Lot acceptance procedures are used in industry for inspecting quality characteristics of raw materials, in-process product, and finished product. These procedures, together with process controls, comprise a quality control program. For additional information on process control see Practice E 2281 dealing with process capability evaluation and Practice E 2587 dealing with the use of control charts in statistical process control.
Lot inspection procedures classify quality characteristics as either attributes (measured on discrete scales such as percent defective) or variables (measured on continuous scales such as length, weight, or concentration).
Operating characteristic curves, which plot the relationship of the lot acceptance probability versus the true lot percent defective, are used to evaluate the discriminatory power of a given lot inspection procedure, or acceptance sampling plan, and are discussed in Practice E 2234.
This practice considers inspection procedures that may involve multiple-stage sampling, where at each stage one can decide to accept the lot or to continue sampling, and the decision to reject the lot is deferred until the last stage.
At each stage there are one or more acceptance criteria on the test results; for example, limits on each individual test result, or limits on statistics based on the sample of test results, such as the average, standard deviation, or coefficient of variation (relative standard deviation).
The methodology in this practice defines an acceptance region for a set of test results from the lot such that, at a prescribed confidence level, the probability that a sample from the lot will pass the original lot acceptance procedure is greater than or equal to a prespecified lower bound.
Having test results fall in the acceptance region is not equivalent to passing the original lot acceptance procedure, but provides assurance that a sample would pass the lot acceptance procedure with a specified probability.
This information can be used for process demonstration or validation.
This information can be used for lot release (acceptance), but the lower bound may be conservative in some cases.
If the results are to be applied to test results from future lots from the same process, then it is assumed that the process is in a state of statistical control (see 4.1). If this is not the case then there can be no guarantee that the probability estimates would be valid predictions of future process performance.
This methodology was originally developed by J. S. Bergum (1-4) for use in two specific quality characteristics of drug products in the pharmaceutical industry: content uniformity and dissolution, as respectively defined in chapters x003C;905> and x003C;711> of the United States Pharmacopeia (5) .
Mathematical derivations would be required that are specific to the individual criteria of each test.
1.1 This practice provides a general methodology for evaluating single-stage or multiple-stage lot acceptance procedures which involve a quality characteristic measured on a numerical scale. This methodology computes, at a prescribed confidence level, a lower bound on the probability of passing a lot acceptance procedure, using estimates of the parameters of the distribution of test results from the lot.
1.2 For a prescribed lower probability bound, the methodology can also generate an acceptance limit table, which defines a set of test method outcomes (e.g., sample averages and standard deviations) that would pass the multiple-stage procedure at a prescribed confidence level.
1.3 This approach may be used for demonstrating compliance with in-process, validation, or lot-release specifications.