Standard Practice for Quality Control of Routine Testing in a Laboratory

Importancia y uso:

4.1 Test Method Monitoring Program—A testing laboratory conducting a test method on a routine basis should ensure the quality of its test results through a test method monitoring program (TMMP). A TMMP provides repeated test results on one or more control sample materials to validate the results for samples (for example, batch release or operations monitoring) and to assess the performance of a test method in the laboratory in terms of test method precision and bias. The program also gives guidance to detect, identify, and correct changes in the testing process. This practice discusses design, implementation, and operation of the TMMP, and statistical analysis of the resulting data. The program is a form of statistical process control (SPC) on the testing process. For general information on SPC and control charts see Practice E2587 and Manual 7 on Presentation of Data and Control Chart Analysis.4 Practice D6299 also describes SPC applications to the testing process.

4.2 Results for a control sample in a run of the test method are often used as a criterion for validity of sample test results for that run. When used for this purpose, the control sample result must be within statistical control limits for the sample test results to be accepted. If outside limits, the test results will be evaluated and may be replaced.

4.3 The TMMP uses control charts to monitor test method behavior and signal drift and other deviations from a stable state of the testing process. It includes an out-of-control action plan (OCAP; see Section 8) to identify and address causes of any testing process departure from the stable state.

4.4 Data from the TMMP program are used to assess the performance of a test method in a laboratory in terms of test method precision and bias (see Practice E177). The laboratory precision estimate generated from test results acquired over an extended period of time, possibly by multiple operators and including other sources of variation, may be useful to laboratory customers that need estimates of measurement uncertainty. This estimate may also be useful in product stability and environmental monitoring, or process development experiments.

4.5 The TMMP aids on-going qualification of laboratory personnel. Personnel demonstrate their proficiency for the test method when they produce consistent test results for control samples. They should also be able to detect, identify, and implement corrections when the testing process drifts out of statistical control.

4.6 Implementation of a TMMP has three stages following Practice E2587:

4.6.1 Stage A: Planning and Process Evaluation—Control sample materials are selected and prepared. Initial control data are collected. Data are tested for trends, autocorrelation, and normal distribution. Initial control limits are calculated.

4.6.2 Stage B: Process Improvement—Special causes of variation are detected and corrected. After this initial period, control charts and limits may be updated.

4.6.3 Stage C: Process Monitoring—The control charts are used to monitor test method behavior and signal deviations from the stable state of the testing process. Additional control chart evaluation rules and additional charts may be applied to detect smaller shifts, drift, and other changes. An OCAP addresses causes for testing process departure from stable state. Test precision estimates and control limits are updated periodically or following significant process changes.

Subcomité:

E11.20

Volúmen:

14.01

Palabras clave:

control chart; laboratory management; measurement; quality control;