4.1 Laboratories conducting routine testing have a continuing need to make improvements in their testing processes. In these situations it must be demonstrated that any changes will neither cause an undesirable shift in the test results from the current testing process nor substantially affect a performance characteristic of the test method. This standard provides guidance on experiments and statistical methods needed to demonstrate that the test results from a modified testing process are equivalent to those from the current testing process, where equivalence is defined as agreement within a prescribed limit, termed an equivalence limit.

4.1.1 The equivalence limit, which represents a worst-case difference or ratio, is determined prior to the equivalence test and its value is usually set by consensus among subject-matter experts.

4.1.2 Examples of modifications to the testing process include, but are not limited, to the following:

(1) Changes to operating levels in the steps of the test method procedure,

(2) Installation of new instruments, apparatus, or sources of reagents and test materials,

(3) Evaluation of new personnel performing the testing, and

(4) Transfer of testing to a new location.

4.1.3 Examples of performance characteristics directly applicable to the test method include bias, precision, sensitivity, specificity, linearity, and range. Additional characteristics are test cost and elapsed time needed to conduct the test procedure.

4.2 Equivalence testing is performed by a designed experiment that generates test results from the modified and current testing procedures on the same types of materials that are routinely tested. The design of the experiment depends on the type of equivalence needed as discussed below. Experiment design and execution for various objectives is discussed in Section 5.

4.2.1 Means equivalence is concerned with a potential shift in the mean test result in either direction due to a modification in the testing process. Test results are generated under repeatability conditions by the modified and current testing processes on the same material, and the difference in their mean test results is evaluated.

4.2.1.1 In situations where testing cannot be conducted under repeatability conditions, such as using in-line instrumentation, test results may be generated in pairs of test results from the modified and current testing processes, and the mean differences among paired test results are evaluated.

4.2.2 Slope equivalence evaluates the slope of the linear statistical relationship between the test results from the two testing procedures. If the slope is equivalent to the value one (1), then the two testing processes meet slope equivalence.

4.2.3 Range equivalence evaluates the differences in means over a selected wider range of test results and the experiment uses materials that cover that range. The combination of slope equivalence and means equivalence defines range equivalence.

4.2.4 Non-inferiority is concerned with a difference only in the direction of an inferior outcome in a performance characteristic of the modified testing procedure versus the current testing procedure. Non-inferiority may involve the comparisons of means, standard deviations, or other statistical parameters.

4.2.4.1 Non-inferiority testing may involve trade-offs in performance characteristics between the modified and current procedures. For example, the modified process may be slightly inferior to the established process with respect to assay sensitivity or precision but may have off-setting advantages such as faster delivery of test results or lower testing costs.

4.3 Risk Management-Guidance is provided for determining the amount of data required to control the risks of making the wrong decision in accepting or rejecting equivalence (see 5.4 and Section A1.2).

4.3.1 The consumer's risk is the risk of falsely declaring equivalence. The probability associated with this risk is directly controlled to a low level so that accepting equivalence gives a high degree of assurance that the true difference is less than the equivalence limit.

4.3.2 The producer's risk is the risk of falsely rejecting equivalence. The probability associated with this risk is controlled by the amount of data generated by the experiment. If valid improvements are rejected by equivalence testing, this can lead to opportunity losses to the company and its laboratories (the producers) or cause unnecessary additional effort in improving the testing process.

This practice provides statistical methodology for conducting equivalence testing on numerical data from two sources to determine if their true means or variances differ by no more than predetermined limits. This standard provides guidance on experiments and statistical methods needed to demonstrate that the test results from a modified testing process are equivalent to those from the current testing process, where equivalence is defined as agreement within a prescribed limit, termed an equivalence limit.