《Method Validation方法确认.ppt》由会员分享，可在线阅读，更多相关《Method Validation方法确认.ppt（61页珍藏版）》请在三一办公上搜索。
1、Method Validation方法确认,2,Method Validation-Objectives方法确认 目标,By the end of this session you will be able to通过学习你将能够:Evaluate a new method with respect to what type of validation data are required获取相关确认数据，对新方法进行评估Assess validation data to determine specific requirements that must be met for the valida
2、tion to be verified and accepted评价确认数据，以确定方法确认应满足的特定要求已经验证并可接受,3,Presentation Outline报告讲述要点,Definitions and ISO 17025 requirements相关定义及ISO 17025 要求Validation process方法确认过程Examples of validation procedures方法确认举例Parameters to be calculated参数计算LOD,LOQ,precision,accuracy,uncertainty检出限定量限,精密度,准确度,不确定度QA
3、 verification of data数据验证的质量保证,4,What is a Method何谓方法?,An analytical method is to be regarded as the set of written instructions completely describing the procedure to be adopted by the analyst in order to obtain the required analytical result.分析方法是一套书面说明,完整描述分析人员为获得所需要的分析结果所采用程序。11Wilson,A.L.“The P
4、erformance Characteristics of Analytical Methods-I”(1970)Talanta 17,21-29,5,What is Method Validation何谓方法确认?,Validation is the process of determining the suitability of methodology for providing useful analytical data.方法确认是确定方法能够提供有效分析数据的过程。22 Taylor,J.K.“Validation of Analytical Methods”(1983)Anal.
5、Chem.55,600A-608A,6,ISO 17025:,Validation is the confirmation by examination and the provision of objective evidence that the particular requirements for a specific intended use are fulfilled.方法确认是通过检查和提供客观证据确证某一特定用途的特殊要求已得到满足33 ISO 17025 Standard,element 5.4.5.1,7,Objective of Method Validation方法确认
6、的目标,应具有合理设计的,充分研究的,形成完整文件的分析方法能为既定领域提供有用的数据,To have properly designed,adequately studied,well documented analytical methodscapable of producing useful data for theintended application,8,Sequence of Events方法确认工作的顺序:,The process is not necessarily linear it may be necessary to cycle back through metho
7、d development at least once before the method can be successfully validated.方法确认不一定是直线进程 它可能需要随着方法的开发至少重复一次直到方法成功确认.,Method Development方法改进,Method Validation方法确认,Method Implementation方法执行,9,General Classification of Method Validation Studies方法确认的分类,Interlaboratory(between laboratories)实验室间At least t
8、wo independent laboratories involved至少两个独立实验室Sponsored by professional organizations(AOAC,ASTM)由专业组织承办AOAC Official Methods官方方法 at least 8 labs至少8个实验室AOAC Peer Verified Methods同行验证方法 at least 2 labs至少2个实验室Intralaboratory(within a laboratory)实验室内,10,Benefits of Method Validation方法确认的益处,Assessment and
9、 optimization of method performance方法性能的评价和优化 Generation of baseline data for comparison to long-term QC data生成基准数据与长期的质控数据进行比较Identification of instruments,other equipment,supplies and chemicals required识别分析仪器、其它设备、耗材和化学品的需求Assessment of purchasing needs,cost and potential suppliers评价购买需求、价格和可能的供应商
10、,11,Benefits of Method Validation方法确认的益处,Identification of personnel needs个人需求的识别 Expertise and training专业技能和培训Consideration of safety and hazardous waste disposal issues对实验安全和有害废物处理的考虑Demonstration of the laboratorys ability to meet the requirements of the customer实验室对消费者需求的满足能力的展示,12,Benefits of M
11、ethod Validation方法确认的益处,Operational efficiency when method is implemented方法执行后的操作效率Design of QA/QC strategy QA/QC方案设计Avoid embarrassment and damage to labs credibility when the method is implemented避免方法执行后造成的对实验室信誉的不良影响Allows initial estimation of measurement uncertainty对测量不确定度进行初步评估,13,When is Meth
12、od Validation Necessary何时需要方法确认?,Official methods(eg.USP,AOAC,Health Canada Compendium,etc.)do not require full internal validation,however the laboratory must官方方法不需要完整内部确认，不过实验室必须:Have procedures to demonstrate that the method is performing in the particular laboratory environment有程序证明在特定实验室环境下该方法可
13、行.Retain documented evidence that the method is performing as expected and the lab has accepted the performance保留有关文件证明所采用的方法在实验室可行.Ensure that the most current method is available(eg.subscribe to an update service,or use other means to ensure the most current version of the method is available to t
14、he analyst).确保有最新有效的方法（例如定购更新服务，或使用其它方法确保分析人员能得到最新有效版的方法）VERIFICATION验证,14,When is Method Validation Necessary何时需要方法确认?,Laboratory-developed(“in-house”)methods and modified official methods must be validated in order to comply with ISO 17025实验室内部制定的方法和改进的官方方法必须经过确认,以符合ISO 17025,15,The Validation Pro
15、cess方法确认过程,What degree of validation is required确认的程度Depends on several factors,including由几项因素决定，包括:customer needs顾客需求intended use of the data数据的用途type of sample样品类型regulatory requirements and法规要求whether a standard reference method is available for comparison是否标准参考方法能够进行对比,16,The Validation Process方
16、法确认过程,Careful planning of the validation process is essential方法确认过程需认真设计，这很关键:Often one set of analyses can be used to determine several parameters常常一组分析可测定几个参数.Need to ensure that all required parameters can be calculated from the data generated需确保所有所需的参数能够通过获得的数据计算出来.,17,Selecting an Appropriate M
17、ethod选择一个合适的方法,Information required to determine if reference method is appropriate通过各种信息来决定方法是否合适:What are the analytes of interest分析对象是什么?What is the sample matrix样品基体是什么?What is the concentration range of interest分析对象的浓度范围?May not be known可能不知道,18,Selecting an Appropriate Method选择一个合适的方法,Informat
18、ion required需要以下信息:What are the qualitative and quantitative needs定性和定量要求?What level of detection and quantitation are required检出限和定量限是多少?Are the samples/analytes stable样品或被分析物稳定是否稳定?,19,Qualitative Needs定性要求,Are analytes predictable分析对象可预见吗?Is concentration predictable浓度可预见吗?Is matrix composition p
19、redictable基体成分可预见吗?Are matrix blanks available and representative基体空白可测定?具有代表性吗?,20,Quantitative Needs定量要求,Different needs for different applications不同应用的不同需求:Conformity to upper and lower certified limits最高和最低限的确认 Formulation or product assay公式表达或产物鉴定High precision,un-biased高精度，无偏差,21,Quantitative
20、Needs定量要求,Conformity to tolerance levels or regulatory thresholds与允许量水平或极限量符合性Contaminants,impurities,degradation products,regulated chemicals污染物、杂质、降解产物,法规中规定的化学品High precision,unbiased at threshold在极限水平上的高精度、无偏差,22,Quantitative Needs定量要求,Chemical behavior studies化学行为研究Degradation studies,stability
21、 studies,environmental fate studies降解、稳定性和环境归属研究Usually requires high sensitivity常需高灵敏度Trace level measurements痕量测定Adequate detection and quantitation levels适当的定性定量水平,23,Preparing for a Method Validation Study方法确认研究的准备,Necessary to consider the following须考虑:Type of samples to be used to test the per
22、formance characteristics of the method测试方法性能特性所用样品类型:Reference standards参考标准CRMs(Certified Reference Materials)有证参考物质Matrix blanks基体空白Positives(appropriate matrices)阳性样品（匹配基体）Other types of control samples其它类型控制样品,24,Preparing for a Method Validation Study方法确认研究的准备,Chemicals and Supplies化学试剂和耗材:Sour
23、ce(s)来源Cost费用Quality质量Quantities available可得到的数量Special handling or storage requirements特殊处理和保管要求?Special training required for analysts对分析人员的专门培训?Waste disposal废物处理,25,Preparing for a Method Validation Study方法确认研究的准备,Equipment/Instruments仪器/设备:Source(s)来源Cost价值Space,maintenance and training require
24、ments空间、维护和培训要求Service and support维修和技术支持Availability of consumables易耗品,26,Preparing for a Method Validation Study方法确认研究的准备,Method Validation SOP方法确证的标准操作程序Available to staff performing the validation负责方法确认的人员可以得到Up to date及时更新Descriptive 叙述Includes technical references,as appropriate适当时,包括专业参考文献,27
25、,Personnel人员,Senior R&D Chemist responsible for method validation高级研发人员负责方法确认Interpretation and assessment of technical literature对专业文献的解释和评价Knowledge of QA/ISO requirements QA/ISO 相关知识Able to critically review the data(unbiased)审查数据（无偏见）,28,Personnel人员,Laboratory Analyst实验室分析人员Understand customer r
26、equirements知晓顾客要求Understand study design and performance characteristics being investigated理解目前研究的特点和相应设计Appropriate laboratory skills适当实验室技能Set-up,optimization,troubleshooting实验准备、优化、疑难解决Observant,good documentation skills遵守职责及规章制度，做好完备记录Able to assess data quality能够评估数据质量,29,Scheduling Considerati
27、ons计划中的注意事项,Anticipate difficulty预见困难Be prepared to do troubleshooting准备解决困难If problems,modify the method to improve the data 如遇问题，改进方法提高数据质量dont continue to repeat analysis with poor method until sufficient amount of adequate data has been obtained在得到充分的数据前,不必大量重复比较差的方法,30,Validation of Food Analys
28、is Methods食品分析方法的确认,Suggest that guidance for validation of a method for food analysis be consistent with the validation strategy used by AOAC International.对食品分析方法的确认建议参照AOAC International使用的确认规则,31,Validation Procedure AOAC确认程序 AOAC,AOAC method validation includes determination of包括:Selectivity选择性
29、Linearity线性limit of detection(LOD)检出限limit of quantitation(LOQ)定量限Recovery回收率Precision精密度Accuracy准确度,32,Validation Procedure AOAC确认程序 AOAC,Must provide evidence of必须提供:Analyte identification被分析物的识别Separation of analyte from interfering substances将被分析物从干扰物质中分离Applicability of the method for measuring
30、 the analyte in the particular matrix在特殊基体中该法测定被分析物的能力Typically,a minimum of 5 analyses must be conducted on different days 特别要注意，最少5个分析结果必须在5个不同的工作日完成,33,Selectivity选择性,Definition定义:The method responds primarily to the analyte of interest and is little affected by other substances 方法主要对被分析物响应，其它物质的
31、影响很小May be demonstrated using reference standards,blanks,and spikes可用参考标准、空白和加标论证:Identification by comparison to behavior of reference standard与参考标准的行为比较进行识别Most conclusive if matrix blank is available基体空白更具说服力,34,Selectivity选择性,Blank空白:Depending on the type of blank,provides different information
32、on sources of interferences and contamination根据不同类型的空白，可给出不同种类的干扰和污染的信息Field blanks场地空白Trip blanks路途空白Matrix blanks基体空白Laboratory reagent blanks实验室试剂空白Instrument blanks 仪器空白,35,Selectivity选择性,For method validation,a variety of blanks should be assessed 在方法确认中，需要对一系列的空白进行评估Identify and control source
33、s of interferences识别和控制干扰来源The purpose of the blanks and the information that will be acquired from analyzing blank data should be well defined and understood空白的目的和空白数据分析信息的意义要实现准确定义和理解,36,Selectivity选择性,Qualitatively demonstrating selectivity from blank and spike data从空白和加标数据定性证明选择性:Compare the dat
34、a(blank vs.spike)数据比对（空白与加标）Ideally,no analyte detected in blank理想条件下，空白中不得检出被测物If interference present identify source and eliminate如果干扰存在，找到来源并去除If source is sample matrix modify method to eliminate interference如果来源于样品基体-改进方法排除干扰,37,Selectivity选择性,Supporting qualitative data may be provided by定性分析
35、数据的提供:Spectral data(UV,MS,fluorescence,IR,atomic emission,etc.)光谱数据（紫外、质谱、荧光、红外、原子发射，等等）Elemental analysis元素分析,38,Selectivity选择性,Alternate methods or techniques for qualitative confirmation定性确证的其它方法或技术:Different chemical separation techniques各种化学分离技术Different detection techniques各种检测技术Different samp
36、le preparation techniques各种样品制备技术,39,Calibration and Linearity校准和线性,Calibration curve:y=mx+b(linear)线性校准曲线：y=mx+bDetermination of the response profile of an instrument over a range of analyte concentrations using a suitable solvent or sample matrix使用适当的溶剂或样品基体测定被测物的浓度范围内的分析仪器响应值域,40,Calibration and
37、Linearity,contd:校准和线性(续),Calibration curve is plotted over the expected range of the analyte in the validation matrix在确认基体中期望的浓度范围内绘制校准曲线Requirements for linearity may be specified by the customer or in the reference method线性要求可能由客户规定或参考方法制定Generally the regression coefficient(r)must be 0.99 or 0.99
38、5一般回归系数(r)必须 0.99 或 0.995,41,Detection and Quantitation Levels定性和定量水平,Limit of detection:LOD检出限Limit of quantitation:LOQ定量限There are numerous definitions and approaches for estimating these levels有很多定义和方法评估这两个量LOD and LOQ are important parameters for residue methods在残留分析中LOD 和 LOQ 是重要参数,42,LOD Deter
39、mination:Example 1例一：LOD测定,Common approach for residue analysis:农残分析常用方法spike 5 portions of blank sample at the estimated LOD在估计的LOD水平上测定5个加标样品Calculate the standard deviation,and multiply by 3 to get the LOD.计算标准偏差，乘以3，得到LODMust be able to show that the peaks are identifiable at this calculated LOD
40、(ie.Example chromatogram)必须能够证明在计算出的LOD 值水平上色谱峰能够识别（例如气象色谱图）,43,LOQ Determination:Example 1例一：LOQ测定,The example given for LOD determination may also be used to calculate the LOQ LOD测定的例子也可用于LOQ的计算Multiply the standard deviation by 10 to get the LOQ 标准偏差乘以10即LOQMust also be able to show that the prec
41、ision is within a reasonable range(typically 20%)*at the LOQ LOQ水平的精密度应在合理范围内(典型的 20%-此值可能在参考方法中或由客户规定）*(this value may be specified in the reference method or by the customer),44,LOD and LOQ:Example 2例二：LOD 和 LOQ,Signal-to-noise ratio calculations信噪比计算Not based on statistics非统计学原理Example:Chromatogr
42、aphy data如色谱数据The“Signal”refers to the magnitude of the analyte signal above the background”信号”是指大于背景的被测物信号值“Noise”refers to the fluctuations in the baseline of the chromatogram at or near the retention time of the analyte”噪声”是指被测物保留时间附近色谱基线的波动,45,LOD and LOQ:Example 2例二：LOD 和 LOQ,The signal-to-nois
43、e(S/N)ratio generally must be信噪比通常必须:Greater than 3:1 at the LOD 在LOD水平上大于3:1 Greater than 10:1 at the LOQ 在LOQ水平上大于10:1 The ratios may differ depending on the method and application根据分析方法和应用的差异该比值是不同的,46,LOD and LOQ Determination LOD 和LOQ 的测定,Choose the method of determination most appropriate for
44、the application选择最适宜的方法Document the definition and approach used记录定义和方法If working under regulatory guidelines(eg.EPA,FDA),the approach prescribed for the program needs to be followed如果按照EPA,FDA等法规指南定义进行，须遵守其规定,47,Precision and Accuracy精密度和准确度,May be determined by analysis of reference materials,if a
45、vailable 如果有,可以通过参考物质的分析进行测定If reference materials not available若无参考物质:Determine precision and accuracy using matrix blanks and fortified(spiked)matrix blanks通过基体空白和添加基体空白测定精密度和准确度,48,Preparation of Spikes加标样品的制备,Add a known quantity of the analyte(s)to the matrix blank prior to sample preparation在样
46、品预处理前，将已知量的被测物加入的基体空白中Treat exactly like a sample according to procedures in the method按照方法中的程序将其视为被测样品进行预处理和测定,49,Critical Points Regarding Spikes加标的关键点,Spiking procedure must be very precise and accurate加标过程必须准确Document steps and techniques used for spiking加标步骤和技术形成文件Spike at the right step:typica
47、lly prior to any sample extraction steps按正确步骤加标：应在样品萃取之前加标Spiking may occur at a variety of steps during method development or validation study to aid in troubleshooting recovery problems 在方法开发或确认研究时，加标可在不同步骤中加入，来帮助解决回收率的问题,50,Spike Recovery Data加标回收数据,The assumption is made that the spiked sample i
48、s chemically and physically representative of the actual samples加标样品在理化性质上假设与实际样品相同Spike recovery data do not necessarily reveal the true recovery of the analyte from the sample加标回收数据不必反映样品被测物的实际回收率,51,Precision and Accuracy精密度和准确度,A minimum of 5 recovery determinations at each of 5 calibration leve
49、ls must be determined.在5个校准水平下必须每个水平至少测定5个加标回收样品One set of recovery determinations must include a blank,and 5 or more non-zero spikes,including the LOQ as the low level.每组加标样品必须包括空白、5个以上非零加标，其中低值应为LOQ,52,Precision and Accuracy精密度和准确度,The linear range of the method,as well as precision and accuracy a
50、re calculated from these determinations.通过这些测定得到方法线性范围，以及精密度和准确度。Intra-batch(within-batch)accuracy and inter-batch(between batch)accuracy may be compared as a measure of robustness.批内和批间准确度可认为是方法稳健度的测定,53,Accuracy准确度,The accuracy required for a particular method may be specified in the reference met
