您需要 登录 才可以下载或查看，没有帐号？立即注册

3.3 METHOD VALIDATION EXPERIMENTS 方法验证试验

In this section we outline the requirements for method validation according to current ICH guidelines.

在本部分，我们列出了现行ICH指南对方法验证的要求。

3.3.1 Specificity 专属性

ICH definition: Specificity is the ability to assess unequivocally the analyte in the presence of components that may be expected to be present.

ICH定义：专属性是方法在样品中常规所有成分存在时对被分析物质进行独立评估的能力。

Most related substance methods will be used in a stability study, and therefore they have to be stability indicating.Stability indicating means that the method has sufficient specificity to resolve all related substances and the drug substance from each other. Typically, for the related substance method for a drug product, degradation products are the most critical related substances. Therefore, as a minimum requirement, the method should have sufficient specificity to resolve the degradation products and the drug substance. In addition, all degradation products should be resolved from potential interference with the excipients.

大部分有关物质方法会被用于稳定性试验，因此应该具有稳定性指示性。稳定性指示性指方法具有足够的专属性来将所有有关物质和原料药彼此分开。典型情况下，对于制剂中有关物质方法，降解产物是最关键的有关物质。因此，作为最低要求，方法应具备足够的专属性将降解产物与原料药分离开。另外，所有降解产物均不应受到辅料的潜在干扰。

Samples for Specificity 专属性样品

? Blank solution to show no interference with any HPLC system artifact peak.

空白溶液（无样品）显示无干扰

? Placebo to demonstrate the lack of interference from excipients.

空白配制（只有辅料，无原料药）显示无辅料干扰

? Drug substance to show that all significant related substances are resolved from the drug substance.

原料药显示所有明显的有关物质均与原料药分离

? Authentic samples of critical related substances to show that all known related substances are resolved from each other.

关键有关物质的认可样品显示所有已知有关物质均彼此分离

? Typically, a stressed sample of about 10 to 20% degradation is used to demonstrate the resolution among degradation products. A 10 to 20% degraded sample is used because it has a sufficiently high concentration level of critical related substance. Therefore, these related substances can be detected easily. In addition, 10 to 20% degradation is not too excessive, and the related substance profile should be close to that of a typical stability sample.

一般来说，降解10-20%的样品用于测试降解产物质的分离度。为什么降解程度在10-20%是因为这种降解样品含有较高浓度的关键有关物质，这样这些有关物质可很容易地被检测到。另外，10-20%不是太过分，这时有关物质谱与典型的稳定性样品相近。

? Stressed placebo to show that the degradation products from the excipients will not interfere with the degradation products of the drug substance.

强降解空白制剂用于显示来自于辅料的降解产物不会对原料药的降解产物形成干扰。

Different Approaches 不同方法

1. When authentic samples of related substance are available. Analyze stressed drug product, placebo, drug substance, stressed placebo, and solutions spiked with authentic samples of related substances. The HPLC chromatograms are used to show the resolution among related substances, drug substance, and other potential interferences. In addition, check the peak homogeneity of the significant degradation products and drug substance by a photodiode array detector (PDA) or mass spectrometer.

可以获得有关物质认可样品时。分析强降解产品、空白制剂、原料药、强降解空白制剂， 以及加入有关物质认可样品的溶液。采用HPLC色谱显示有关物质、原料药和其它可能的干扰间的分离度。另外，采用PDA或质谱检查显著降解产物和原料药的峰纯度。

This verifies that no significant related substance coelute with each other.

这样样能证明较大的有关物质没有在同一时间流出。

2. When authentic samples of impurities are not available. A stressed drug product can be analyzed to show separation of the most significant related substances. In addition, the peak homogeneity of the stressed sample should be investigated by PDA or mass spectrometry. Alternatively, one may use an orthogonal procedure to verify the method specificity. The orthogonal method can be a different technique (e.g., capillary electrophoresis, thin layer chromatography) or different type of HPLC analysis (e.g., reversed phase versus normal phase). For example, compare the related substance profile in the original HPLC method and that of the orthogonal method. To demonstrate method specificity, the significant related substances should be consistent in these methods.

没有杂质认可样品时。对强降解药品进行检测，以显示最显著有关物质之间的分离情况。另外，强降解产品的峰纯度应该用PDA或质谱进行检查。公司也可以使用矩形程序验证方法的专属性。矩形程序可以采用不同技术（例如，毛细管电泳、薄层色谱）或不同的HPLC方法（例如，反相—正相）。例如，将原始HPLC方法中有关物质谱与对照方法中的有关物质谱进行比较。要证明方法的专属性，这些方法中的显著有关物质结果应一致。

3.3.2 Quantitation Limit (and/or Detection Limit) 定量限（和/或检测限）

ICH definition: The quantitation limit of an individual analytical procedure is the lowest amount of analyte in a sample that can be determined quantitatively with suitable precision and accuracy. The detection limit of an individual analytical procedure is the lowest amount of analyte in a sample that can be detected but not necessarily quantitated as an exact value.

ICH定义：一个检验方法的定量限是具有适当的精密度和准确度时可以定量检测的样品中的被分析物质的量。一个分析方法的的检测限是样品中的被分析物质可以被检测到，但不一定可以进行定量为一个确定值的最低量。

Two types of approaches can be used to determine the quantitation limit or detection limit, as described below.

两种方法可以用来决定检测限，说明如下。

Signal-to-Noise Approach. Quantitation limit (QL; Figure 3.9) is defined as the concentration of related substance in the sample that will give a signal-to-noise (S/N) ratio of 10 : 1. Detection limit (DL) corresponds to the concentration that will give a signal-to-noise ratio of 3 : 1. The quantitation limit of a method is affected by both the detector sensitivity and the accuracy of sample preparation at such a low concentration. In practice, the quantitation limit should be lower than the corresponding ICH reporting limit (Table 3.1).

信噪比法。定量限（QL，见图3.9）定义信噪比为10:1时的样品中有关物质的浓度。检测限（DL）对应于信噪比为3:1时的浓度。方法定量限受到检测器灵敏度和样品在这么低的浓度配制准确度的影响。在实践中，定量限应低于对应的ICH报告限（见表3.1）。

To investigate the effect of both factors (i.e., sample preparation and detector sensitivity), solutions of different concentrations near the ICH reporting limits are prepared by spiking known amounts of related substances into excipients.

为了调查两个因素的影响（即样品制备和检测限灵敏度），采用加入已知量有关物质到辅料中的方法，配制接近ICH报告限浓度的不同浓度溶液。

Each solution is prepared according to the procedure and analyzed repeatedly to determine the S/N ratio. The average S/N ratio from all analyses at each concentration level is used to calculate the QL or DL. The following equation can be used to estimate the QL at each concentration level. Since different concentration levels give different QLs, typically the worst-case QL will be reported as the QL of the method.

根据检验方法配制各溶液，重复检测以决定信噪比。使用各浓度水平的所有检测得到的平均信噪比，计算检测限。以下公式可以用于估算在不同浓度水平的检测限。由于不同浓度水平给出不同的检测限，一般会将最差情况时的检测限报告为方法的检测限。

图3.9 定量限

Alternatively, the spike solution can be diluted serially to lower concentrations.

也可以采用加样溶液逐级稀释至较低浓度。

The S/N ratio at each concentration level is determined. The concentration level (in percent related substance) that gives an S/N value of about 10 will be reported as the QL.

检验各浓度水平的信噪比。信噪比约为10时的浓度水平（有关物质百分比）即为定量限。

Standard Deviation Approach. The following equations can be used to determine quantitation limit and detection limit by standard deviation of the response at low concentrations:

标准偏差方法。可以采用以下公式，根据低浓度响应值的标准偏差计算定量限和检测限，

where SD is the standard deviation of the response near QL and S is the slope of the linearity curve near QL.

这里，SD是接近定量限响应的标准误差，S是接近定量限的线性曲线的斜率

There are two ways to determine SD: 有两种方法计算SD

1. Using experiments similar to those given for the signal-to-noise approach, determine the standard deviation of the responses by repeat analysis of a solution near the targeted QL.

采用与信噪比相似的试验，对接受目标定量限溶液进行重复分析，计算标准误差

2. Construct a calibration curve near the targeted QL:

构建接近目标定量限的校正曲线

a. Determine the residual standard deviation of the regression line of calibration, or

计算校正曲线的回归相对标准误差

b. Determine the standard deviation of the y-intercept.

计算Y轴截距的标准误差

Other Considerations for QL. To account for instrument-to-instrument variation, one may need to verify the QL in multiple runs using different instruments.

定量限的其它考虑。考虑到仪器之间的差异，公司可能需要确认采用不同仪器进行多次测试来确认定量限。

The desired QL should be less than the ICH reporting limit (e.g., 50% of ICH reporting limit). QL should be appropriate; too high indicates that the method is not sensitive enough to report results at the ICH reporting limit. Too low indicates that insignificant degradation products, even though much lower than the ICH reporting limit, may need to be reported.

要求定量限应低于ICH报告限（例如，报告限的50%）。定量限应适当，太高说明方法不够灵敏，不能在ICH报告限报告结果；太低说明可能会将不显著的降解产物，甚至比ICH报告限低很多的降解产物均进行了报告。

To ensure that the HPLC system in each analysis is sufficiently sensitive to report results at the ICH reporting limit, one may use a detector sensitivity solution as part of the system suitability test. Since the ICH reporting limit corresponds to QL (i.e., S/N = 10), one-third of the ICH reporting limit should correspond to DL (i.e., S/N = 3). Therefore, as part of the system suitability test, a detector sensitivity solution of a concentration of about one-third of the ICH reporting limit level would be injected. The response of the detector sensitivity solution should meet the detection limit and should be visually distinguishable from baseline.

为保证HPLC系统每次检测时均具有足够的灵敏度，能在ICH报告限来报告结果，公司可以采用检测器灵敏度溶液作为系统适用性测试的一部分。由于ICH报告限对应定量限（即信噪比=10），ICH报告限的1/3对应检测限（即信噪比=3）。因此，作为系统适用性的一部分，进样检测器灵敏度溶液，浓度为ICH报告限水平1/3。检测器灵敏度溶液的响应符合检测限标准，应目视可以从基线上识别出。

Alternatively, one may evaluate the S/N ratio of the standard solution during method development or validation. Part of the routine system suitability test is to determine the S/N of the standard solution before each analysis. Therefore, the S/N of each analysis needs to be greater than the established limit.

公司也可以在方法开发或验证期间采用标准溶液的信噪比来进行评估。作为常规系统适用性测试的一部分，每次检测前测试标准溶液的信噪比。这样，每次分析的信噪比要比所建立的限度更大。

3.3.3 Linearity 线性

ICH definition: The linearity of an analytical procedure is its ability (within a given range) to obtain test results that are directly proportional to the concentration (amount) of analyte in the sample.

ICH定义：分析方法的线性是其所得结果与样品中被测物的浓度（数量）直接成比例的能力（在给定的范围内）。

General Requirements 通用要求

Range. Ideally, linearity should be established from 50% of the ICH reporting limit to the nominal concentration of drug substance in the sample solution (for area percent method). If the linearity does not support such a wide range of concentration, determine the linearity from 50% of the ICH reporting level to 150% of the proposed shelf life specifications of the related substance (for the high–low and external standard methods) as a minimum. This will ensure a linear response for related substances at all concentration levels to be detected during stability.

范围。理想情况下，所建立的线性应覆盖ICH报告限的50%至样品溶液中原料药名义浓度（对于面积百分比方法）。如果线性不支持较宽的浓度范围，则最少在ICH报告限50%至有关物质货架期限度的150%内检测线性（对于高-低浓度和外标法）。这样可以保证有关物质在稳定性期间所有浓度水平响应均为线性。

Experimental Requirements. Solutions of known concentrations are used to determine the linearity. A plot of peak area versus concentration (in percent related substance) is used to demonstrate the linearity. Authentic samples of related substances with known purity are used to prepare these solutions. In most cases, for the linearity of a drug product, spiking the related substance authentic sample into excipients is not necessary, as the matrix effect should be investigated in method accuracy.

试验要求。采用已知浓度的溶液进行线性测定。将峰面积对浓度（有关物质百分比）作图，确认线性。采用已知纯度的有关物质认可样品配制这些线性测试溶液。大多数情况下，试验制剂的线性时，不需要向辅料中加入有关物质，因为在方法准确度中，要调查基底效应。

Acceptance Criteria. Visual inspection is the most sensitive method for detecting nonlinearity. Therefore, the plot has to be linear by visual inspection. In addition, according to ICH guidelines, the following results should be reported: slope, correlation coefficient, y-intercept, and residual sum of squares.

可接受标准。目视检查是检测是否线性最灵敏的方法。因此，目视所绘图应为直线。另外，根据ICH指南，以下结果应该进行报告：斜率，相关系数、Y轴截距、误差项平方和（残差平方和）。

y-Intercept. There are several approaches to evaluating the significance of the y-intercept.

Y轴截距。有几个方法可以用来评估Y轴截距的显著性。

? Intercept/slope ratio. The intercept/slope ratio is used to convert the yintercept from the response unit (peak area) to the unit of percent related substance. The intercept/slope ratio should be compared to the proposed specifications to determine its significance. For example, if the shelf life specification is 2.0%, an intercept/slope ratio of 0.2% may be considered significant, as 0.2% represents 10% relative to the specification.

截距斜率比。截距斜率比用于将Y轴截距上响应单位（峰面积）转化为有关物质的百分比单位。截距斜率比应与所提议的质量标准进行比较，以测试其显著性。例如，如果货架期限度为2.0%，截距斜率比为0.2%，可以认为是显著的，因为0.2%代表限度的10%。

? Statistical approach. The linearity results can be subjected to statistical analysis (e.g., use of statistical analysis in an Excel spreadsheet). The pvalue of the y-intercept can be used to determine if the intercept is statistically significant. In general, when the p-value is less than 0.05, the y-intercept is considered statistically significant. Thep-value, which compares the y-intercept with the variation of responses, indicates the probability that the y-intercept to be not equal to zero. For example, when the p-value is less than 0.05, this indicates that it is 95% confident that the y-intercept is not equal to zero. In other words, it is 95% certain that the y-intercept is significant.

统计学方法。线性结果可以进行统计学分析（例如，采用EXCEL表格中的统计分析）。Y轴截距上的p值可以用于决定截距是否具有统计学显著性。一般来说，如果p值小于0.05，则Y轴截距被认为是统计学显著的。p值，其将Y轴截距与响应变动进行比较，指出的是其Y轴截距不等于0的可能性。例如，如果p值小于0.05，则说明具有95%置信度时Y轴截距不等于0，换句话说，95%肯定Y轴截距是显著的。

Typically, a positive y-intercept indicates the existence of interference with the response or the saturation of responses at high concentrations. A negative y-intercept indicates the possibility of method sensitivity problem (i.e., a low response cannot be detected) or analytes get retained in the glassware or HPLC system (i.e., a compatibility issue between sample solvent and mobile phase).

一般情况下，正的Y轴截距表示高浓度响应的饱和度或响应处有干扰存在。负的Y轴截距说明可能方法灵敏度有问题（即，检测不到低响应值），或被分析物质残留在玻璃容器中或HPLC系统中（即，样品溶剂与流动相有相容性问题）。

Different Approaches for Linearity Determination. The first approach is to weigh different amounts of authentic sample directly to prepare linearity solutions of different concentrations. Since solutions of different concentration are prepared separately from different weights, if the related substances reach their solubility limit, they will not be completely dissolved and will be shown as a nonlinear response in the plot. However, this is not suitable to prepare solutions of very low concentration, as the weighing error will be relatively high at such a low concentration. In general, this approach will be affected significantly by weighing error in the preparation.

确定线性的不同方法。第一个方法是称量不同数量的认可样品，直接配制不同浓度的线性溶液。由于不同浓度的溶液采用不同重量的样品分别配制而成，如果有关物质达到其溶解极限，可能会溶解不完全，则制图时显示出非线性。不管怎样，这种方法不适用于配制非常低浓度的溶液，因为浓度低时，称量误差会相对较高。一般来说，这种方法会在配制过程中，受到称量误差的重大影响。

Another approach is to prepare a stock solution of high concentration, then perform serial dilution from the stock solution to obtain solutions of lower concentrations for linearity determination. This is a more popular approach, as serial dilution can be used to prepare solutions of very low concentrations. Since the low concentrations are prepared by serial dilution, this approach does not need to weigh a very small quantity of related substance. In addition, since all solutions are diluted from the same stock solution, weighing error in preparing the stock solution will not affect the linearity determination.

另一个方法是制备高浓度贮备液，然后进行系列稀释获得低浓度溶液，进行线性测定。这种方法更常用，因为逐级稀释可以用于配制非常低浓度的溶液。由于低学及其溶液是采用逐级稀释获得，这种方法不需要称量微量的有关物质。另外，由于所有溶液均从同一个贮备液稀释而来，在贮备液配制过程中的称量误差不会对线性测定产生影响。

Relative Response Factor. The relative response factor (RRF) can be used to correct for differences in relative response between the related substances and the drug substance. In the area percent and high–low method, the related substances are calculated against the response of the drug substance. In the external standard calculation, the standard curve of drug substance is generally used in the calculation. Since the related substances are calibrated by the response of the drug substance, it is necessary to determine the relative response of the related substance to that of the drug substance. After the linearity of the related substances and the drug substance are determined, one can calculate the relative response factor by comparing the slope of the related substance to that of the drug substance. If the relative response factor is significantly different from unity, a correction factor may need to be used in the calculation. Otherwise, the reported results will be grossly over- or underestimated (Figure 3.10).

相对响应因子。相对响应因子（RRF）可以用于校正有关物质和原料药的响应之间的差异。采用面积百分比和高-低浓度方法时，有关物质基于原料药响应进行计算。采用外标法计算时，一般采用原料药标准曲线进行计算。由于有关物质经过原料药响应校正，因此有必要测试有关物质相对于原料药的相对响应值。在有关物质和原料药的线性测试完成后，公司可以通过比较有关物质和原料药的斜率来计算相对响应因子。如果相对响应因子显示有显著差异，则在计算中可能需要使用校正因子。否则，报告结果可能会大大超出或低于实际值（图3.10）。

图3.10 相对响应因子的影响

3.3.4 Accuracy 准确度

Definition: The accuracy of an analytical procedure expresses the closeness of agreement between the value that is accepted either as a conventional true value or as an accepted reference value and the value found.

定义：一个分析方法的准确度是实测值与接受为真值或可接受参照值之间的接近程度。

General Considerations 一般考虑

Range. Accuracy for the area percent method should be established from 50% of the ICH reporting limit to the nominal concentration of drug substance in the sample solution. For the high–low and external standard methods, determine accuracy from 50% of the ICH reporting level to 150% of the proposed shelf life specification of the related substances. In addition, for the area percent and high–low methods, it is necessary to determine the accuracy of the related substances and the drug substance. For the external standard method, only the accuracy of related substances is required. Since the response of the drug substance in the sample solution is not used in the external standard calculation, it is not necessary to determine accuracy for the drug substance.

范围。面积百分比方法的准确度范围应为ICH报告限的50%至样品溶液中原料药名义浓度。高-低浓度和外标法，其范围为ICH报告限水平的50%至货架期有关物质质量标准的150%。另外，面积百分比和高-低浓度方法，必须要测试有关物质和原料药的准确度。对于外标方法，只要求有关物质的准确度。由于样品溶液中原料药的响应不会用于外标法计算，因此不需要测试原料药的准确度。

Experimental Considerations. Typically, known amounts of related substances and the drug substance in placebo are spiked to prepare an accuracy sample of known concentration of related substance. According to the ICH, accuracy should be determined using a minimum of nine determinations over a minimum of three concentration levels covering the range specified. Similar to the experiments used in linearity (see Section 3.3.3) the related substances and the drug substance can be weighed directly into the placebo or serial dilution from a stock solution can be used. When no authentic sample is available for preparing the spiked solutions, one may determine method accuracy by comparing the results of the original method with that of the orthogonal method (e.g., capillary electrophoresis, thin layer chromatography, normal-phase HPLC).

实验考虑。典型情形下，将已知数量的有关物质和空白制剂中的原料药加样，以配制已知有关物质浓度的准确度样品。根据ICH指南，准确度应在指定范围内，采用最少3个浓度水平进行最少9次检测。与线性测试（见第3.3.3部分）中所用的试验类似，有关物质和原料药可以直接称重，加到空白制 剂中，或采用贮备液逐级稀释。如果没有认可样品用于配制加标溶液，公司可以采用将原始方法与替代方法结果比较的方式来测试准确度（例如，毛细管电泳 、薄层色谱、正相HPLC）。

Intrinsic Accuracy. Intrinsic accuracy indicates the bias caused by sample matrix and sample preparation. In this approach, a stock solution is prepared by using known quantities of related substance and drug substance. The stock solution is further diluted to obtained solutions of lower concentrations. These solutions are used to generate linearity results. In addition, these linearity solutions of different concentrations are spiked into placebo. The spiked solutions are prepared according to the procedure for sample analysis. The resulting solutions, prepared from the spiked solution, are then analyzed. If the same stock solution is used for both linearity and accuracy and all of these solutions are analyzed on the same HPLC run, the response of linearity (without spike into matrix) and accuracy (with spike into matrix) can be compared directly. Any differences in response indicate the bias caused by matrix interference or sample preparation. To determine the intrinsic accuracy at each concentration level, one can compare the peak area of accuracy (with matrix) with that of linearity (without matrix) at the same concentration (Figure 3.11). This is the simplest approach, and one would expect close to 100% accuracy at all concentration levels.

固有准确度。固有准确度指由于样品成份和样品制备引起的误差。在此方法中，采用已知数量有关物质和原料药配制贮备液，再将贮备液稀释至低浓度。这些溶液用于检测获得线性结果。另外，这些不同浓度的线性溶液被加入空白制剂中，按分析方法配制加标溶液。如果采用了同一份贮备液进行线性和准确度测试，这些溶液在同一次HPLC运行中检测，线性响应（不加样至基底样）和准确度（加样至基底样）可以直接进行比较。响应中的任何差异都是由于矩阵干扰或样品配制误差引起的。为测试各浓度水平的固有准确度，公司可以比较相同浓度时准确度（有基底）与线性测试（无基底）的峰面积。这是最简单的方法，公司可以在所有浓度水平得到接受100%的准确度。

Overall Accuracy. In addition to the matrix effect and sample preparation error, method accuracy can also be affected by calculation error: for example, difference in relative response, significant y-intercept, and nonlinearity. Therefore, a more vigorous approach is to determine the overall accuracy, which incorporates the effect from all aspects of the method:

总体准确度。除了基底效应和样品配制误差之外，方法准确度还可能受到计算误差的影响。例如，相对响应差异、显著Y轴截距和非线性。因此，更好的方法是测试总体准确度，它是方法所有方面影响的综合表现。

? Matrix effect

基底效应

? Sample preparation

样品配制

? Calculation error

计算误差

In this approach, similar to the determination of intrinsic accuracy, the accuracy solutions are prepared by spiking a known quantity of authentic samples and drug substance into excipients. The accuracy solutions are then analyzed using HPLC.

这种方法与固有准确度相似，加入已知数量认可样品和原料药至辅料中，制备准确度溶液，采用HPLC对准确度溶液进行检测。

图3.11 固有准确度

The percent related substance results are calculated by the proposed method calculation procedure (e.g., high–low, area percent, external standard). Any correction factors are applied according to the procedure proposed. The overall accuracy is determined by the following equation (Figure 3.12):

根据建立的方法计算有关物质结果（例如，高低浓度法、面积百分比法、外标法），代入校正因子。采用以下公式计算总体准确度（见图3.12）

This is a more stringent approach, as this indicates the bias caused by matrix interference, sample preparation, and calculation. For example, related substance (found) = 1.20% and related substance (theory) = 1.40% (calculated from the weight of authentic sample used in the spiked solution); therefore,

这是一种比较严格的方法，因为这种方法显示出了由于矩阵干扰、样品制备和计算误差产生的偏差。例如，有关物质（实测）= 1.20%，有关物质（理论）=1.40% （从认可样品称样量计算），这时

3.3.5 Precision 精密度

Repeatability. ICH definition: Repeatability expresses the precision under the same operating conditions over a short interval of time. Repeatability is also termed intraassay precision.

重复性。ICH定义：重复性表示在同一操作条件下，较短时间段内的精密度。重复性也可以称为批内精密度。

Repeatability of a method can be determined by multiple replicate preparations of the same sample. This can be done either by multiple sample preparations (n = 6) in the same experiment or by preparing three replicates at three different concentrations. In general, one should evaluate results of individual related substances, total related substances, and the consistency of related substance profiles in all experiments. The percent RSD and confidence level of these results are reported to illustrate the method repeatability.

方法的重复性可以对同一样品平行制备进行多次检测。这时，可以采用同一试验多个样品溶液检测（n=6），或3个不同浓度各配制3份平行样。一般来说，公司要评估所有试验中单个有关物质、总有关物质结果，以及有关物质谱一致性。这些结果的RSD百分比和置信水平要进行报告，以说明方法的重复性。

Typically, an aged sample should be used to ensure that there are significant levels of related substance in the sample. Alternatively, if different samples are available with different levels of related substance (e.g., fresh sample and sample at expiry), one can determine the repeatability by performing three replicate preparations for each sample. ICH guidelines require a minimum of three samples with three different levels of related substance.

典型情况下，应使用一个放置过的样品，以保证样品中有关物质具有较显著的水平。如果可以获得具有不同有关物质水平的样品（例如新鲜样品和过期样品），公司也可以对每一样品进行3次平行制备来测试重复性。ICH指南要求至少在3个不同有关物质水平测试各3个样品。

图3.12 总体准确度（外标法）

Instead of using spike samples (as in accuracy determination), drug product lots that are representative of the commercial products should be used for precision (repeatability, intermediate precision). This is to ensure that the commercial drug product is used in at least one part of the method validation and that the repeatability results are representative of those that can be expected in the future.

在精密度测试中（重复性、中间精密度），应采用具有代表性的商业生产批次制剂，而不是加样样品（在准确度检测中所用）。这是为了保证方法验证中至少有一部分是采用了商业生产的制剂，重复性结果具有代表性，可代表对将来生产批次检验的结果。

Intermediate Precision. ICH definition: Intermediate precision expresses, within laboratories variations, different days, different analysts, different equipment, and so on.

中间精密度。ICH定义：不同化验室内，不同天，不同化验员，不同设备的中间精密度。

Intermediate precision is to determine method precision in different experiments using different analysts and/or instrument setup. Similar to that of repeatability, one should evaluate the results of individual related substances, total related substances, and the consistency of related substance profiles in all experiments.

中间精密度用于决定不同分析员和/或不同仪器时的方法精密度。与重复性相似，在所有试验中，公司要评估单个有关物质结果，总有关物质结果，有关物质谱的一致性。

The percent RSD and confidence level of these results are reported to illustrate the intermediate precision.

报告这些结果的RSD百分值和可置信水平，说明其中间精密度。

Reproducibility. ICH definition: Reproducibility expresses the precision between laboratories (collaborative studies are generally used, for standardization of methodology).

再现性。ICH定义：再现性表示的是不同化验室之间的精密度（对于标准方法，一般采用联合化验室研究）。

This is an optional validation parameter that requires demonstration of laboratory-to- laboratory variation only if multiple laboratories use the same procedure. The reproducibility data can be obtained during method transfer between laboratories.

这是一个可选的验证项，只有当不同化验室使用相同检验方法时才需要进行不同化验室比较。再现性数据可以在方法在不同化验室间进行转移获得。