less than a 5% probability that the most recent set of QC checks come from a different population of QC checks (based on a comparison of both mean and variability) than either the original post- calibration set or the set from the most recent QC interval , the calibration period is permitted to extended beyond 12 months. The results of such analyses are to be available to auditors for review and determination of acceptability. 7.8 Laboratory Spikes Note—Lab spikes assess sources of error in the entire laboratory measurement system relative to a recognized reference radon concentration. These laboratory QC checks are independent of calibration operations and are designed to measure and track variability within the entire laboratory’s quality system . To do this, spikes are handled and analyzed just as routine detectors. These QC checks are in addition to laboratory control samples that track variability of equipment performance or internal processes. 7.8.1 Analysis Laboratory—Spike Frequency Laboratory spikes are to be conducted with at least three laboratory spikes per every 100 devices processed, or a maximum necessary of six per month. If a laboratory analyzes different configurations, the same requirements apply for each different configuration, which includes both the detector design as well as the type and source of the charcoal or alpha track-detection plastic. The definition of significant type difference can be used to determine the need for sample spikes for devices with different characteristics.