Assuring of Test Results based on Statistical tools.pptx

for Physico chemical laboratory

1. Assuring of Test Results based on Statistical tools Sep 2022

2. The objective are: Understand statistical application in testing laboratories Select and apply the statistics required laboratory activities § Describe categories of statistical quality control (SQC). Explain the use of descriptive statistics in measuring quality characteristics

3. What is Statistics Statistics is an area of science concerned with the design of experiments or sampling procedures, the analysis of data and the making of inferences about a population of measurements from information contained in a sample . population is the set representing all measurements of interest to the sample collector. A sample is a subset of measurements selected from the population of interest. Statistics helps in studying various inferential procedures, in looking for the best predictor or decision making process for a given situation. Even more important, it provides information concerning the goodness of an inferential procedure. When predicting, it is important to know something about the error in such prediction.

4. INTRODUCTION Testing and calibration laboratories, including analytical chemical laboratories, are continually requested to provide evidence on the quality of their operations The general ISO definition of “ quality ” is given as “totality of characteristics of an entity that bears on its ability to satisfy stated and implied needs” Issues related to the reliability of data are often grouped under the general heading of “ QUALITY ASSURANCE and QUALITY CONTROL” (QA/QC) a description that captures the idea that data quality can not only be documented

5. Quality Assurance (QA) is that part of the management function that ensures the quality of results. It is a function that should be carried out to the extent necessary, no more and no less, and should be a fully integrated part of managerial duties on a daily basis. It is important to note that it involves not only getting the right answer but being able to prove the right answer has been obtained as well as maintaining documentary evidence of this. When introduced in the right way, QA leads to improved morale as people gain confidence in their results and pleasure at being able to prove the veracity of those results. Quality assurance focuses attention on relevant aspects of daily activities and training needs and helps staff to develop their skills and further their careers.

6. What is QUALITY ASSURANCE ( QA): QA describes the overall measures that a laboratory uses to ensure the quality of its operations. Typically this might include: A quality system Suitable laboratory environment Educated, trained and skilled staff Training procedures and records Equipment suitably maintained and calibrated Quality control procedures Documented and validated methods Traceability and measurement uncertainty Checking and reporting procedures ü Preventative and corrective actions Proficiency testing Internal audit and review procedures Complaints procedures Requirements for reagents, calibrants , measurement standards & reference materials

7. QUALITY CONTROL (QC) : ‘The operational techniques and activities that are used to fulfill requirements for quality’. Quality control procedures relate to ensuring the quality of specific samples or batches of samples and include: Analysis of reference materials/measurement standards Analysis of blind samples Use of quality control samples & control charts Analysis of blanks Analysis of spiked samples Analysis in duplicate Proficiency Testing/Inter-laboratory comparisons

8. Advantages of a quality assurance programme It provides a tracking record to ensure sample integrity, with documentation to verify that laboratory instruments are functioning properly and that laboratory data were generated according to approved written protocols. Saves analysis time and costs. Although the quality assurance programme may initially seem to reduce a laboratory's productivity, it may actually save analytical time and costs over a long period, since analyses would tend to be done correctly the first time. Aids in identifying training needs of analysts. This training would not be restricted to new employees; it would also apply to present employees whose performance may be deficient or needs updating. Increases in analyst confidence derived from knowing that results are reliable. This increased confidence, in turn, would lead to improved morale and performance. Ensuring errors are minimized or eliminated.

9. It is impossible to eliminate all errors but it is possible to ensure that very, very few serious errors are made without discovery before the results are transmitted outside the laboratory. Ensuring credibility. E.g. if the laboratory implement QA There is usually a strong legal tradition in the about the test applied to evidence in court. Give confidence to the management. e.g. If there is dispute, enquiry on the results from the customer the QA Ensures for any events of enquiry, dispute or error : records are available to resolve the issue. As per the QA Records should be kept for a considerable length of time. Five years is often chosen. Providing a review of deficiencies, errors and complaints so that remedial action can be systematic and lead to intrinsic improvement.

10. Components of Quality Assurance Documented commitment & policy Define laboratory purpose Staff structure that shows the responsibilities & roles of laboratory personnel to meet purpose. Identify someone with responsibility for QA / QC we call it QM. Must be free of internal/external pressures that could influence testing / reporting Sub-contracting tests – define criteria for selecting sub-contractors. List of approved sub-contractors. Criteria for, and list of approved suppliers. Process for addressing nonconforming testing – who, what, when?

11. Control of records – consistent, secure, retrievable, visible alterations, retention. Regular INTERNAL AUDITS need to check if we are doing everything that we “say” we are doing. (Must be documented and kept – evidence). Personnel - must SHOW that staff are competent (qualifications, experience, knowledge of the tests, able to operate equipment, do tests, etc ) – evidence. S taff training program – with personalized records Documentation : all major processes / procedures must be documented. Document control : only ONE version of ALL documents – no confusion, no mix-ups. (Can’t have conflicting processes).

12. Examples of documented procedures sampling ,sample receipt(sample register) test methods, Work instructions, SOPs, manuals ü quality control More documentation - FORMS – to record: Receipt of samples - SAMPLE REGISTER Analytical observations – can be visual or instrumental QC observations – e.g. temperatures, records Preparation of reagents & media (what, how much?) Receipt of consumables Audits findings Calibration checks (e.g. balances, thermometers, pH) Staff training, etc , etc ………. Can be ELECTRONIC - but how to personalise ? Need to demonstrate Proficiency testing INTRAlab - between staff INTERlab – between laboratories All outcomes must be recorded / documented

13. Laboratory environment must be suitable for its operations (e.g. temperature, humidity, tidiness) - equipment operations - storage of chemicals reagents incubation temperatures separate office work area no food / drinks

14. Internal quality control Internal quality control: QCs are used to measure accuracy, precision, contamination, and matrix effects. Generally, QCs are run per batch or set of samples at a frequency of 5% or one every twenty samples is recommended This level sufficiently demonstrates the validity of results. The laboratory determines, where feasible, the accuracy and precision of all analyses performed.

15. Blanks: either matrix or reagent, to determine and measure contamination and interferences ü Blanks should be below the method detection limit where possible. Blank results are evaluated and corrected where possible. If blank results are consistently above the method detection level (MDL) established, the MDL should be re-established. High blank results may also indicate contamination either from the solvent, laboratory equipment or laboratory environment. 2. Matrix spikes Matrix spikes measure the effects the sample matrix may have on the analytical method, usually the analyte recovery. Method accuracy is documented and controlled based on the percent recovery of matrix spikes for quantitative analysis and the positive response of the analyte for qualitative analysis.

16. .3. Duplicate samples or matrix spike duplicates It measures the precision of the analytical process. Duplicate analysis usually involves a replicate sample, subsampled in the laboratory, but for some methods it is in the form of a matrix spike duplicate. Method precision is documented and controlled based on the relative percent difference (RPD) or the positive response for qualitative analysis. 4. Quality control samples It measures the method performance. The matrix of the QCS should match the matrix of the samples being analyzed and should pass through the entire sample preparation process. The QCS, therefore, measures both the sample preparation process and the analytical process. 5. Standards ( CRM and RM) Calibration check standards referred to as initial calibration verification (ICV) and continuing calibration verification (CCV) are used to determine whether an analytical procedure is in control and stays within control. They are used to detect analytical method errors from procedural or operator errors or contamination from laboratory sources. 6. Accuracy and precision control charts

17. External Quality Control Program Participation in proficiency testing is an important means of quality control and assessing laboratory performance Proficiency surveys are used as a tool to assist personnel in the identification of laboratory problems that may exist and that have eluded the internal quality control program. Other Quality Control Monitoring Activities Replicate testing Retesting of retained items Correlation - Checking for correlation means evaluating the interrelated characteristics ( analytes ) of the sample. By comparing results from different analyses on the same test item, one checks for reasonableness (i.e. Does the data make sense or correspond as anticipated?).

18. statistical problem involves the: design of the experiment or, sampling procedure, collection and analysis of data, and making of inferences about the population based upon information in the sample

19. Error and Deviation; Mean and Standard Deviation The concepts of accuracy and precision can be put on a mathematical basis by defining equivalent terms: error and deviation . This will allow the understanding of somewhat more complicated statistical formulations used commonly in the testing laboratory. If a set of N replicate measurements x1, x2, x3,…, x n , were made (examples: weighing a vial N times, determining HPLC peak area of N injections from a single solution, measuring the height of a can N times, …), then:

20. DESCRIPTIVE STATISTICS

21. CONFIDENCE LIMIT/INTERVAL Defines the probability of finding the true mean within certain confidence interval as defined by the standard deviation Two important ways: Z-value Statistical t tables Z- Value is another way of understanding the normal distribution curve used for large numbers of samples, used to determine the confidence limit or interval How? Through realizing that the probability of finding the true mean is within certain confidence intervals as defined by the standard deviation. How is this calculation done? Look at the Z value from statistical tables after deciding the desired degree of certainty

22. Values for Z for Checking Both Upper and Lower Levels Degree of Certainty (Confidence) Z Value 80% 1.29 90% 1.64 95% 1.96 99% 2.58 99.9% 3.29__ Example : Calculate the confidence limit (or interval) for percent protein content of a certain product where measurement was done four times and the rresults are: 64.53% 64.45% 65.10% 64.78%

23. Note: Since this calculation is not valid for small numbers, So, assume we had run 25 samples instead of four.

24. OUTLIER TEST s Where: • Xi = suspected single outlier • Xava= mean of the measurements • Sd= standard deviation

25. . ERRORS

26. Significance tests

27. Comparison of two experimental means

28. Evaluation of proficiency tests

29. Measurement uncertainty Is the result of the evaluation aimed at characterizing the range with in which the true value of a measurand is estimated to lie. Uncertainty, on the other hand, takes the form of a range or interval ( eg 2.08±0.06), and, if estimated for an analytical procedure may apply to all determinations so described.   Component: Each of the separate contributions to uncertainty is referred to as an uncertainty component.   • Standard Uncertainty u(xi) : When expressed as standard deviation, an uncertainty component is known as a standard uncertainty.   • Combined Standard Uncertainty uc (y): Square root of the sum of the squares of all the standard uncertainties.   • Expanded Uncertainty (U): The multiple of uc (y) and coverage factor, k, which is approx. equal

30. DATA ANALYSIS IN MEASUREMENT UNCERTAINTY Testing laboratories shall have and shall apply procedures for estimating uncertainty of measurement. In certain cases, the nature of the test method may preclude rigorous, metro logically and statistically valid, calculation of uncertainty measurement. In these cases, the laboratory shall at least attempt to identify all the components of uncertainty and make a reasonable estimation, and shall ensure that the form of reporting of the result does not give a wrong impression of the uncertainty. Note 1: The degree of rigor needed in an estimation of uncertainty of measurement depends on factors such as: • The requirements of the test method; • The requirements of the customer; • The existence of narrow limits on which decisions on conformity to a specification are based.

31. Sources of MU Sampling and sub sampling Sample preparation Certified material used as part of the measurement system Equipment calibration Analysis and data acquisition

32. Data elaboration and rounding reference materials Information collected during method development and validation Results from methods quality control samples Inter-laboratory comparison of measurements Specifications from instrument manufacturers Theoretical models and literature information The personnel carrying out the tests Environmental conditions Uncertainty arising from correction of the measurement results for systematic effects

33. . MEASURNMENT UNCERTAINITY APPROACHES 6.1Type A (Top Down): Evaluation of components using statistical probability distributions of the results of a series of measurements and can be characterized by standard deviations of the respective distributions. Type A approach MU determination process; Specify Measurand. Quantify within laboratory Reproducibility. Quantify Bias. Convert components to standard uncertainties of all the same units or relative standard uncertainties. Calculate combined standard uncertainty. Calculate expanded uncertainty Combined Standard Uncertainty Expanded Uncertainty: at a 95% confidence level.

34. WHAT’S COVERED BY TYPE A? All components, except the reference material, are included in a Type A approach where the uncertainties associated with the Reproducibility and the Bias are evaluated using statistical data (usually from validation or verification trials). Precision and bias studies take into account the influence of equipment set-up, calibration, QC, environmental factors and personnel. The only external component that needs to be additionally included is the uncertainty associated with the Reference Material. WITHIN-LAB REPRODUCIBILITY QC chart history – stable standards over prolonged period of time covering the working range of the method. Validation/verification data for within-lab reproducibility. PT performance over a number of studies. 6.1.3 BIAS QC chart history CRMs Spiked or fortified samples Standard Addition PT Studies/Inter-laboratory studies Validation/Verification studies.

35. Type B (Bottom Up): Evaluation of components, and characterizing as standard deviations, by estimating their “assumed” probability distributions using: Previous measurement data; Experience with or general knowledge of the behaviour and properties of relevant materials or instruments; Manufacturer’s specifications; Data provided in calibration and other certificates; Uncertainties assigned to reference data taken from handbooks. Estimate all the contributing components to the measurement uncertainty Type B approach Measurement uncertainty process; Convert them to relative standard uncertainties. Combine the relative standard uncertainties as the square root of the sum of the squares of the relative standard uncertainties. Determine the expanded uncertainty using the relevant coverage factor.

36. Quality control true value setting for melting point of shortening product Physiochemical testing laboratory uses Quality Control (QC) samples to assure the quality of the data. The sample was prepared at Chemical laboratory and the true value of the QC sample is assigned after analyzing homogeneity and precision tests. Homogeneity Test The homogeneity of the QC honey sample was checked by analyzing the MELTING POINT. Four different sub samples were prepared and from each 2 duplicate (total 8) were analyzed by two analysts in reproducible conditions. The data are illustrated in table 1 below. Table 1: homogeneity check for MELTING POINT content in shortening product

37. homogeneity check and true value setting for MELTING POINT content in shortening product