11

Standardisation & Quality Assurance - Part 2

• This lecture covers aspects of standardisation and quality assurance, specifically focusing on validation reports, internal quality control, external quality assessments, and control charts.

Learning Objectives

• For validation reports, describe how to measure precision and accuracy.
• Define the principles of internal quality control and external quality assessments.
• Explain how to use control charts.

Validation Report - Assessment of Precision

• Precision is the closeness of agreement between independent test results under stipulated conditions.
• Measured as repeatability (r), intermediate precision (Rw), and reproducibility (R).
• Precision is hard to quantify so imprecision (inversely related) is commonly reported.
• The Coefficient of Variation (%CV) expresses imprecision as a percentage of the mean ( %CV = s / mean x 100).
• Precision data is needed for method/instrument validation for routine use.
• Assay precision determined over time using standardized procedures (internal quality control).

Assessment of Precision

• Repeatability (within-run or within-day): variability observed when laboratory, technician, days, instrument, and reagent lot are held constant; time between measurements kept to a minimum.
• Reproducibility: All factors varied; measurements carried out over several days.
• Intermediate precision (between-run or between-day): All factors except the laboratory are allowed to vary. (factors that changed should be stated in the validation report).

Example Procedure for Validation

• Collect samples with known high and low concentrations. Pool samples if necessary.
• Make 25 aliquots of each sample and store at -80°C.
• At days 1-5, measure 5 replicates of each sample. Days need not be consecutive.
• Insert data into an Excel file, with separate days in different rows.
• Calculate mean, standard deviation (SD), and %CV to assess repeatability and intermediate precision.

Evaluation of precision of methods

• Intra-assay precision: Analyze ~20 replicate samples/standards in one batch/run to get within-assay %CV. Repeat using replicate specimens of different concentrations.
• Intra-assay precision depends on the nature and concentration of samples, and the performance of the analysis at that time. It is not a definitive measure of precision.
• Inter-assay/inter-day precision: Select 3-5 samples/standards across the analytical range, and analyze one per day for 20 consecutive days. Calculate inter-day %CV from the 20 results.

Assessment of Accuracy

• Accuracy is the closeness of a measured value to a standard or known value.
• Calibration standards (calibrant): Quantitative measurements require standards. A solution containing a known amount of the component to be assayed.
• Primary standard: a substance where the concentration is determined by accurately weighing an amount and diluting to the appropriate volume/weight.
• Accuracy depends on the purity of the standard material and quality of preparation.
• Secondary standard: The concentration is determined using a known analytical method.
• Internal standard: A substance not normally present in the sample but is added to it in a known quantity, allowing clear differentiation from the test analyte.

Assessment of Accuracy (continued)

• Analytical methods are classified into 3 categories based on accuracy:
  • Definitive method: No known source of inaccuracy.
  • Reference method: After exhaustive comparison with a definitive method, it has negligible inaccuracy.
  • Method of known bias: Needs testing against a reference method. Methods with little/no bias are referred to as standard methods.
• Recovery experiments depend on determining a substance in identical samples, one with a known added amount. The difference is expressed as a percentage of the added amount (% recovery).
• Control materials (e.g., certified reference materials): Specimens with known values used for comparison.
• Statistical comparison of methods: The best method assesses accuracy by comparing results with a method of known accuracy (e.g., t-test or ANOVA).

Principles of IQC for immunoassay

• IQC samples should match routinely used samples in type, concentration, volume, and stability (sufficient for a reasonable time – e.g. 6 months).
• Commercially available IQC material is often expensive, but unassayed material may be available for precision studies; errors in reconstitution can occur.
• In-house preparation of pools can be hazardous (depending on the source); the addition of exogenous analyte or dilution may be needed.

Principles of IQC for immunoassay (continued)

• IQC procedures should reliably estimate intra-batch and inter-batch errors.
• Precision is defined as the agreement between replicate measurements (random error).
• Drift refers to progressive/non-random change in results due to e.g., temperature variations, unmixed reagents, or incubation times.

Basic IQC statistics

• Intra-batch precision: Measured by mean differences between repeat analytical controls (RAC) at the beginning and end of a batch or using standard curves at the beginning and end of the batch.
• Intra-batch drift: Measured by mean differences between repeat analytical controls (RAC) at the beginning and end of the batch.
• Inter-batch precision: Measured by the standard deviation between batches or using the standard deviation of IQC pools.
• Inter-batch drift: Measured by the current mean of IQC pools compared with previous means.

Precision Profile Plot

• Three applications:
  • Determining the working range of the assay.
  • Rejecting sample outliers in raw data.
  • Shape may reveal unsatisfactory batch quality.

Presentation & Interpretation of IQC data

• Quality control charts (Shewhart charts) provide visual displays for day-to-day analysis and forecasting trends over longer periods. Plot a measured quantity or statistic versus time.

Control Charts

• Any of the following statistics can be used for plotting control charts:
  • Results from control samples or the difference between a result and a mean/target value.
  • The mean of duplicate analyses of a control sample.
  • The standard deviation or %CV calculated for the batch, day, week, or month.

Control Charts (continued)

• Control limits are often set at +/- 2SD and +/- 3SD.
• Warning Limits: +/- 2SD – if a value falls outside limits, investigate method error.
• Action Limits: +/- 3SD – if a value falls outside limits, results need rejection. Investigate, rectify and repeat.
• Control charts should show randomly distributed points about the central mean.

Control Charts - Common Mistakes

• Misinterpreting random variation as a sign of a problem.
• Ignoring special causes (outliers).
• Incorrect control limits.
• Overcomplication of charts.
• Lack of training.

Operation of IQC programme

• IQC results evaluated by someone other than the analyst.
• Rejection of batches due to imprecision in techniques, instrumentation, or inaccuracy in standard preparation.
• Cumulative IQC data is reviewed to monitor long-term performance trends.
• Comparisons with external quality assessment schemes (EQAS).
• Review meetings.

Principles of EQAS

• Scheme Design: Frequent sample distribution, statistically valid data return.
• Accurate documentation of methods.
• Common units for reporting results.
• Assessed by participants in precisely the same way as routine specimens.

Principles of EQAS (continued)

• Sample Material: Identical behavior to real samples, appropriate analyte concentrations, stable, and free of infectious hazards.
• Definition of target values: Reference method values, validation of accuracy and stability.
• Assessment of performance includes appropriate statistics, overall and individual laboratory performance, and aspects like interference and interpretation.

Other aspects of QA in immunoassays

• Methods, protocols, and equipment: Protocol transfer from R&D to routine, adequate facilities such as incubators, plate washers, and calibrated instruments (pipettes, dispensers, microtitre plate readers, automated plate washers).

This quiz focuses on the critical elements of standardisation and quality assurance in laboratory settings. It covers validation reports, internal and external quality control principles, and the application of control charts. Test your understanding of precision, accuracy, and quality assessment techniques.