Method Validation Work Instruction

Thread Content

1. Purpose: To verify the various methods selected for use in the laboratory, in order to confirm that the laboratory is capable of applying these methods correctly, and to ensure that they are suitable for their intended purposes and can be used within acceptable error ranges within this laboratory. 2. Scope: Applies to standard methods adopted by laboratories, as well as non-standard methods, methods developed by the laboratory itself, standard methods used outside their intended scope, and standardized methods that have been expanded or modified ； It is also applicable to developing new methods for the research of new methods/technologies. 3. Responsibilities 3.1 The technical supervisor designates a specific person to be responsible for method validation, and reviews and approves the results of such validation ； 3.2 Engineers involved in the method validation should meticulously record the test phenomena and data, summarize the experimental results, and document them ； 3.3 Relevant personnel must strictly follow this operation manual. 4. Glossary 4.1 Detection limit: The method detection limit refers to the lowest concentration (relative detection limit) or lowest mass (absolute detection limit) of the substance being measured that can be detected by a particular method at a given confidence level. There are various methods for determining the method detection limit, and the appropriate method should be chosen based on the specific testing method used. 4.2 Linearity Linearity refers to the ability of an analytical method to yield measurement values that are in a linear relationship with the amount or concentration of the substance being measured. 4.3 Precision Precision refers to the degree of consistency among a series of measurement values obtained when multiple samples to be tested are analyzed repeatedly from a homogeneous sample. The error of the measured value is expressed in the form of deviation, standard deviation, or relative standard deviation. 4.4 Accuracy Accuracy refers to the degree to which the value obtained by an analytical method corresponds to the true value (standard value). It is usually expressed as the ratio of the difference between the mean value obtained from repeated measurements and the true value to the true value itself. The true value is generally the theoretical value; if such a value does not exist or is difficult to determine, a validated or accepted value can be used in its place. 4.5 Uncertainty of analysis results: Measurement uncertainty: A parameter related to the measurement result, which characterizes the dispersion of values that can be reasonably assigned to the quantity being measured. In practical work, the uncertainty of results can arise from many sources, such as incomplete definitions, sampling, matrix effects and interferences, environmental conditions, uncertainties in mass and containers, reference values, estimates and assumptions in measurement methods and procedures, as well as random variations. When validating a method, efforts should be made to identify all components affecting uncertainty, provide a comprehensive assessment, and ensure that the presentation of the results does not create any misconceptions. A sound assessment should be based on knowledge of method implementation and the measurement scope, as well as past experience. There should be no omissions nor repetitions of the uncertainty components arising from each influencing quantity. When evaluating measurement uncertainty, all significant uncertainty components under the given conditions shall be considered using appropriate analytical methods. 4.6 Sensitivity of the method: Sensitivity can be described as the ratio of the instrument’s response value or some other indicator to the concentration or amount of the substance being measured. The sensitivity of a method can vary depending on changes in the experimental conditions. Under certain experimental conditions, sensitivity exhibits relative stability. 5. Confirmation of contents 5.1 Linear evaluation 5.1.1 Prepare at least 5 standard solutions at different levels to create a standard curve, and evaluate its regression equation and coefficient of linear correlation. 5.1.2 The concentration at the peak of the standard curve shall not be more than 50 times the concentration at the trough. 5.1.3 The lowest point of the curve should preferably be close to the reported detection limit. 5.1.4 The correlation coefficient for organic tests such as GC should be ≥0.99, while the correlation coefficient for inorganic tests such as ICP should be ≥0.995. 5.2 Confirmation of detection limit 5.2.1 Method detection limit (MDL): Repeat the measurement of the standard substance in the reagent blank at least 7 times, calculate the standard deviation s, and multiplying this value by 3 yields an estimate of the MDL. Select samples with uniform texture and ensure that they show no readings; add standards with concentrations similar to those of the MDL estimate, and repeat the experiment at least 7 times, with the average of these 7 tests to be obtained over three days. Calculate the standard deviation of the spiking test: MDL = t(n-1) × s, where the t-value is the coefficient from the t-distribution with degrees of freedom of n-1 at a 99% confidence level, which can be found in tables. Values for n: 73.14, 83.00, 92.90, 102.82. Note: The calculated MDL value should not exceed 10 times the spiking amount; otherwise, a lower concentration should be chosen and the above steps repeated. Calculate the recovery rate and RSD value of the spiked sample; the recovery rate should be within the range of 70–130%, with an RSD ≤ 20%. Recovery rate = Actual value of the standard / Theoretical value after spiking × 100% 5.3 Verification of accuracy and precision 5.3.1 Verification of accuracy is carried out using the spiked recovery method, in which a certain amount of standard substance is added to the sample to determine its recovery rate. At least 7 spiked samples should be tested, and these 7 spiked samples should be tested on average over three days. Record the test results and calculate the recovery rate, which should be within the range of 80–120%, with an RSD ≤ 20%. The recovery rate range for elemental analysis using ICP and similar methods should be controlled between 90 and 110%. 5.3.2 The accuracy verification tests should cover, as much as possible, the different material samples mentioned in the method; the spiking levels used should encompass the entire range of values that can be tested by the method, including concentration levels slightly above the MDL. The tests at the highest points of linearity should be repeated at least seven times, with an RSD of ≤20%. 5.4 Assessment of the uncertainty associated with the tests 5.4.1 The quantities to be measured are specified in the corresponding SOPs. 5.4.2 Identifying sources of uncertainty: Consider the uncertainties related to the factors affecting the overall performance of the method, such as observable precision and the standard deviation measured relative to appropriate reference materials; these constitute the main components in the assessment of uncertainty. Then, the other possible components are evaluated, and those that are significant are quantified. 5.4.3 Quantification of uncertainty components: Measurement uncertainty is classified into Category A assessment and Category B assessment: Statistical analysis of the observed data is conducted based on the actual measurement results; the standard deviation of repeated measurements is calculated to determine the Category A uncertainty ； The Type B uncertainty shall be evaluated using a method different from the Type A evaluation, based on relevant technical data, characteristics of measuring instruments, or information such as calibration certificates ； 5.4.4 Calculation of the combined uncertainty: The components of the uncertainties are combined in accordance with relevant rules to obtain the combined uncertainty; by multiplying this combined uncertainty by the coverage factor, the expanded uncertainty can be obtained. 5.5 Data Recording: During the experiment, the tester should record in detail the entire process of the test, including the method of sample preparation, the grade of reference standards and the preparation of reagents, the experimental environment, analysis and testing parameters, precautions, and analysis results. 5.6 Preparation of the method validation report: After the tests are completed, engineers should prepare a method validation report. This report should include at least the following information: a) Scope; b) Description of the types of items to be tested; c) Parameters and values that need to be determined; d) Devices, equipment, and reference materials required; e) Required environmental conditions; f) Records of method validation parameters, as well as the criteria and/or requirements for determining compliance. g) Data to be recorded and methods for analysis and presentation h) Uncertainty or procedures for assessing uncertainty and the results thereof i) Instructions for the procedure, including: identification markers of the samples to be tested, their handling, transportation, storage, and preparation; inspections prior to testing; checks to ensure that the equipment is functioning properly, with calibration or adjustment carried out before each use if necessary; methods for recording data and results; safety measures to be followed, etc. 5.7 Review of the method validation report The method validation report is submitted to the technical supervisor or other relevant personnel for review, and only after approval can the testing method be implemented in the testing laboratory.