Statistical methods quality control

Adverse event reporting Trial medication Premature withdrawal Subject replacement policy Criteria for excluding data Data analysis/statistical methods Quality control/assurance Data handling and record keeping Ethics (e.g. IRB/IEC approval)... 

The monitoring of analytical quality by the use of statistical methods and control charts. 

The FOCUS-PDCA/PMAIO Process Improvement Model is a statistical-based quality-control method for improving processes. This approach to problem solving could be used by all process action teams to ensure uniformity within an organization. FOCUS-PDCA/PMAIO is as follows ... 

Quality assurance and quality control are the unifying themes of this chapter as well as the previous one. Uni- and multivariate statistics provide the quantitative framework to measure reliability and uncertainty, two indispensable descriptors of any forensic data. Uncertainty and errors are inevitable in even the best-designed and validated method quality control and statistics provide the methods to characterize and incorporate those errors into data analysis and reporting. If problems arise during an anal5reis, quality assurance and quality control procedures and practices serve to raise a red flag and allow the forensic chemist to take corrective action. Thus, QA/QC are as much a part of forensic chemistry as are stoichiometry, organic chemistry, and sample preparation. 

Several methods have evolved to achieve, sustain, and improve quality, they are quality control, quality improvement, and quality assurance, which collectively are known as quality management. This trilogy is illustrated in Figure 2.1. Techniques such as quality planning, quality costs, Just-in-time , and statistical process control are all elements of... 

Control laboratories in the canned food industry are usually divorced from the research organization to a lesser degree than is the case in the chemical and allied industries. For this reason, a closer relationship exists between the problems of the control laboratory and the research laboratory. Although from a research standpoint this condition is often considered undesirable, it has considerable merit in the case of the canned food industry, in which production may be seasonal and often of rather short duration. The collection of control data in many instances may also serve for research purposes—for example, in the case of soil analyses, which may be correlated with agricultural research designed to improve crop yields. Because the variables which affect the quality of canned foods must usually be investigated rather extensively, and often over a period of more than one year, the application of statistical methods to data collected for control purposes can conceivably make a substantial contribution to a research program. 

Statistical process control (SPC) is an important on-line method in real time by which a production process can be monitored and control plans can be initiated to keep quality standards within acceptable limits. Statistical quality control (SQC) provides off-line analysis of the big picture such as what was the impact of previous improvements. It is important to understand how SPC and SQC operate. 

In passing we remark that there are well-known statistical methods of hypothesis testing and parameter estimation used in decisionmaking. Sequential analysis is a method of sampling used to decide whether to accept or reject a lot with defective items, or whether to continue sampling. Also, there are various statistical methods used in quality control of a manufacturing process, to decide on how much the quality should be improved to be acceptable. 

The Production Department was not amused, because lower values had been expected. Quality Control was blamed for using an insensitive, unse-lective, and imprecise test, and thereby unnecessarily frightening top management. This outcome had been anticipated, and a better method, namely polarography, was already being set up. The same samples were run, this time in duplicate, with much the same results. A relative confidence interval of 25% was assumed. Because of increased specificity, there were now less doubts as to the amounts of this particular heavy metal that were actually present. To rule out artifacts, the four samples were sent to outside laboratories to do repeat tests with different methods X-ray fluorescence (XRFi °) and inductively coupled plasma spectrometry (ICP). The confidence limits were determined to be 10% resp. 3%. Figure 4.23 summarizes the results. Because each method has its own specificity pattern, and is subject to intrinsic artifacts, a direct statistical comparison cannot be performed without first correcting the apparent concentrations in order to obtain presumably true... 

Of the various methods of data presentation, the one with which starting analysts may be least familiar is trend analysis and statistical quality control. In an industrial environment, analysis is often centered around the production of batches of material. The properties of those batches may change over time due to random effects or to subtle changes in the production process. In either case, the quality of the product may change. Analysis is used to track the change in the properties of batches over time. Industrial analytical methods, therefore, need to be extremely rugged. Millions of dollars may depend on the analyst s judgment as to batch equivalence. 

Ad-hoc approaches—Methods of estimating should be borrowed from other problems whenever applicable. For example, statistical techniques for quality control theory can probably be applied to chemicals by viewing discharges as "faulty" production. 

On the one hand, statistical quality control is an important tool for quality assurance within analytical chemistry itself (monitoring of test methods), and on the other for quality control of processes and products by means of analytical methods. 

Method validation provides information concerning the method s performance capabilities and limitations, when applied under routine circumstances and when it is within statistical control, and can be used to set the QC limits. The warning and action limits are commonly set at twice and three times the within-laboratory reproducibility, respectively. When the method is used on a regular basis, periodic measurement of QC samples and the plotting of these data on QC charts is required to ensure that the method is still within statistical control. The frequency of QC checks should not normally be set at less than 5% of the sample throughput. When the method is new, it may be set much higher. Quality control charts are discussed in Chapter 6. 

As shown above, these include a laboratory to be third-party assessed to international accreditation standards, to demonstrate that it is in statistical control by using appropriate internal quality control procedures, to participate in proficiency testing schemes which provide an objective means of assessing and documenting the reliability of the data it is producing and to use methods of analysis that are fit-for-purpose . These requirements are summarised below and then described in greater detail later in this chapter. 

Gunter, B., Brideau, C., Pikounis, B., and Liaw, A., Statistical and graphical methods for quality control determination of high-throughput screening data, J. Biomol. Screen., 8, 624, 2003. 

International Organization for Standardization (ISO), Statistical methods for quality control, Vol. 2, 4th Edition, Accuracy (trueness and precision) of measurement methods and results - Part 2 Basic method for the determination of repeatability and reproducibility of a standard measurement method, ISO 1994(E), 5725-2. 

Shewhart, W.A. (1939), Statistical Method from the Viewpoint of Quality Control, The Graduate School, The Agriculture Department, Washington, DC. 

Data have been collected since 1970 on the prevalence and levels of various chemicals in human adipose (fat) tissue. These data are stored on a mainframe computer and have undergone routine quality assurance/quality control checks using univariate statistical methods. Upon completion of the development of a new analysis file, multivariate statistical techniques are applied to the data. The purpose of this analysis is to determine the utility of pattern recognition techniques in assessing the quality of the data and its ability to assist in their interpretation. 

Statistical Control for a New Method To implement a new method, a laboratory must produce a preliminary track record of its success so that quality control charts can be established and then maintained. Aside from acquiring the space, supplies, equipment, instrumentation, and manpower required, the method must be tested, modified, tested again, etc., until it is ready to go "online." Gillis and Callio (listed in Bibliography) recommend the following sequence for preparing an instrumental method for routine use. 

Unlike SPC techniques, standard feedback control methods such as PID-control, do exert control upon a process, in an effort to minimize y, — yk. Control in Statistical Process Control is as such not regulatory control, but a semantic means of relating SPC to quality control—a means that often leads to the hybrid term SQC. Ogunnaike and Ray [14, Sec. 28.4] offer advice on when to use SPC and when to use standard feedback control methods When the sampling interval is much greater than the process response time, when zero-mean Gaussian measurement noise dominates process disturbances, and when the cost of regulatory control action is considerable, SPC is preferred. 

One approach for using DOE on more complex processes is to do the majority of the process development on smaller, representative sections of material, such as test panels, rather than on full-scale parts, and then to scale up with a more limited experimental matrix. There is no guarantee that experience on small-scale test panels will directly translate to large parts because dimensions and thickness of the part are important variables in their own right. Another way to save on costs is to start with a satisfactory process and to continue, via careful monitoring of process variations and results, to extend the range of experience. This method is variously called statistical process control or statistical quality control. 

SPC or statistical quality control (SQC) is similar to DOE in that it is a statistical, rather than mechanistic, method. Both SPC/SQC and DOE rely on the theory that there is a direct relationship between variations in process controls and resulting changes in product quality. In SPC, however, the experiments are not forced on the process like they are in DOE. The variations in product quality and the random process variations are traced over time instead. The variations in end product are then correlated, if possible, with changes in the process that have occurred during that time. SPC techniques are usually applied to the process after some baseline process has been established by other methods. 

The PAT guidance facilitates introduction of new measurement and control tools in conjunction with well-established statistical methods such as design of experiments and statistical process control. It, therefore, can provide more effective means for product and process design and control, alternate efficient approaches for quality assurance, and a means for moving away from the corrective action to a continuous improvement paradigm. 

Includes all information on analytical quality control, such a precision clauses (repeatability and reproducibility data), table of statistical data outlining accuracy (trueness and precision) of method... 

Neither one-point nor two-point calibrations have room to test the model or statistical assumptions, but as long as the model has been rigorously validated its use in the laboratory has been verified, these methods can work well. Typical use of this calibration is in process control in pharmaceutical companies, where the system is very well known and controlled and there are sufficient quality control samples to ensure on-going performance. 

---