欢迎访问『 博普智库 』制药人必备知识工具
收录 被收录2次
质量管理 质量管理其他

歪翻书之:ECA OOE&OOT指南(3)

请注意本系列第一篇原文链接 及 感谢语。本系列能够最终成册,感谢 915_雨 在最后阶段的审核校对。仅以本文分享对 该文件的编写机构和专家 致以最诚挚的敬意。ECA OOE&OOT
评分 评分评分评分评分评分
阅读 513 收藏 2
手机端查看
使用微信 “扫一扫” 即可在手机上查看
请注意本系列第一篇原文链接 及 感谢语。本系列能够最终成册,感谢 915_雨 在最后阶段的审核校对。仅以本文分享对 该文件的编写机构和专家 致以最诚挚的敬意。ECA OOE&OOT指南(1)博普智库 (bopuyun.com)

Trend Analysis for Statistical Process Control

Overview

A control chart provides the simplest means of visuallytracking a processto identify trends.It consists of ahorizontal plot of an ongoing performancecharacteristic -- for example, analytical result for a particularparameter -- witha new data point added for each new measurement. Overlaid lines show evaluationcriteria such as allowed tolerances. The control charthighlights poor quality by showing when ameasurement lies outside the expectedvariation. More importantly, it shows when a process is trendingtoward failure.There are many different types of control charts, a number of these arediscussed in thisguideline.
控制图提供了直观跟踪过程以确定趋势的最简单方法。
它由性能特征持续的水平图组成--例如,特定参数的分析结果--为每个新测量添加一个新数据点。重叠线显示评估标准,如允许的公差。
控制图用测量值超出预期变化,来标记质量差。
更重要的是,它显示流程何时趋于失败。
有许多不同类型的控制图,本指南将讨论其中的一些。
As mentioned earlier,all measurements have variation. There are two types of variation.
如前所述,所有测量值都有变异原因,有两种类型的变异原因。

普通原因变异或噪音
2.Special Cause variation such as process shift,drift or excessive noise.
特殊原因变异,如工艺移位、漂移或过度噪音。

控制图的目的是检测特殊原因变异。对过程的期望是它处于统计控制之下,即变异的唯一来源是测试结果噪声(注:随机误差)。
Control of continuousdata

质量控制(QC)在制药和生物制药行业的相关流程中起着至关重要的作用。质量控制很大一部分侧重于监控工艺的持续性能,以发现问题或确定改进机会。一个理想的质量控制系统会在问题出现之前就标记出问题。有许多统计和图形技术用于监控持续的质量性能。
Under certain circumstances, if not investigated and or corrected, an OOT may lead in time to an OOS and thereforean identification of an OOT may be an early indicator of a potentialfuture OOS and so facilitate action being taken designed to reduce the incidence ofnon-random OOS results.
在某些情况下,如果不进行调查和/或纠正,OOT可能会及时导致OOS,因此,OOT的识别可能是未来潜在OOS的早期指标,因此有助于采取旨在降低非随机OOS结果发生率的措施。
These processes may only be effective wherethere is a suitable control strategy in place.
因此,趋势数据的生成是生产和质量管理问题的重要工具。只有在适当的控制策略到位的情况下,这些过程才可能有效。
A control strategyis a planned set of controls, derivedfrom current productand process understanding, thatensures process performanceand product quality. These controls can include parameters and attributes related to drug substance and drug product materialsand components, facility and equipment operating conditions, in-process controls, finishedproduct specifications, and the associated methods and frequency of monitoring and control.
控制策略是从当前产品和工艺理解中衍生出来的一组计划控制,以确保工艺性能和产品质量。这些控制可包括与原料药和药品材料和成分、设施和设备操作条件、过程中控制、成品规范以及相关监测和控制方法和频率相关的参数和属性。

当今制药行业产品质量和工艺性能生命周期的典型控制策略可能包括以下要素:

·In-Process Monitoring and control of ProcessPerformance Attributes
·Monitoring and control of Critical ProcessParameters linked to Critical Quality Attributes
·Controls for facility and equipment
·Monitoring the Drug Substance (API) andexcipients against purchasing specification
·Monitoring and trending of stability data forproduct and raw materials including the API
·了解和识别关键工艺参数
·对工艺性能属性的工艺内监测和控制
·监控与关键质量属性相关的关键工艺参数
·设施和设备的控制
·根据采购标准监控原料药(API)和赋形剂
·产品和原材料(包括API)稳定性数据的监测和趋势分析
An Out of Trent (OOT) result is a non-random event that is identified as test result or patternof results that are outside of pre-defined limits. For continuous dataevaluation, this guideline recommends using simple Shewhart type control charts in the firstinstance. These control charts developed in the 1930s have been widely applied in engineering and manufacturing industries.
超趋势(OOT)结果是一个非随机事件,被确定为超出预先定义限值的测试结果或结果模式。对于连续数据评估,本指南建议首先使用简单的休哈特型控制图。1930年代开发的这些控制图已广泛应用于工程和制造业。

控制图以适当方式收集的数据,应用于标准或基于历史数据的理想结果。任何控制图上的中心线都相当于统计期内收集的数值的平均值(平均的)。

一条(或多条)线位于中心线上方和下方,作为控制限值。这些限值,即上限控制限(UCL)和下限控制限(LCL),提供了一个可以接受的结果范围。因此,控制图用于确定即将出现的结果是否在可接受的范围内,或者过程是否失控。在可能的情况下,这些控制上限和下限必须基于为工艺的经验证可接受范围(PAR)和正常工作范围(NOR)确定的值。
In investigational circumstances it may berequired to analyse historical data to see if there have been special causevariations. In this instance a postmortem CuSum approach is to be recommended
在研究环境中,可能需要分析历史数据,以查看是否存在特殊原因变异。在这种情况下,建议采用事后CuSum方法
CuSum stands for "cumulativesum." A CuSum chart is related to a standard control chart and is made in much the same manner, except that the vertical axis isused to plot the cumulative sum of the variability (differences between successive values) ina process. This CuSum is plotted on the vertical (Y) axis against time on the horizontal (X) axis.
CuSum表示“累积总和”。CuSum图与标准控制图相关,其制作方式大致相同,但纵轴用于绘制过程中变异性(连续值之间的差异)的累积总和。即CuSum绘制在垂直(Y)轴上,时间绘制在水平(X)轴上。
(注:Minitab17帮助“累积和:显示每个样本值与目标值之间偏差的累积和”。)
This type of plot is helpful in spotting abiased process, in which the process misses the calculated mean value high or misses it low, since repeated misses onone side of the ideal value will force the cumulative sum away from the idealvalue or benchmark value (which may be zero ) which is the ideal low variance (no variance) objective.
这类的图有助于发现有偏过程,其忽略过程计算平均值的高或低,从理想值一侧的重复失败起,将迫使累积总和远离理想值或基准值(可能为零),这就远离了理想的低方差(无方差)目标。
The minimum number of data values fromwhich a suitable statistical mean can be calculated for use in a CuSum chart is 10 individual values. The maximumnumber of valuesto limit variationin the data set, should be set at 30 to 100 data values.
在CuSum图中,用于计算适当统计平均值的最少数量为10个单独值。要最大限度的描述数据集变异,应设置为30到100个数值。
This technique is discussed in detail in on page 44with a worked example in Appendix 4.
第44页详细讨论了该技术,附录4中有一个工作示例。
Determination of a Trend using Statistical ProcessControl (SPC)

统计过程控制(SPC)是一种使用统计方法和数据可视化展示的方法,使我们能够了解过程随时间的变化。通过了解过程变化的类型和大小,我们可以对过程进行改进,并预测这将带来更好的结果。SPC还可以用来确认我们的预测是否正确。这些方法是由休哈特(WalterShewhart)和戴明(WEdwards Deming)(以及其他人)在二十世纪上半叶发展起来的。

所有结果和过程的测量值会随着时间的推移而变化,但数据管理中的当前做法往往会隐藏这种变化,在这种情况下,数据是聚合(平均)的,并在较长的时间段内呈现(如按季度)。连续绘制数据(每周/每月)可以提供大量信息。如果我们这样做,我们就会揭示变异的来源和程度。

Control of continuousdata


Shewhart identified two sources of processvariation: common cause variation (chance variation) that isinherent in process,and stable over time, and special cause variation (assignable, or uncontrolled variation),which is unstable over time - the result of specificevents outside the system.
A process that is operating only withcommon causes of variation is said to be in statistical control. Aprocess that is operating in the presence of assignable causesis said to be outof control. The eventual goalof SPC is the elimination of variability in the process.
当处理的质量特性是变量时,我们希望确保该特性处于控制之下。
休哈特确定了过程变异的两个来源:过程固有且随时间稳定的共同原因变异(偶然变化)和随时间不稳定的特殊原因变异(可分配或不受控变化)——系统外特定事件的结果。
一个仅在共同变异原因下运行的过程称为统计控制。在存在可指定原因的情况下运行的过程称为失控。SPC的最终目标是消除过程中的可变性。

控制图的设计旨在区分过程中普通和特殊的变异原因,并提供一条规则,以最大限度地降低对特殊原因(事实上是常见原因)作出反应的风险,以及在存在特殊原因时不对其作出反应的风险。它对一组观测结果的中心趋势和分散性进行可视化观测。
A typical controlchart has controllimits set at values such that if theprocess is in control, nearly all points will lie betweenthe upper control limit (UCL) and the lower control limit (LCL).
A control chart is typically constructed as follows:
典型控制图的控制限值设置为,如果过程处于控制状态,几乎所有点都位于控制上限(UCL)和控制下限(LCL)之间。
控制图的构造通常如下所示:
(注:原文错误,把下控制限LCL写成了UCL) (1.2)
When the assignable causesare eliminated and the pointsplotted are withinthe control limits,the process
is in state of control.Further improvement can be obtainedthrough changing basic process, system. Depending on the datathan can be collected and on the purpose (detect small shift or large shift, investigation orcontinuous process verification), different control charts can be used. The following flowchart gives anindication of which chart to use when.
当可分配的原因被消除,且绘制的点在控制范围内时,过程处于控制状态。通过改变基本工艺、系统,可以得到进一步的改进。根据可收集的数据和目的(检测小班组或大班组、调查或连续过程验证),可使用不同的控制图。下面的流程图给出了何时使用哪个图表的指示。
歪翻书之:ECA OOE&OOT指南(3)

Figure 6: Control Charting selectionprocess
6:控制图选择流程
[redrawn & based on frontis illustration in D. C. Montgomery – Introduction to Statistical QualityControl) 6th Edition2009]
[重新绘制并基于蒙哥马利特区的frontis插图-统计质量控制导论)2009年第6]
I-Moving Range (MR) Control Charts
Individual control charts (or Shewhart control charts) are usedwhenever the sample size for process monitoring isn=1, for example one observation per batch. The moving range (MR) of two consecutive observations is usedas an estimation of process variability:
当过程监控的样本量为n=1时,使用单值的控制图(或休哈特控制图),例如每批观察一次。两次连续观测的移动极差(MR)作为过程可变性的估计:
过程平均值的估计值x̅为:
(1.4)
The Individuals chart control limits
(1.5)
The MR chart control limits
(1.6)
X-bar and R/S Control Charts
发布于 2022-04-26 13:54:53 © 著作权归作者所有
评分
评论
点赞
收藏
更多