Laserfiche WebLink
<br />ACZ Laboratories, Inc. August 10, 2007 <br />Quality Assurance Plan Version 12 <br />SOPAD018.08.07.12 Page 48 of 96 <br />12 EVALUATING QUALITY CONTROL SAMPLES <br />In general, acceptance criteria for quality control samples are method-specific; however, requirements of <br />clients and regulatory or other accrediting agencies must also be included. Immediate corrective action must <br />be taken if any quality control is outside of the acceptance criteria. Appropriate corrective actions are <br />described in the test SOP. To the extent possible, client samples are reported only if all quality control <br />measurements are acceptable. If a quality control measure is outside of acceptance criteria, and the data must <br />be reported, then all samples associated with the failed QC must be reported to the client with the appropriate <br />data qualifier(s). Clients will occasionally request limits different from those in a published method. If a client <br />has data quality objectives that require modification of our guidelines then we may deviate from those <br />guidelines only if more stringent controls are requested. ACZ's policy is to adhere to the strictest limits as a <br />means of meeting all agency and client requirements. <br />For methods that do not specify acceptance criteria for any type of quality control measurement, limits may be <br />generated by plotting historical data in a control chart once a minimum of 20 data points is available. A control <br />chart application may be accessed through LIMS and allows the user to create limits, either from a specified <br />number of data points or for a specific time period, that are set at + 3 times the standard deviation from the <br />mean percent recovery. Current control limits are also plotted to provide a direct comparison of the two sets of <br />data. New limits developed from a control chart must be documented on FRMQA039 and must be reviewed by <br />the QA/QC department prior to implementation. If the new limits are approved, then QA/QC personnel will <br />update LIMS. Refer to ACZ's SOP' Control Charting Application and Procedure (SOPAD041) for further <br />details. Default acceptance criteria established by the Arizona Department of Health Services (ADHS) may be <br />used in lieu of generating a control chart to establish limits; however the SOP must specify which limits are in <br />use. 4 NOTE: For all data evaluation, final results-ending with 1 - 4 are rounded down and results ending <br />with 5 - 9 are rounded up. <br />12.1 Accuracy <br />Accuracy is defined as "the degree of agreement of a measured value with the true or expected value of the <br />quantity of concern." Control samples (LCS or LFB) and spiked samples are analyzed with every batch of <br />samples or as stipulated by the specific test SOP to assess accuracy and matrix effects. <br />Percent Recovery (%R) for a control sample is calculated as follows: <br />%R M, x 100 Where: M = Measured concentration of the control sample <br />SP Sp= True value of the control sample <br />Percent Recovery (%R) for a spike is calculated as follows: <br />%R= M - S x100 Where: M = Measured concentration of the spiked sample <br />Sp S = Measured concentration of the sample aliquot <br />Sp= True value of the spike concentration <br />4 ADHS Information Update #87 (issued July 7, 2005) <br />* "Quality Assurance of Chemical Measurements," Taylor, J., 1987 <br /> <br /> <br /> <br /> <br />t rc <br /> <br /> <br />2773 Downhill Drive 970-879-6590 ' <br />Steamboat Springs, CO 80487 www.acz.com <br />